Sign in with
Sign up | Sign in
Your question

Help - TZ Clock Optos

Last response: in Video Games
Share
Anonymous
July 13, 2005 4:54:51 PM

Archived from groups: rec.games.pinball (More info?)

Ok, before I get hate mail, let me tell you that I know this subject has
been discussed to death before. But I just can't find a good explanation to
my problem. Please help me!

My clock's been broken for the longest and I need it fixed now. I made my
own minutes board (with the extra diodes) to replace the old burnt board.
Replaced all the optos with OPB804 (I know it's not supposed to be a good
replacement part). These actually work wonderfully as the hour optos but
just not as the minutes. Well actually that's not completely true, they do
work when I run the clock test and trigger the optos with a piece of paper,
I can see all 8 working well. Once I put the clock back together and run the
clock test (doing forward), the 15 and 30 minute optos keep on flickering
all the time for no reason. So what's the deal here? Why aren't the hour
optos doing the same? I thought it might have been RF interference from the
motor so I tried a ceramic cap on it, no luck. Other problem, I don't quite
understand the thing with this sensitivity problem, I can't even find a
friggin datasheet with graphs about this component (anybody has it?). Why is
it that the .0086 sensitivity is required? How will the circuit differ if I
use the OPB804 (electrically)? What I figure is that the phototransistor
will trigger easier than with lower sensitivity, but that doesn't quite make
sense since my OPB804 is triggerring all the time for no reason (I thought
it had a lesser sensitivity...). I heard those 339s die all the time, could
this possibly be the case?

Any comment or answer is highly appreciated!
Thanks
Nic

More about : clock optos

Anonymous
July 13, 2005 4:54:52 PM

Archived from groups: rec.games.pinball (More info?)

On Wed, 13 Jul 2005 12:54:51 -0400, "Nicolas Legare"
<no_email_please@hotmail.com> wrote:

>Ok, before I get hate mail, let me tell you that I know this subject has
>been discussed to death before. But I just can't find a good explanation to
>my problem. Please help me!
>
>My clock's been broken for the longest and I need it fixed now. I made my
>own minutes board (with the extra diodes) to replace the old burnt board.
>Replaced all the optos with OPB804 (I know it's not supposed to be a good
>replacement part). These actually work wonderfully as the hour optos but
>just not as the minutes. Well actually that's not completely true, they do
>work when I run the clock test and trigger the optos with a piece of paper,
>I can see all 8 working well. Once I put the clock back together and run the
>clock test (doing forward), the 15 and 30 minute optos keep on flickering
>all the time for no reason. So what's the deal here? Why aren't the hour
>optos doing the same? I thought it might have been RF interference from the
>motor so I tried a ceramic cap on it, no luck. Other problem, I don't quite
>understand the thing with this sensitivity problem, I can't even find a
>friggin datasheet with graphs about this component (anybody has it?). Why is
>it that the .0086 sensitivity is required? How will the circuit differ if I
>use the OPB804 (electrically)? What I figure is that the phototransistor
>will trigger easier than with lower sensitivity, but that doesn't quite make
>sense since my OPB804 is triggerring all the time for no reason (I thought
>it had a lesser sensitivity...). I heard those 339s die all the time, could
>this possibly be the case?
>
>Any comment or answer is highly appreciated!
>Thanks
>Nic
>
Here is the datasheet for the original I had it posted in an old post with a bad
link so I'll repost it for the Google record.

http://mywebpages.comcast.net/eavedesian/opto/DSCF0143....
http://mywebpages.comcast.net/eavedesian/opto/DSCF0144....
http://mywebpages.comcast.net/eavedesian/opto/DSCF0145....
http://mywebpages.comcast.net/eavedesian/opto/DSCF0146....




--
Eric A.
eavedesianNOSPAM@comcast.net
313-268-0541

Mr. Propane Torch!!

ELV!S,TZ,Cy(l*ne,FH,STTNG,RFM,BH,HH,HS,HS2,CV,B@by-P@c M@n,IJ,RS,BK
Remove NOSPAM to reply
July 13, 2005 5:00:30 PM

Archived from groups: rec.games.pinball (More info?)

Put the problem to bed with these:

http://www.pbliz.com/id31.htm

If you want plug and play, send the assembly and Joel will send it back 100%
plug and play.



Otto

CARGPB11

My web page: http://home.bellsouth.net/p/PWP-Ottoslanding

"Nicolas Legare" <no_email_please@hotmail.com> wrote in message
news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
> My clock's been broken for the longest and I need it fixed now.
Related resources
Can't find your answer ? Ask !
Anonymous
July 13, 2005 5:04:06 PM

Archived from groups: rec.games.pinball (More info?)

Thanks for the tip, but 120+Ship is a bit too much for me right now, if I
can figure out a way to get away with a couple optos, I'd be much happier.
And to be honest, I don't quite like the blue LED look.

Thanks
Nic

"Otto" <ottondebremove&%$*@bellsouth.net> wrote in message
news:ELbBe.96291$du.69737@bignews1.bellsouth.net...
> Put the problem to bed with these:
>
> http://www.pbliz.com/id31.htm
>
> If you want plug and play, send the assembly and Joel will send it back
100%
> plug and play.
>
>
>
> Otto
>
> CARGPB11
>
> My web page: http://home.bellsouth.net/p/PWP-Ottoslanding
>
> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
> > My clock's been broken for the longest and I need it fixed now.
>
>
Anonymous
July 13, 2005 5:04:07 PM

Archived from groups: rec.games.pinball (More info?)

How many hours do you have in this project so far? I don't know about your
time, but clearly it is cheaper to go with a proven superior solution and
fix it once.
Go with pin liz and FAGETABOUTIT

--

John Arriola

"Nicolas Legare" <no_email_please@hotmail.com> wrote in message
news:aQbBe.2888$p5.1010@nnrp.ca.mci.com!nnrp1.uunet.ca...
> Thanks for the tip, but 120+Ship is a bit too much for me right now, if I
> can figure out a way to get away with a couple optos, I'd be much happier.
> And to be honest, I don't quite like the blue LED look.
>
> Thanks
> Nic
>
> "Otto" <ottondebremove&%$*@bellsouth.net> wrote in message
> news:ELbBe.96291$du.69737@bignews1.bellsouth.net...
>> Put the problem to bed with these:
>>
>> http://www.pbliz.com/id31.htm
>>
>> If you want plug and play, send the assembly and Joel will send it back
> 100%
>> plug and play.
>>
>>
>>
>> Otto
>>
>> CARGPB11
>>
>> My web page: http://home.bellsouth.net/p/PWP-Ottoslanding
>>
>> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
>> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
>> > My clock's been broken for the longest and I need it fixed now.
>>
>>
>
>
Anonymous
July 13, 2005 6:30:30 PM

Archived from groups: rec.games.pinball (More info?)

Nicolas,

You cannot assume the OPB804's will work in any WMS application. The
OPB804s are rated at 0.5mA of current in the detector side. WMS tries to
pump 12.0mA into there. WMS actually had Quality Technologies (QT) do a
special fabrication which was stamped .0086. If you were to change your
OPB804s to QVE1223.0086 you problem will go away!

Jim Knight - Who is painfully aware of the differences between OPBs and
QVEs.

"Nicolas Legare" <no_email_please@hotmail.com> wrote in message
news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
> Ok, before I get hate mail, let me tell you that I know this subject has
> been discussed to death before. But I just can't find a good explanation
> to
> my problem. Please help me!
>
> My clock's been broken for the longest and I need it fixed now. I made my
> own minutes board (with the extra diodes) to replace the old burnt board.
> Replaced all the optos with OPB804 (I know it's not supposed to be a good
> replacement part). These actually work wonderfully as the hour optos but
> just not as the minutes. Well actually that's not completely true, they do
> work when I run the clock test and trigger the optos with a piece of
> paper,
> I can see all 8 working well. Once I put the clock back together and run
> the
> clock test (doing forward), the 15 and 30 minute optos keep on flickering
> all the time for no reason. So what's the deal here? Why aren't the hour
> optos doing the same? I thought it might have been RF interference from
> the
> motor so I tried a ceramic cap on it, no luck. Other problem, I don't
> quite
> understand the thing with this sensitivity problem, I can't even find a
> friggin datasheet with graphs about this component (anybody has it?). Why
> is
> it that the .0086 sensitivity is required? How will the circuit differ if
> I
> use the OPB804 (electrically)? What I figure is that the phototransistor
> will trigger easier than with lower sensitivity, but that doesn't quite
> make
> sense since my OPB804 is triggerring all the time for no reason (I thought
> it had a lesser sensitivity...). I heard those 339s die all the time,
> could
> this possibly be the case?
>
> Any comment or answer is highly appreciated!
> Thanks
> Nic
>
>
Anonymous
July 13, 2005 7:37:05 PM

Archived from groups: rec.games.pinball (More info?)

Hmmmm ok, any ideas why they work on the hours side then??? I could probably
just add a transistor to each of the minutes optos to give 12mA to the
LM339... if I understand what you are saying...?

Thanks
Nic



"Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
> Nicolas,
>
> You cannot assume the OPB804's will work in any WMS application. The
> OPB804s are rated at 0.5mA of current in the detector side. WMS tries to
> pump 12.0mA into there. WMS actually had Quality Technologies (QT) do a
> special fabrication which was stamped .0086. If you were to change your
> OPB804s to QVE1223.0086 you problem will go away!
>
> Jim Knight - Who is painfully aware of the differences between OPBs and
> QVEs.
>
> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
> > Ok, before I get hate mail, let me tell you that I know this subject has
> > been discussed to death before. But I just can't find a good explanation
> > to
> > my problem. Please help me!
> >
> > My clock's been broken for the longest and I need it fixed now. I made
my
> > own minutes board (with the extra diodes) to replace the old burnt
board.
> > Replaced all the optos with OPB804 (I know it's not supposed to be a
good
> > replacement part). These actually work wonderfully as the hour optos but
> > just not as the minutes. Well actually that's not completely true, they
do
> > work when I run the clock test and trigger the optos with a piece of
> > paper,
> > I can see all 8 working well. Once I put the clock back together and run
> > the
> > clock test (doing forward), the 15 and 30 minute optos keep on
flickering
> > all the time for no reason. So what's the deal here? Why aren't the hour
> > optos doing the same? I thought it might have been RF interference from
> > the
> > motor so I tried a ceramic cap on it, no luck. Other problem, I don't
> > quite
> > understand the thing with this sensitivity problem, I can't even find a
> > friggin datasheet with graphs about this component (anybody has it?).
Why
> > is
> > it that the .0086 sensitivity is required? How will the circuit differ
if
> > I
> > use the OPB804 (electrically)? What I figure is that the phototransistor
> > will trigger easier than with lower sensitivity, but that doesn't quite
> > make
> > sense since my OPB804 is triggerring all the time for no reason (I
thought
> > it had a lesser sensitivity...). I heard those 339s die all the time,
> > could
> > this possibly be the case?
> >
> > Any comment or answer is highly appreciated!
> > Thanks
> > Nic
> >
> >
>
>
Anonymous
July 13, 2005 7:39:04 PM

Archived from groups: rec.games.pinball (More info?)

Thought about it, but like I said before, don't really like the price and
look of it.
Time... I've got lots of time... : ) If I can save a hundred bucks, I'll do
it. Worse
comes to worse, I'll probably end up with the pin liz boards.

Nic

"John Arriola" <nospam@nowhere.com> wrote in message
news:L5adnbNqY9trxEjfRVn-oA@speakeasy.net...
> How many hours do you have in this project so far? I don't know about your
> time, but clearly it is cheaper to go with a proven superior solution and
> fix it once.
> Go with pin liz and FAGETABOUTIT
>
> --
>
> John Arriola
>
> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
> news:aQbBe.2888$p5.1010@nnrp.ca.mci.com!nnrp1.uunet.ca...
> > Thanks for the tip, but 120+Ship is a bit too much for me right now, if
I
> > can figure out a way to get away with a couple optos, I'd be much
happier.
> > And to be honest, I don't quite like the blue LED look.
> >
> > Thanks
> > Nic
> >
> > "Otto" <ottondebremove&%$*@bellsouth.net> wrote in message
> > news:ELbBe.96291$du.69737@bignews1.bellsouth.net...
> >> Put the problem to bed with these:
> >>
> >> http://www.pbliz.com/id31.htm
> >>
> >> If you want plug and play, send the assembly and Joel will send it back
> > 100%
> >> plug and play.
> >>
> >>
> >>
> >> Otto
> >>
> >> CARGPB11
> >>
> >> My web page: http://home.bellsouth.net/p/PWP-Ottoslanding
> >>
> >> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
> >> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
> >> > My clock's been broken for the longest and I need it fixed now.
> >>
> >>
> >
> >
>
>
Anonymous
July 13, 2005 9:04:21 PM

Archived from groups: rec.games.pinball (More info?)

I believe Happ Controls have the opto you need which is a drop in
replacement.
Anonymous
July 13, 2005 9:30:18 PM

Archived from groups: rec.games.pinball (More info?)

Special optos were made exclusively for WMS because
they had the _exact_problem_you_are_having_. It is
a problem in the design of the WPC switch matrix that
causes these issues. Replacement optos that are currently
being offered for sale may or may not meet the specs of
those originally made for WMS. At worst, they will fail
again, and you will have to replace them.....again.

These things were dealt with, and solved with the redesigned
replacement boards done by the Pinball Lizard. Joel identified
the problem, and corrected the design. The special part is
not required in his design.

In other words, you are chasing your tail....

Fred
TX
CARGPB#8

===================================
"Nicolas Legare" <no_email_please@hotmail.com> wrote in message news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...

Any comment or answer is highly appreciated!
Thanks
Nic
July 13, 2005 9:41:15 PM

Archived from groups: rec.games.pinball (More info?)

Go for the transistor mod, as Jim suggested. You built the hours board,
you are close.

Most of the optos in the game have a dedicated 339 circuit to drive the
switch matrix. For the clock, they connected the optos direct to the
switch matrix, and regular optos might work in practoce, but not in
theory or production. They doin't pull enough current to guarantee a
low voltage at the comparator. So you get flicker.
July 13, 2005 9:41:17 PM

Archived from groups: rec.games.pinball (More info?)

Go for the transistor mod, as Jim suggested. You built the hours board,
you are close.

Most of the optos in the game have a dedicated 339 circuit to drive the
switch matrix. For the clock, they connected the optos direct to the
switch matrix, and regular optos might work in practoce, but not in
theory or production. They doin't pull enough current to guarantee a
low voltage at the comparator. So you get flicker.
Anonymous
July 13, 2005 10:44:05 PM

Archived from groups: rec.games.pinball (More info?)

Yes you can. Thinking of adding these to the next version of our clock
boards. All you need to do is to have the OPB act as a pre-driver for the
12mA sink current. It is not brain surgery. Good luck.

As to why hours vs minutes, vs other applications, it is a very marginal
design. you have the detector being driven from a 12V rail through a 1k
resistor (some applications a 1.2k resistor). For VCEsat of less than one
volt, the OPBs (an others) struggle.

Jim Knight - Who hates optos


"Nicolas Legare" <no_email_please@hotmail.com> wrote in message
news:B3eBe.2912$p5.871@nnrp.ca.mci.com!nnrp1.uunet.ca...
> Hmmmm ok, any ideas why they work on the hours side then??? I could
> probably
> just add a transistor to each of the minutes optos to give 12mA to the
> LM339... if I understand what you are saying...?
>
> Thanks
> Nic
>
>
>
> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
> news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
>> Nicolas,
>>
>> You cannot assume the OPB804's will work in any WMS application. The
>> OPB804s are rated at 0.5mA of current in the detector side. WMS tries to
>> pump 12.0mA into there. WMS actually had Quality Technologies (QT) do a
>> special fabrication which was stamped .0086. If you were to change your
>> OPB804s to QVE1223.0086 you problem will go away!
>>
>> Jim Knight - Who is painfully aware of the differences between OPBs and
>> QVEs.
>>
>> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
>> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
>> > Ok, before I get hate mail, let me tell you that I know this subject
>> > has
>> > been discussed to death before. But I just can't find a good
>> > explanation
>> > to
>> > my problem. Please help me!
>> >
>> > My clock's been broken for the longest and I need it fixed now. I made
> my
>> > own minutes board (with the extra diodes) to replace the old burnt
> board.
>> > Replaced all the optos with OPB804 (I know it's not supposed to be a
> good
>> > replacement part). These actually work wonderfully as the hour optos
>> > but
>> > just not as the minutes. Well actually that's not completely true, they
> do
>> > work when I run the clock test and trigger the optos with a piece of
>> > paper,
>> > I can see all 8 working well. Once I put the clock back together and
>> > run
>> > the
>> > clock test (doing forward), the 15 and 30 minute optos keep on
> flickering
>> > all the time for no reason. So what's the deal here? Why aren't the
>> > hour
>> > optos doing the same? I thought it might have been RF interference from
>> > the
>> > motor so I tried a ceramic cap on it, no luck. Other problem, I don't
>> > quite
>> > understand the thing with this sensitivity problem, I can't even find a
>> > friggin datasheet with graphs about this component (anybody has it?).
> Why
>> > is
>> > it that the .0086 sensitivity is required? How will the circuit differ
> if
>> > I
>> > use the OPB804 (electrically)? What I figure is that the
>> > phototransistor
>> > will trigger easier than with lower sensitivity, but that doesn't quite
>> > make
>> > sense since my OPB804 is triggerring all the time for no reason (I
> thought
>> > it had a lesser sensitivity...). I heard those 339s die all the time,
>> > could
>> > this possibly be the case?
>> >
>> > Any comment or answer is highly appreciated!
>> > Thanks
>> > Nic
>> >
>> >
>>
>>
>
>
Anonymous
July 13, 2005 11:16:51 PM

Archived from groups: rec.games.pinball (More info?)

Whoop, forgot to put in my 2c on the pre-driver. Would put a 10k resistor
to the 12V into the collector of the OPB804, that would reduce the current
to 1.2mA which should be fine. Do a emitter follower onto the npn driver to
the switch matrix. You could use 2n2222 or anything like that.

Jim Knight
"Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
news:4L2dneWCNbSqBEjfRVn-pQ@comcast.com...
> Yes you can. Thinking of adding these to the next version of our clock
> boards. All you need to do is to have the OPB act as a pre-driver for the
> 12mA sink current. It is not brain surgery. Good luck.
>
> As to why hours vs minutes, vs other applications, it is a very marginal
> design. you have the detector being driven from a 12V rail through a 1k
> resistor (some applications a 1.2k resistor). For VCEsat of less than one
> volt, the OPBs (an others) struggle.
>
> Jim Knight - Who hates optos
>
>
> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
> news:B3eBe.2912$p5.871@nnrp.ca.mci.com!nnrp1.uunet.ca...
>> Hmmmm ok, any ideas why they work on the hours side then??? I could
>> probably
>> just add a transistor to each of the minutes optos to give 12mA to the
>> LM339... if I understand what you are saying...?
>>
>> Thanks
>> Nic
>>
>>
>>
>> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
>> news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
>>> Nicolas,
>>>
>>> You cannot assume the OPB804's will work in any WMS application. The
>>> OPB804s are rated at 0.5mA of current in the detector side. WMS tries
>>> to
>>> pump 12.0mA into there. WMS actually had Quality Technologies (QT) do a
>>> special fabrication which was stamped .0086. If you were to change your
>>> OPB804s to QVE1223.0086 you problem will go away!
>>>
>>> Jim Knight - Who is painfully aware of the differences between OPBs and
>>> QVEs.
>>>
>>> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
>>> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
>>> > Ok, before I get hate mail, let me tell you that I know this subject
>>> > has
>>> > been discussed to death before. But I just can't find a good
>>> > explanation
>>> > to
>>> > my problem. Please help me!
>>> >
>>> > My clock's been broken for the longest and I need it fixed now. I made
>> my
>>> > own minutes board (with the extra diodes) to replace the old burnt
>> board.
>>> > Replaced all the optos with OPB804 (I know it's not supposed to be a
>> good
>>> > replacement part). These actually work wonderfully as the hour optos
>>> > but
>>> > just not as the minutes. Well actually that's not completely true,
>>> > they
>> do
>>> > work when I run the clock test and trigger the optos with a piece of
>>> > paper,
>>> > I can see all 8 working well. Once I put the clock back together and
>>> > run
>>> > the
>>> > clock test (doing forward), the 15 and 30 minute optos keep on
>> flickering
>>> > all the time for no reason. So what's the deal here? Why aren't the
>>> > hour
>>> > optos doing the same? I thought it might have been RF interference
>>> > from
>>> > the
>>> > motor so I tried a ceramic cap on it, no luck. Other problem, I don't
>>> > quite
>>> > understand the thing with this sensitivity problem, I can't even find
>>> > a
>>> > friggin datasheet with graphs about this component (anybody has it?).
>> Why
>>> > is
>>> > it that the .0086 sensitivity is required? How will the circuit differ
>> if
>>> > I
>>> > use the OPB804 (electrically)? What I figure is that the
>>> > phototransistor
>>> > will trigger easier than with lower sensitivity, but that doesn't
>>> > quite
>>> > make
>>> > sense since my OPB804 is triggerring all the time for no reason (I
>> thought
>>> > it had a lesser sensitivity...). I heard those 339s die all the time,
>>> > could
>>> > this possibly be the case?
>>> >
>>> > Any comment or answer is highly appreciated!
>>> > Thanks
>>> > Nic
>>> >
>>> >
>>>
>>>
>>
>>
>
>
Anonymous
July 13, 2005 11:49:32 PM

Archived from groups: rec.games.pinball (More info?)

Perfect, I'll try this out, thanks for the help, really appreciate it!
Finally makes sense.

Nic

"Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
news:4L2dneWCNbSqBEjfRVn-pQ@comcast.com...
> Yes you can. Thinking of adding these to the next version of our clock
> boards. All you need to do is to have the OPB act as a pre-driver for the
> 12mA sink current. It is not brain surgery. Good luck.
>
> As to why hours vs minutes, vs other applications, it is a very marginal
> design. you have the detector being driven from a 12V rail through a 1k
> resistor (some applications a 1.2k resistor). For VCEsat of less than one
> volt, the OPBs (an others) struggle.
>
> Jim Knight - Who hates optos
>
>
> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
> news:B3eBe.2912$p5.871@nnrp.ca.mci.com!nnrp1.uunet.ca...
> > Hmmmm ok, any ideas why they work on the hours side then??? I could
> > probably
> > just add a transistor to each of the minutes optos to give 12mA to the
> > LM339... if I understand what you are saying...?
> >
> > Thanks
> > Nic
> >
> >
> >
> > "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
> > news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
> >> Nicolas,
> >>
> >> You cannot assume the OPB804's will work in any WMS application. The
> >> OPB804s are rated at 0.5mA of current in the detector side. WMS tries
to
> >> pump 12.0mA into there. WMS actually had Quality Technologies (QT) do
a
> >> special fabrication which was stamped .0086. If you were to change
your
> >> OPB804s to QVE1223.0086 you problem will go away!
> >>
> >> Jim Knight - Who is painfully aware of the differences between OPBs and
> >> QVEs.
> >>
> >> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
> >> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
> >> > Ok, before I get hate mail, let me tell you that I know this subject
> >> > has
> >> > been discussed to death before. But I just can't find a good
> >> > explanation
> >> > to
> >> > my problem. Please help me!
> >> >
> >> > My clock's been broken for the longest and I need it fixed now. I
made
> > my
> >> > own minutes board (with the extra diodes) to replace the old burnt
> > board.
> >> > Replaced all the optos with OPB804 (I know it's not supposed to be a
> > good
> >> > replacement part). These actually work wonderfully as the hour optos
> >> > but
> >> > just not as the minutes. Well actually that's not completely true,
they
> > do
> >> > work when I run the clock test and trigger the optos with a piece of
> >> > paper,
> >> > I can see all 8 working well. Once I put the clock back together and
> >> > run
> >> > the
> >> > clock test (doing forward), the 15 and 30 minute optos keep on
> > flickering
> >> > all the time for no reason. So what's the deal here? Why aren't the
> >> > hour
> >> > optos doing the same? I thought it might have been RF interference
from
> >> > the
> >> > motor so I tried a ceramic cap on it, no luck. Other problem, I don't
> >> > quite
> >> > understand the thing with this sensitivity problem, I can't even find
a
> >> > friggin datasheet with graphs about this component (anybody has it?).
> > Why
> >> > is
> >> > it that the .0086 sensitivity is required? How will the circuit
differ
> > if
> >> > I
> >> > use the OPB804 (electrically)? What I figure is that the
> >> > phototransistor
> >> > will trigger easier than with lower sensitivity, but that doesn't
quite
> >> > make
> >> > sense since my OPB804 is triggerring all the time for no reason (I
> > thought
> >> > it had a lesser sensitivity...). I heard those 339s die all the time,
> >> > could
> >> > this possibly be the case?
> >> >
> >> > Any comment or answer is highly appreciated!
> >> > Thanks
> >> > Nic
> >> >
> >> >
> >>
> >>
> >
> >
>
>
Anonymous
July 14, 2005 1:12:54 AM

Archived from groups: rec.games.pinball (More info?)

if you don't want to pay pbliz board price. why not buy from
rottendog. His are around $75 and better design. Don't know if he has
regular and led versions thou. forgot :/ 
Anonymous
July 14, 2005 1:53:46 AM

Archived from groups: rec.games.pinball (More info?)

The OPB804's are not rated 0.5mA but are rated far higher. The worst case
current of 30mA isn't even approached in 'normal' operation (assuming no
coil shorts to switch matrix, etc). Within the TZ clock, these are operated
at (12volts / 1200 ohms) = 10mA. The 1200 ohm value is the switch matrix's
LM339 input pullup resistor to 12 volts. Actually, the current will be less
than 10mA as you need to subtract switch column diode voltage drop (D5 on
clock board) and the switch driver ULN2803 transistor array's VOL from the
12 volts. Actual current flow is probably closer to 8mA. So, both OPB804
and QVE11233 can handle the current just fine.

Both QVE11233 and OPB804 are rated at the exact same VCE(sat) of 0.4 volts -
plenty low enough for the LM339 comparator. Both are spec'd at this voltage
with a diode forward current of 20mA and a transistor collector of 250uA.
Both have nearly identical current transfer ratio's (and both are pretty
bad). Electrically, they are quite interchangeable.

The main differences between the standard QVE11233 and the OPB804 is the
slit width (0.125 versus 0.150), the apertures for the QVE11233 are 0.05 and
0.01 for the emitter & detector whereas the OPB804 is wide open on both
emitter and detector. And a few other minor package dimensions that don't
matter. The aperture's is probably what makes the biggest difference here.
The OPB804 has a wider gap AND a wide open detector and therefore picks up
interference IR from many directions. There are also a few absolute max
ratings but none that make any difference in this application.

The difference with the QT parts as ordered by Williams, so I was told by
tech support at Fairchild (who owns QT Opto), between the QVE11233 and the
QVE11233.0086 is *only* the receiver sensitivity. And the sensitivity
adjustment was performed by adjusting the sensor's aperture size. This is
used to reduce IR interference from other sources... i.e. lamps. I would
guess that the reason an OPB804 doesn't work properly is due to excessive IR
interferrence. This excessive IR interferrence is forcing the OPB804 into
the linear zone where it's conducting but not going into saturation -
forcing the output voltage going into the LM339 to be at some mid level
rather than a distinct 'on' or 'off' voltage.

The thing that makes the QVE11233.0086 so hard to replace with a newer and
more sensitive opto is the lead spacing. Most opto's are made with 0.32"
and 0.22" lead spacings. The QVE11233 has a 0.3" lead spacing.

-- Ed



"Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
> Nicolas,
>
> You cannot assume the OPB804's will work in any WMS application. The
> OPB804s are rated at 0.5mA of current in the detector side. WMS tries to
> pump 12.0mA into there. WMS actually had Quality Technologies (QT) do a
> special fabrication which was stamped .0086. If you were to change your
> OPB804s to QVE1223.0086 you problem will go away!
>
> Jim Knight - Who is painfully aware of the differences between OPBs and
> QVEs.
>
> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
>> Ok, before I get hate mail, let me tell you that I know this subject has
>> been discussed to death before. But I just can't find a good explanation
>> to
>> my problem. Please help me!
>>
>> My clock's been broken for the longest and I need it fixed now. I made my
>> own minutes board (with the extra diodes) to replace the old burnt board.
>> Replaced all the optos with OPB804 (I know it's not supposed to be a good
>> replacement part). These actually work wonderfully as the hour optos but
>> just not as the minutes. Well actually that's not completely true, they
>> do
>> work when I run the clock test and trigger the optos with a piece of
>> paper,
>> I can see all 8 working well. Once I put the clock back together and run
>> the
>> clock test (doing forward), the 15 and 30 minute optos keep on flickering
>> all the time for no reason. So what's the deal here? Why aren't the hour
>> optos doing the same? I thought it might have been RF interference from
>> the
>> motor so I tried a ceramic cap on it, no luck. Other problem, I don't
>> quite
>> understand the thing with this sensitivity problem, I can't even find a
>> friggin datasheet with graphs about this component (anybody has it?). Why
>> is
>> it that the .0086 sensitivity is required? How will the circuit differ if
>> I
>> use the OPB804 (electrically)? What I figure is that the phototransistor
>> will trigger easier than with lower sensitivity, but that doesn't quite
>> make
>> sense since my OPB804 is triggerring all the time for no reason (I
>> thought
>> it had a lesser sensitivity...). I heard those 339s die all the time,
>> could
>> this possibly be the case?
>>
>> Any comment or answer is highly appreciated!
>> Thanks
>> Nic
>>
>>
>
>
July 14, 2005 2:03:34 AM

Archived from groups: rec.games.pinball (More info?)

The way that the circuit works, the opto has to pull enough current
through the 1K2 resistor to drop 7V. That's just about 5.8mA. If you
look at the QVE data sheet, it is absolutely guaranteed to pull 0.5mA -
yes, 5.3mA short of a trigger. Now, you may be lucky and get more
current, but you are looking for over 1,000% better than spec.

The OBP data sheet promises 5mA. Yes, still not enough but close enough
for that Williams designer. Brand new, they will do better.

I've never found an opto spec with enough current to guarantee pulling
that line to 5V. And if your 5V is low, your 12V high, and your
resistor only 1K1 you could need almost 8mA.

I am a bit puzzled about the saturation voltage spec. I am not sure
that you could get enough light from the LED to drive enough base
current for saturation.

Bottom line: help needed. These things can't guarantee good results on
their own.

Jim Knight wrote:
> ;-)
>
> Jim Knight
>
> PS Do you want me to send you the spec for the QVE11223.0086? It took me
> about 1 year to get it and at least 10 service calls.
>
> Optek has sent me specially built OPB804 type prototypes that meet the WMS
> spec, but, without an order of x00k of them, we are out of luck.
>
> All of my newly layed out boards with optos (if any) will have
> pre-drivers. I agree, what new grad at WMS designed these into ALL of their
> optos.
>
> Jim Knight
>
>
> "GPE" <See_my_website_for_email@cox.net> wrote in message
> news:a5lBe.16244$fV.11459@okepread06...
> >
> > "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
> > news:E9ednZX5gb6WREjfRVn-1w@comcast.com...
> >> Ed,
> >> Been on the phone with Fairchild engineers for long periods of time as
> >> well. Here is the data sheet:
> >>
> >> http://www.optekinc.com/pdf/OPB804.pdf
> >>
> >> On page 2. VCEsat of 0.4V is only guaranteed for phototransistor current
> >> of 0.25mA. The WMS currents are on the order of 12mA. I have measured
> >> the VCEsat in WMS games at over 2V.
> >>
> >
> > I've also been on the phone to the Fairchild guys. Took about ten of them
> > guys before one finally had a clue about the .0086
> >
> > Take a look at the QVE11223 at Fairchild:
> > http://www.fairchildsemi.com/ds/QV/QVE11233.pdf
> > On page 2. VCEsat of 0.4V is also only guaranteed for a phototransistor
> > current of 0.25mA. hmmm, this looks quite familiar, eh? And it will work
> > just fine with a phototransistor current of upto 30mA.
> >
> > The tech said the electrical specs for the two (QVE11233 vs .0086) would
> > be identical. Only the aperture sizes changed.
> >
> > HOWEVER, after looking at the spec sheets for the .0086 that somebody else
> > posted (I never saw these until a couple minutes ago) -- At higher
> > currents, the QVE11223.0086 will maintain a lower VCE(SAT) than the
> > OPB804. At 10mA (just over the WMS current) - the .0086 has a VCE(SAT) of
> > 0.8V max. This is the point not mentioned by the Fairchild tech. The OPB
> > doesn't list the VCE(SAT) at this voltage but extrapolation would put it
> > pretty high. You'd still think this would be lower than the 6 volts
> > expected by the comparator, though.
> >
> > So, based on what I saw with this spec sheet - the guy I talked to at
> > Fairchild apparently didn't have as much of a clue as I hoped.
> >
> > -- Ed .. still shaking his head at how badly designed the williams boards
> > were... Thank God these guys don't design military or life support
> > hardware!
> >
> >
> >
> >> Please let me know where I am wrong.
> >>
> >> Jim Knight
> >>
> >>
> >> "GPE" <See_my_website_for_email@cox.net> wrote in message
> >> news:XskBe.16235$fV.3416@okepread06...
> >>>
> >>> The OPB804's are not rated 0.5mA but are rated far higher. The worst
> >>> case current of 30mA isn't even approached in 'normal' operation
> >>> (assuming no coil shorts to switch matrix, etc). Within the TZ clock,
> >>> these are operated at (12volts / 1200 ohms) = 10mA. The 1200 ohm value
> >>> is the switch matrix's LM339 input pullup resistor to 12 volts.
> >>> Actually, the current will be less than 10mA as you need to subtract
> >>> switch column diode voltage drop (D5 on clock board) and the switch
> >>> driver ULN2803 transistor array's VOL from the 12 volts. Actual current
> >>> flow is probably closer to 8mA. So, both OPB804 and QVE11233 can handle
> >>> the current just fine.
> >>>
> >>> Both QVE11233 and OPB804 are rated at the exact same VCE(sat) of 0.4
> >>> volts - plenty low enough for the LM339 comparator. Both are spec'd at
> >>> this voltage with a diode forward current of 20mA and a transistor
> >>> collector of 250uA. Both have nearly identical current transfer ratio's
> >>> (and both are pretty bad). Electrically, they are quite
> >>> interchangeable.
> >>>
> >>> The main differences between the standard QVE11233 and the OPB804 is the
> >>> slit width (0.125 versus 0.150), the apertures for the QVE11233 are 0.05
> >>> and 0.01 for the emitter & detector whereas the OPB804 is wide open on
> >>> both emitter and detector. And a few other minor package dimensions
> >>> that don't matter. The aperture's is probably what makes the biggest
> >>> difference here. The OPB804 has a wider gap AND a wide open detector and
> >>> therefore picks up interference IR from many directions. There are also
> >>> a few absolute max ratings but none that make any difference in this
> >>> application.
> >>>
> >>> The difference with the QT parts as ordered by Williams, so I was told
> >>> by tech support at Fairchild (who owns QT Opto), between the QVE11233
> >>> and the QVE11233.0086 is *only* the receiver sensitivity. And the
> >>> sensitivity adjustment was performed by adjusting the sensor's aperture
> >>> size. This is used to reduce IR interference from other sources... i.e.
> >>> lamps. I would guess that the reason an OPB804 doesn't work properly is
> >>> due to excessive IR interferrence. This excessive IR interferrence is
> >>> forcing the OPB804 into the linear zone where it's conducting but not
> >>> going into saturation - forcing the output voltage going into the LM339
> >>> to be at some mid level rather than a distinct 'on' or 'off' voltage.
> >>>
> >>> The thing that makes the QVE11233.0086 so hard to replace with a newer
> >>> and more sensitive opto is the lead spacing. Most opto's are made with
> >>> 0.32" and 0.22" lead spacings. The QVE11233 has a 0.3" lead spacing.
> >>>
> >>> -- Ed
> >>>
> >>>
> >>>
> >>> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
> >>> news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
> >>>> Nicolas,
> >>>>
> >>>> You cannot assume the OPB804's will work in any WMS application. The
> >>>> OPB804s are rated at 0.5mA of current in the detector side. WMS tries
> >>>> to pump 12.0mA into there. WMS actually had Quality Technologies (QT)
> >>>> do a special fabrication which was stamped .0086. If you were to
> >>>> change your OPB804s to QVE1223.0086 you problem will go away!
> >>>>
> >>>> Jim Knight - Who is painfully aware of the differences between OPBs and
> >>>> QVEs.
> >>>>
> >>>> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
> >>>> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
> >>>>> Ok, before I get hate mail, let me tell you that I know this subject
> >>>>> has
> >>>>> been discussed to death before. But I just can't find a good
> >>>>> explanation to
> >>>>> my problem. Please help me!
> >>>>>
> >>>>> My clock's been broken for the longest and I need it fixed now. I made
> >>>>> my
> >>>>> own minutes board (with the extra diodes) to replace the old burnt
> >>>>> board.
> >>>>> Replaced all the optos with OPB804 (I know it's not supposed to be a
> >>>>> good
> >>>>> replacement part). These actually work wonderfully as the hour optos
> >>>>> but
> >>>>> just not as the minutes. Well actually that's not completely true,
> >>>>> they do
> >>>>> work when I run the clock test and trigger the optos with a piece of
> >>>>> paper,
> >>>>> I can see all 8 working well. Once I put the clock back together and
> >>>>> run the
> >>>>> clock test (doing forward), the 15 and 30 minute optos keep on
> >>>>> flickering
> >>>>> all the time for no reason. So what's the deal here? Why aren't the
> >>>>> hour
> >>>>> optos doing the same? I thought it might have been RF interference
> >>>>> from the
> >>>>> motor so I tried a ceramic cap on it, no luck. Other problem, I don't
> >>>>> quite
> >>>>> understand the thing with this sensitivity problem, I can't even find
> >>>>> a
> >>>>> friggin datasheet with graphs about this component (anybody has it?).
> >>>>> Why is
> >>>>> it that the .0086 sensitivity is required? How will the circuit differ
> >>>>> if I
> >>>>> use the OPB804 (electrically)? What I figure is that the
> >>>>> phototransistor
> >>>>> will trigger easier than with lower sensitivity, but that doesn't
> >>>>> quite make
> >>>>> sense since my OPB804 is triggerring all the time for no reason (I
> >>>>> thought
> >>>>> it had a lesser sensitivity...). I heard those 339s die all the time,
> >>>>> could
> >>>>> this possibly be the case?
> >>>>>
> >>>>> Any comment or answer is highly appreciated!
> >>>>> Thanks
> >>>>> Nic
> >>>>>
> >>>>>
> >>>>
> >>>>
> >>>
> >>>
> >>
> >>
> >
> >
July 14, 2005 2:03:52 AM

Archived from groups: rec.games.pinball (More info?)

The way that the circuit works, the opto has to pull enough current
through the 1K2 resistor to drop 7V. That's just about 5.8mA. If you
look at the QVE data sheet, it is absolutely guaranteed to pull 0.5mA -
yes, 5.3mA short of a trigger. Now, you may be lucky and get more
current, but you are looking for over 1,000% better than spec.

The OBP data sheet promises 5mA. Yes, still not enough but close enough
for that Williams designer. Brand new, they will do better.

I've never found an opto spec with enough current to guarantee pulling
that line to 5V. And if your 5V is low, your 12V high, and your
resistor only 1K1 you could need almost 8mA.

I am a bit puzzled about the saturation voltage spec. I am not sure
that you could get enough light from the LED to drive enough base
current for saturation.

Bottom line: help needed. These things can't guarantee good results on
their own.

Jim Knight wrote:
> ;-)
>
> Jim Knight
>
> PS Do you want me to send you the spec for the QVE11223.0086? It took me
> about 1 year to get it and at least 10 service calls.
>
> Optek has sent me specially built OPB804 type prototypes that meet the WMS
> spec, but, without an order of x00k of them, we are out of luck.
>
> All of my newly layed out boards with optos (if any) will have
> pre-drivers. I agree, what new grad at WMS designed these into ALL of their
> optos.
>
> Jim Knight
>
>
> "GPE" <See_my_website_for_email@cox.net> wrote in message
> news:a5lBe.16244$fV.11459@okepread06...
> >
> > "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
> > news:E9ednZX5gb6WREjfRVn-1w@comcast.com...
> >> Ed,
> >> Been on the phone with Fairchild engineers for long periods of time as
> >> well. Here is the data sheet:
> >>
> >> http://www.optekinc.com/pdf/OPB804.pdf
> >>
> >> On page 2. VCEsat of 0.4V is only guaranteed for phototransistor current
> >> of 0.25mA. The WMS currents are on the order of 12mA. I have measured
> >> the VCEsat in WMS games at over 2V.
> >>
> >
> > I've also been on the phone to the Fairchild guys. Took about ten of them
> > guys before one finally had a clue about the .0086
> >
> > Take a look at the QVE11223 at Fairchild:
> > http://www.fairchildsemi.com/ds/QV/QVE11233.pdf
> > On page 2. VCEsat of 0.4V is also only guaranteed for a phototransistor
> > current of 0.25mA. hmmm, this looks quite familiar, eh? And it will work
> > just fine with a phototransistor current of upto 30mA.
> >
> > The tech said the electrical specs for the two (QVE11233 vs .0086) would
> > be identical. Only the aperture sizes changed.
> >
> > HOWEVER, after looking at the spec sheets for the .0086 that somebody else
> > posted (I never saw these until a couple minutes ago) -- At higher
> > currents, the QVE11223.0086 will maintain a lower VCE(SAT) than the
> > OPB804. At 10mA (just over the WMS current) - the .0086 has a VCE(SAT) of
> > 0.8V max. This is the point not mentioned by the Fairchild tech. The OPB
> > doesn't list the VCE(SAT) at this voltage but extrapolation would put it
> > pretty high. You'd still think this would be lower than the 6 volts
> > expected by the comparator, though.
> >
> > So, based on what I saw with this spec sheet - the guy I talked to at
> > Fairchild apparently didn't have as much of a clue as I hoped.
> >
> > -- Ed .. still shaking his head at how badly designed the williams boards
> > were... Thank God these guys don't design military or life support
> > hardware!
> >
> >
> >
> >> Please let me know where I am wrong.
> >>
> >> Jim Knight
> >>
> >>
> >> "GPE" <See_my_website_for_email@cox.net> wrote in message
> >> news:XskBe.16235$fV.3416@okepread06...
> >>>
> >>> The OPB804's are not rated 0.5mA but are rated far higher. The worst
> >>> case current of 30mA isn't even approached in 'normal' operation
> >>> (assuming no coil shorts to switch matrix, etc). Within the TZ clock,
> >>> these are operated at (12volts / 1200 ohms) = 10mA. The 1200 ohm value
> >>> is the switch matrix's LM339 input pullup resistor to 12 volts.
> >>> Actually, the current will be less than 10mA as you need to subtract
> >>> switch column diode voltage drop (D5 on clock board) and the switch
> >>> driver ULN2803 transistor array's VOL from the 12 volts. Actual current
> >>> flow is probably closer to 8mA. So, both OPB804 and QVE11233 can handle
> >>> the current just fine.
> >>>
> >>> Both QVE11233 and OPB804 are rated at the exact same VCE(sat) of 0.4
> >>> volts - plenty low enough for the LM339 comparator. Both are spec'd at
> >>> this voltage with a diode forward current of 20mA and a transistor
> >>> collector of 250uA. Both have nearly identical current transfer ratio's
> >>> (and both are pretty bad). Electrically, they are quite
> >>> interchangeable.
> >>>
> >>> The main differences between the standard QVE11233 and the OPB804 is the
> >>> slit width (0.125 versus 0.150), the apertures for the QVE11233 are 0.05
> >>> and 0.01 for the emitter & detector whereas the OPB804 is wide open on
> >>> both emitter and detector. And a few other minor package dimensions
> >>> that don't matter. The aperture's is probably what makes the biggest
> >>> difference here. The OPB804 has a wider gap AND a wide open detector and
> >>> therefore picks up interference IR from many directions. There are also
> >>> a few absolute max ratings but none that make any difference in this
> >>> application.
> >>>
> >>> The difference with the QT parts as ordered by Williams, so I was told
> >>> by tech support at Fairchild (who owns QT Opto), between the QVE11233
> >>> and the QVE11233.0086 is *only* the receiver sensitivity. And the
> >>> sensitivity adjustment was performed by adjusting the sensor's aperture
> >>> size. This is used to reduce IR interference from other sources... i.e.
> >>> lamps. I would guess that the reason an OPB804 doesn't work properly is
> >>> due to excessive IR interferrence. This excessive IR interferrence is
> >>> forcing the OPB804 into the linear zone where it's conducting but not
> >>> going into saturation - forcing the output voltage going into the LM339
> >>> to be at some mid level rather than a distinct 'on' or 'off' voltage.
> >>>
> >>> The thing that makes the QVE11233.0086 so hard to replace with a newer
> >>> and more sensitive opto is the lead spacing. Most opto's are made with
> >>> 0.32" and 0.22" lead spacings. The QVE11233 has a 0.3" lead spacing.
> >>>
> >>> -- Ed
> >>>
> >>>
> >>>
> >>> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
> >>> news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
> >>>> Nicolas,
> >>>>
> >>>> You cannot assume the OPB804's will work in any WMS application. The
> >>>> OPB804s are rated at 0.5mA of current in the detector side. WMS tries
> >>>> to pump 12.0mA into there. WMS actually had Quality Technologies (QT)
> >>>> do a special fabrication which was stamped .0086. If you were to
> >>>> change your OPB804s to QVE1223.0086 you problem will go away!
> >>>>
> >>>> Jim Knight - Who is painfully aware of the differences between OPBs and
> >>>> QVEs.
> >>>>
> >>>> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
> >>>> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
> >>>>> Ok, before I get hate mail, let me tell you that I know this subject
> >>>>> has
> >>>>> been discussed to death before. But I just can't find a good
> >>>>> explanation to
> >>>>> my problem. Please help me!
> >>>>>
> >>>>> My clock's been broken for the longest and I need it fixed now. I made
> >>>>> my
> >>>>> own minutes board (with the extra diodes) to replace the old burnt
> >>>>> board.
> >>>>> Replaced all the optos with OPB804 (I know it's not supposed to be a
> >>>>> good
> >>>>> replacement part). These actually work wonderfully as the hour optos
> >>>>> but
> >>>>> just not as the minutes. Well actually that's not completely true,
> >>>>> they do
> >>>>> work when I run the clock test and trigger the optos with a piece of
> >>>>> paper,
> >>>>> I can see all 8 working well. Once I put the clock back together and
> >>>>> run the
> >>>>> clock test (doing forward), the 15 and 30 minute optos keep on
> >>>>> flickering
> >>>>> all the time for no reason. So what's the deal here? Why aren't the
> >>>>> hour
> >>>>> optos doing the same? I thought it might have been RF interference
> >>>>> from the
> >>>>> motor so I tried a ceramic cap on it, no luck. Other problem, I don't
> >>>>> quite
> >>>>> understand the thing with this sensitivity problem, I can't even find
> >>>>> a
> >>>>> friggin datasheet with graphs about this component (anybody has it?).
> >>>>> Why is
> >>>>> it that the .0086 sensitivity is required? How will the circuit differ
> >>>>> if I
> >>>>> use the OPB804 (electrically)? What I figure is that the
> >>>>> phototransistor
> >>>>> will trigger easier than with lower sensitivity, but that doesn't
> >>>>> quite make
> >>>>> sense since my OPB804 is triggerring all the time for no reason (I
> >>>>> thought
> >>>>> it had a lesser sensitivity...). I heard those 339s die all the time,
> >>>>> could
> >>>>> this possibly be the case?
> >>>>>
> >>>>> Any comment or answer is highly appreciated!
> >>>>> Thanks
> >>>>> Nic
> >>>>>
> >>>>>
> >>>>
> >>>>
> >>>
> >>>
> >>
> >>
> >
> >
Anonymous
July 14, 2005 2:10:01 AM

Archived from groups: rec.games.pinball (More info?)

Best way to figure out if you need to amplify the current - measure it.
According to the QT and Optek data sheets, their opto's have a better than
50% current transfer ratio and nearly 100% with 20mA going into the emitting
diode. This means the opto *should* be capable of switching a 20mA load.
If the VSAT isn't going low enough, you may need to put in the transistor to
help pull it lower. This would probably be due to the transistor output not
completely going into (or coming out of) saturation.

-- Ed


"Nicolas Legare" <no_email_please@hotmail.com> wrote in message
news:B3eBe.2912$p5.871@nnrp.ca.mci.com!nnrp1.uunet.ca...
> Hmmmm ok, any ideas why they work on the hours side then??? I could
> probably
> just add a transistor to each of the minutes optos to give 12mA to the
> LM339... if I understand what you are saying...?
>
> Thanks
> Nic
>
>
>
> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
> news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
>> Nicolas,
>>
>> You cannot assume the OPB804's will work in any WMS application. The
>> OPB804s are rated at 0.5mA of current in the detector side. WMS tries to
>> pump 12.0mA into there. WMS actually had Quality Technologies (QT) do a
>> special fabrication which was stamped .0086. If you were to change your
>> OPB804s to QVE1223.0086 you problem will go away!
>>
>> Jim Knight - Who is painfully aware of the differences between OPBs and
>> QVEs.
>>
>> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
>> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
>> > Ok, before I get hate mail, let me tell you that I know this subject
>> > has
>> > been discussed to death before. But I just can't find a good
>> > explanation
>> > to
>> > my problem. Please help me!
>> >
>> > My clock's been broken for the longest and I need it fixed now. I made
> my
>> > own minutes board (with the extra diodes) to replace the old burnt
> board.
>> > Replaced all the optos with OPB804 (I know it's not supposed to be a
> good
>> > replacement part). These actually work wonderfully as the hour optos
>> > but
>> > just not as the minutes. Well actually that's not completely true, they
> do
>> > work when I run the clock test and trigger the optos with a piece of
>> > paper,
>> > I can see all 8 working well. Once I put the clock back together and
>> > run
>> > the
>> > clock test (doing forward), the 15 and 30 minute optos keep on
> flickering
>> > all the time for no reason. So what's the deal here? Why aren't the
>> > hour
>> > optos doing the same? I thought it might have been RF interference from
>> > the
>> > motor so I tried a ceramic cap on it, no luck. Other problem, I don't
>> > quite
>> > understand the thing with this sensitivity problem, I can't even find a
>> > friggin datasheet with graphs about this component (anybody has it?).
> Why
>> > is
>> > it that the .0086 sensitivity is required? How will the circuit differ
> if
>> > I
>> > use the OPB804 (electrically)? What I figure is that the
>> > phototransistor
>> > will trigger easier than with lower sensitivity, but that doesn't quite
>> > make
>> > sense since my OPB804 is triggerring all the time for no reason (I
> thought
>> > it had a lesser sensitivity...). I heard those 339s die all the time,
>> > could
>> > this possibly be the case?
>> >
>> > Any comment or answer is highly appreciated!
>> > Thanks
>> > Nic
>> >
>> >
>>
>>
>
>
Anonymous
July 14, 2005 3:16:24 AM

Archived from groups: rec.games.pinball (More info?)

Ed,
Been on the phone with Fairchild engineers for long periods of time as
well. Here is the data sheet:

http://www.optekinc.com/pdf/OPB804.pdf

On page 2. VCEsat of 0.4V is only guaranteed for phototransistor current
of 0.25mA. The WMS currents are on the order of 12mA. I have measured the
VCEsat in WMS games at over 2V.

Please let me know where I am wrong.

Jim Knight


"GPE" <See_my_website_for_email@cox.net> wrote in message
news:XskBe.16235$fV.3416@okepread06...
>
> The OPB804's are not rated 0.5mA but are rated far higher. The worst case
> current of 30mA isn't even approached in 'normal' operation (assuming no
> coil shorts to switch matrix, etc). Within the TZ clock, these are
> operated at (12volts / 1200 ohms) = 10mA. The 1200 ohm value is the
> switch matrix's LM339 input pullup resistor to 12 volts. Actually, the
> current will be less than 10mA as you need to subtract switch column diode
> voltage drop (D5 on clock board) and the switch driver ULN2803 transistor
> array's VOL from the 12 volts. Actual current flow is probably closer to
> 8mA. So, both OPB804 and QVE11233 can handle the current just fine.
>
> Both QVE11233 and OPB804 are rated at the exact same VCE(sat) of 0.4
> volts - plenty low enough for the LM339 comparator. Both are spec'd at
> this voltage with a diode forward current of 20mA and a transistor
> collector of 250uA. Both have nearly identical current transfer ratio's
> (and both are pretty bad). Electrically, they are quite interchangeable.
>
> The main differences between the standard QVE11233 and the OPB804 is the
> slit width (0.125 versus 0.150), the apertures for the QVE11233 are 0.05
> and 0.01 for the emitter & detector whereas the OPB804 is wide open on
> both emitter and detector. And a few other minor package dimensions that
> don't matter. The aperture's is probably what makes the biggest
> difference here. The OPB804 has a wider gap AND a wide open detector and
> therefore picks up interference IR from many directions. There are also a
> few absolute max ratings but none that make any difference in this
> application.
>
> The difference with the QT parts as ordered by Williams, so I was told by
> tech support at Fairchild (who owns QT Opto), between the QVE11233 and the
> QVE11233.0086 is *only* the receiver sensitivity. And the sensitivity
> adjustment was performed by adjusting the sensor's aperture size. This is
> used to reduce IR interference from other sources... i.e. lamps. I would
> guess that the reason an OPB804 doesn't work properly is due to excessive
> IR interferrence. This excessive IR interferrence is forcing the OPB804
> into the linear zone where it's conducting but not going into saturation -
> forcing the output voltage going into the LM339 to be at some mid level
> rather than a distinct 'on' or 'off' voltage.
>
> The thing that makes the QVE11233.0086 so hard to replace with a newer and
> more sensitive opto is the lead spacing. Most opto's are made with 0.32"
> and 0.22" lead spacings. The QVE11233 has a 0.3" lead spacing.
>
> -- Ed
>
>
>
> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
> news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
>> Nicolas,
>>
>> You cannot assume the OPB804's will work in any WMS application. The
>> OPB804s are rated at 0.5mA of current in the detector side. WMS tries to
>> pump 12.0mA into there. WMS actually had Quality Technologies (QT) do a
>> special fabrication which was stamped .0086. If you were to change your
>> OPB804s to QVE1223.0086 you problem will go away!
>>
>> Jim Knight - Who is painfully aware of the differences between OPBs and
>> QVEs.
>>
>> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
>> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
>>> Ok, before I get hate mail, let me tell you that I know this subject has
>>> been discussed to death before. But I just can't find a good explanation
>>> to
>>> my problem. Please help me!
>>>
>>> My clock's been broken for the longest and I need it fixed now. I made
>>> my
>>> own minutes board (with the extra diodes) to replace the old burnt
>>> board.
>>> Replaced all the optos with OPB804 (I know it's not supposed to be a
>>> good
>>> replacement part). These actually work wonderfully as the hour optos but
>>> just not as the minutes. Well actually that's not completely true, they
>>> do
>>> work when I run the clock test and trigger the optos with a piece of
>>> paper,
>>> I can see all 8 working well. Once I put the clock back together and run
>>> the
>>> clock test (doing forward), the 15 and 30 minute optos keep on
>>> flickering
>>> all the time for no reason. So what's the deal here? Why aren't the hour
>>> optos doing the same? I thought it might have been RF interference from
>>> the
>>> motor so I tried a ceramic cap on it, no luck. Other problem, I don't
>>> quite
>>> understand the thing with this sensitivity problem, I can't even find a
>>> friggin datasheet with graphs about this component (anybody has it?).
>>> Why is
>>> it that the .0086 sensitivity is required? How will the circuit differ
>>> if I
>>> use the OPB804 (electrically)? What I figure is that the phototransistor
>>> will trigger easier than with lower sensitivity, but that doesn't quite
>>> make
>>> sense since my OPB804 is triggerring all the time for no reason (I
>>> thought
>>> it had a lesser sensitivity...). I heard those 339s die all the time,
>>> could
>>> this possibly be the case?
>>>
>>> Any comment or answer is highly appreciated!
>>> Thanks
>>> Nic
>>>
>>>
>>
>>
>
>
Anonymous
July 14, 2005 3:16:25 AM

Archived from groups: rec.games.pinball (More info?)

"Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
news:E9ednZX5gb6WREjfRVn-1w@comcast.com...
> Ed,
> Been on the phone with Fairchild engineers for long periods of time as
> well. Here is the data sheet:
>
> http://www.optekinc.com/pdf/OPB804.pdf
>
> On page 2. VCEsat of 0.4V is only guaranteed for phototransistor current
> of 0.25mA. The WMS currents are on the order of 12mA. I have measured
> the VCEsat in WMS games at over 2V.
>

I've also been on the phone to the Fairchild guys. Took about ten of them
guys before one finally had a clue about the .0086

Take a look at the QVE11223 at Fairchild:
http://www.fairchildsemi.com/ds/QV/QVE11233.pdf
On page 2. VCEsat of 0.4V is also only guaranteed for a phototransistor
current of 0.25mA. hmmm, this looks quite familiar, eh? And it will work
just fine with a phototransistor current of upto 30mA.

The tech said the electrical specs for the two (QVE11233 vs .0086) would be
identical. Only the aperture sizes changed.

HOWEVER, after looking at the spec sheets for the .0086 that somebody else
posted (I never saw these until a couple minutes ago) -- At higher currents,
the QVE11223.0086 will maintain a lower VCE(SAT) than the OPB804. At 10mA
(just over the WMS current) - the .0086 has a VCE(SAT) of 0.8V max. This is
the point not mentioned by the Fairchild tech. The OPB doesn't list the
VCE(SAT) at this voltage but extrapolation would put it pretty high. You'd
still think this would be lower than the 6 volts expected by the comparator,
though.

So, based on what I saw with this spec sheet - the guy I talked to at
Fairchild apparently didn't have as much of a clue as I hoped.

-- Ed .. still shaking his head at how badly designed the williams boards
were... Thank God these guys don't design military or life support
hardware!



> Please let me know where I am wrong.
>
> Jim Knight
>
>
> "GPE" <See_my_website_for_email@cox.net> wrote in message
> news:XskBe.16235$fV.3416@okepread06...
>>
>> The OPB804's are not rated 0.5mA but are rated far higher. The worst
>> case current of 30mA isn't even approached in 'normal' operation
>> (assuming no coil shorts to switch matrix, etc). Within the TZ clock,
>> these are operated at (12volts / 1200 ohms) = 10mA. The 1200 ohm value
>> is the switch matrix's LM339 input pullup resistor to 12 volts.
>> Actually, the current will be less than 10mA as you need to subtract
>> switch column diode voltage drop (D5 on clock board) and the switch
>> driver ULN2803 transistor array's VOL from the 12 volts. Actual current
>> flow is probably closer to 8mA. So, both OPB804 and QVE11233 can handle
>> the current just fine.
>>
>> Both QVE11233 and OPB804 are rated at the exact same VCE(sat) of 0.4
>> volts - plenty low enough for the LM339 comparator. Both are spec'd at
>> this voltage with a diode forward current of 20mA and a transistor
>> collector of 250uA. Both have nearly identical current transfer ratio's
>> (and both are pretty bad). Electrically, they are quite interchangeable.
>>
>> The main differences between the standard QVE11233 and the OPB804 is the
>> slit width (0.125 versus 0.150), the apertures for the QVE11233 are 0.05
>> and 0.01 for the emitter & detector whereas the OPB804 is wide open on
>> both emitter and detector. And a few other minor package dimensions that
>> don't matter. The aperture's is probably what makes the biggest
>> difference here. The OPB804 has a wider gap AND a wide open detector and
>> therefore picks up interference IR from many directions. There are also
>> a few absolute max ratings but none that make any difference in this
>> application.
>>
>> The difference with the QT parts as ordered by Williams, so I was told by
>> tech support at Fairchild (who owns QT Opto), between the QVE11233 and
>> the QVE11233.0086 is *only* the receiver sensitivity. And the
>> sensitivity adjustment was performed by adjusting the sensor's aperture
>> size. This is used to reduce IR interference from other sources... i.e.
>> lamps. I would guess that the reason an OPB804 doesn't work properly is
>> due to excessive IR interferrence. This excessive IR interferrence is
>> forcing the OPB804 into the linear zone where it's conducting but not
>> going into saturation - forcing the output voltage going into the LM339
>> to be at some mid level rather than a distinct 'on' or 'off' voltage.
>>
>> The thing that makes the QVE11233.0086 so hard to replace with a newer
>> and more sensitive opto is the lead spacing. Most opto's are made with
>> 0.32" and 0.22" lead spacings. The QVE11233 has a 0.3" lead spacing.
>>
>> -- Ed
>>
>>
>>
>> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
>> news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
>>> Nicolas,
>>>
>>> You cannot assume the OPB804's will work in any WMS application. The
>>> OPB804s are rated at 0.5mA of current in the detector side. WMS tries
>>> to pump 12.0mA into there. WMS actually had Quality Technologies (QT)
>>> do a special fabrication which was stamped .0086. If you were to change
>>> your OPB804s to QVE1223.0086 you problem will go away!
>>>
>>> Jim Knight - Who is painfully aware of the differences between OPBs and
>>> QVEs.
>>>
>>> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
>>> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
>>>> Ok, before I get hate mail, let me tell you that I know this subject
>>>> has
>>>> been discussed to death before. But I just can't find a good
>>>> explanation to
>>>> my problem. Please help me!
>>>>
>>>> My clock's been broken for the longest and I need it fixed now. I made
>>>> my
>>>> own minutes board (with the extra diodes) to replace the old burnt
>>>> board.
>>>> Replaced all the optos with OPB804 (I know it's not supposed to be a
>>>> good
>>>> replacement part). These actually work wonderfully as the hour optos
>>>> but
>>>> just not as the minutes. Well actually that's not completely true, they
>>>> do
>>>> work when I run the clock test and trigger the optos with a piece of
>>>> paper,
>>>> I can see all 8 working well. Once I put the clock back together and
>>>> run the
>>>> clock test (doing forward), the 15 and 30 minute optos keep on
>>>> flickering
>>>> all the time for no reason. So what's the deal here? Why aren't the
>>>> hour
>>>> optos doing the same? I thought it might have been RF interference from
>>>> the
>>>> motor so I tried a ceramic cap on it, no luck. Other problem, I don't
>>>> quite
>>>> understand the thing with this sensitivity problem, I can't even find a
>>>> friggin datasheet with graphs about this component (anybody has it?).
>>>> Why is
>>>> it that the .0086 sensitivity is required? How will the circuit differ
>>>> if I
>>>> use the OPB804 (electrically)? What I figure is that the
>>>> phototransistor
>>>> will trigger easier than with lower sensitivity, but that doesn't quite
>>>> make
>>>> sense since my OPB804 is triggerring all the time for no reason (I
>>>> thought
>>>> it had a lesser sensitivity...). I heard those 339s die all the time,
>>>> could
>>>> this possibly be the case?
>>>>
>>>> Any comment or answer is highly appreciated!
>>>> Thanks
>>>> Nic
>>>>
>>>>
>>>
>>>
>>
>>
>
>
Anonymous
July 14, 2005 3:48:38 AM

Archived from groups: rec.games.pinball (More info?)

;-)

Jim Knight

PS Do you want me to send you the spec for the QVE11223.0086? It took me
about 1 year to get it and at least 10 service calls.

Optek has sent me specially built OPB804 type prototypes that meet the WMS
spec, but, without an order of x00k of them, we are out of luck.

All of my newly layed out boards with optos (if any) will have
pre-drivers. I agree, what new grad at WMS designed these into ALL of their
optos.

Jim Knight


"GPE" <See_my_website_for_email@cox.net> wrote in message
news:a5lBe.16244$fV.11459@okepread06...
>
> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
> news:E9ednZX5gb6WREjfRVn-1w@comcast.com...
>> Ed,
>> Been on the phone with Fairchild engineers for long periods of time as
>> well. Here is the data sheet:
>>
>> http://www.optekinc.com/pdf/OPB804.pdf
>>
>> On page 2. VCEsat of 0.4V is only guaranteed for phototransistor current
>> of 0.25mA. The WMS currents are on the order of 12mA. I have measured
>> the VCEsat in WMS games at over 2V.
>>
>
> I've also been on the phone to the Fairchild guys. Took about ten of them
> guys before one finally had a clue about the .0086
>
> Take a look at the QVE11223 at Fairchild:
> http://www.fairchildsemi.com/ds/QV/QVE11233.pdf
> On page 2. VCEsat of 0.4V is also only guaranteed for a phototransistor
> current of 0.25mA. hmmm, this looks quite familiar, eh? And it will work
> just fine with a phototransistor current of upto 30mA.
>
> The tech said the electrical specs for the two (QVE11233 vs .0086) would
> be identical. Only the aperture sizes changed.
>
> HOWEVER, after looking at the spec sheets for the .0086 that somebody else
> posted (I never saw these until a couple minutes ago) -- At higher
> currents, the QVE11223.0086 will maintain a lower VCE(SAT) than the
> OPB804. At 10mA (just over the WMS current) - the .0086 has a VCE(SAT) of
> 0.8V max. This is the point not mentioned by the Fairchild tech. The OPB
> doesn't list the VCE(SAT) at this voltage but extrapolation would put it
> pretty high. You'd still think this would be lower than the 6 volts
> expected by the comparator, though.
>
> So, based on what I saw with this spec sheet - the guy I talked to at
> Fairchild apparently didn't have as much of a clue as I hoped.
>
> -- Ed .. still shaking his head at how badly designed the williams boards
> were... Thank God these guys don't design military or life support
> hardware!
>
>
>
>> Please let me know where I am wrong.
>>
>> Jim Knight
>>
>>
>> "GPE" <See_my_website_for_email@cox.net> wrote in message
>> news:XskBe.16235$fV.3416@okepread06...
>>>
>>> The OPB804's are not rated 0.5mA but are rated far higher. The worst
>>> case current of 30mA isn't even approached in 'normal' operation
>>> (assuming no coil shorts to switch matrix, etc). Within the TZ clock,
>>> these are operated at (12volts / 1200 ohms) = 10mA. The 1200 ohm value
>>> is the switch matrix's LM339 input pullup resistor to 12 volts.
>>> Actually, the current will be less than 10mA as you need to subtract
>>> switch column diode voltage drop (D5 on clock board) and the switch
>>> driver ULN2803 transistor array's VOL from the 12 volts. Actual current
>>> flow is probably closer to 8mA. So, both OPB804 and QVE11233 can handle
>>> the current just fine.
>>>
>>> Both QVE11233 and OPB804 are rated at the exact same VCE(sat) of 0.4
>>> volts - plenty low enough for the LM339 comparator. Both are spec'd at
>>> this voltage with a diode forward current of 20mA and a transistor
>>> collector of 250uA. Both have nearly identical current transfer ratio's
>>> (and both are pretty bad). Electrically, they are quite
>>> interchangeable.
>>>
>>> The main differences between the standard QVE11233 and the OPB804 is the
>>> slit width (0.125 versus 0.150), the apertures for the QVE11233 are 0.05
>>> and 0.01 for the emitter & detector whereas the OPB804 is wide open on
>>> both emitter and detector. And a few other minor package dimensions
>>> that don't matter. The aperture's is probably what makes the biggest
>>> difference here. The OPB804 has a wider gap AND a wide open detector and
>>> therefore picks up interference IR from many directions. There are also
>>> a few absolute max ratings but none that make any difference in this
>>> application.
>>>
>>> The difference with the QT parts as ordered by Williams, so I was told
>>> by tech support at Fairchild (who owns QT Opto), between the QVE11233
>>> and the QVE11233.0086 is *only* the receiver sensitivity. And the
>>> sensitivity adjustment was performed by adjusting the sensor's aperture
>>> size. This is used to reduce IR interference from other sources... i.e.
>>> lamps. I would guess that the reason an OPB804 doesn't work properly is
>>> due to excessive IR interferrence. This excessive IR interferrence is
>>> forcing the OPB804 into the linear zone where it's conducting but not
>>> going into saturation - forcing the output voltage going into the LM339
>>> to be at some mid level rather than a distinct 'on' or 'off' voltage.
>>>
>>> The thing that makes the QVE11233.0086 so hard to replace with a newer
>>> and more sensitive opto is the lead spacing. Most opto's are made with
>>> 0.32" and 0.22" lead spacings. The QVE11233 has a 0.3" lead spacing.
>>>
>>> -- Ed
>>>
>>>
>>>
>>> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
>>> news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
>>>> Nicolas,
>>>>
>>>> You cannot assume the OPB804's will work in any WMS application. The
>>>> OPB804s are rated at 0.5mA of current in the detector side. WMS tries
>>>> to pump 12.0mA into there. WMS actually had Quality Technologies (QT)
>>>> do a special fabrication which was stamped .0086. If you were to
>>>> change your OPB804s to QVE1223.0086 you problem will go away!
>>>>
>>>> Jim Knight - Who is painfully aware of the differences between OPBs and
>>>> QVEs.
>>>>
>>>> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
>>>> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
>>>>> Ok, before I get hate mail, let me tell you that I know this subject
>>>>> has
>>>>> been discussed to death before. But I just can't find a good
>>>>> explanation to
>>>>> my problem. Please help me!
>>>>>
>>>>> My clock's been broken for the longest and I need it fixed now. I made
>>>>> my
>>>>> own minutes board (with the extra diodes) to replace the old burnt
>>>>> board.
>>>>> Replaced all the optos with OPB804 (I know it's not supposed to be a
>>>>> good
>>>>> replacement part). These actually work wonderfully as the hour optos
>>>>> but
>>>>> just not as the minutes. Well actually that's not completely true,
>>>>> they do
>>>>> work when I run the clock test and trigger the optos with a piece of
>>>>> paper,
>>>>> I can see all 8 working well. Once I put the clock back together and
>>>>> run the
>>>>> clock test (doing forward), the 15 and 30 minute optos keep on
>>>>> flickering
>>>>> all the time for no reason. So what's the deal here? Why aren't the
>>>>> hour
>>>>> optos doing the same? I thought it might have been RF interference
>>>>> from the
>>>>> motor so I tried a ceramic cap on it, no luck. Other problem, I don't
>>>>> quite
>>>>> understand the thing with this sensitivity problem, I can't even find
>>>>> a
>>>>> friggin datasheet with graphs about this component (anybody has it?).
>>>>> Why is
>>>>> it that the .0086 sensitivity is required? How will the circuit differ
>>>>> if I
>>>>> use the OPB804 (electrically)? What I figure is that the
>>>>> phototransistor
>>>>> will trigger easier than with lower sensitivity, but that doesn't
>>>>> quite make
>>>>> sense since my OPB804 is triggerring all the time for no reason (I
>>>>> thought
>>>>> it had a lesser sensitivity...). I heard those 339s die all the time,
>>>>> could
>>>>> this possibly be the case?
>>>>>
>>>>> Any comment or answer is highly appreciated!
>>>>> Thanks
>>>>> Nic
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
Anonymous
July 14, 2005 3:48:39 AM

Archived from groups: rec.games.pinball (More info?)

If you have a softcopy of the spec that's printable - that would be great.
GPE.at.cox.net

-- Ed.


"Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
news:sp-dnQBoX_8EfUjfRVn-tw@comcast.com...
> ;-)
>
> Jim Knight
>
> PS Do you want me to send you the spec for the QVE11223.0086? It took me
> about 1 year to get it and at least 10 service calls.
>
> Optek has sent me specially built OPB804 type prototypes that meet the WMS
> spec, but, without an order of x00k of them, we are out of luck.
>
> All of my newly layed out boards with optos (if any) will have
> pre-drivers. I agree, what new grad at WMS designed these into ALL of
> their optos.
>
> Jim Knight
>
>
> "GPE" <See_my_website_for_email@cox.net> wrote in message
> news:a5lBe.16244$fV.11459@okepread06...
>>
>> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
>> news:E9ednZX5gb6WREjfRVn-1w@comcast.com...
>>> Ed,
>>> Been on the phone with Fairchild engineers for long periods of time as
>>> well. Here is the data sheet:
>>>
>>> http://www.optekinc.com/pdf/OPB804.pdf
>>>
>>> On page 2. VCEsat of 0.4V is only guaranteed for phototransistor
>>> current of 0.25mA. The WMS currents are on the order of 12mA. I have
>>> measured the VCEsat in WMS games at over 2V.
>>>
>>
>> I've also been on the phone to the Fairchild guys. Took about ten of
>> them guys before one finally had a clue about the .0086
>>
>> Take a look at the QVE11223 at Fairchild:
>> http://www.fairchildsemi.com/ds/QV/QVE11233.pdf
>> On page 2. VCEsat of 0.4V is also only guaranteed for a phototransistor
>> current of 0.25mA. hmmm, this looks quite familiar, eh? And it will
>> work just fine with a phototransistor current of upto 30mA.
>>
>> The tech said the electrical specs for the two (QVE11233 vs .0086) would
>> be identical. Only the aperture sizes changed.
>>
>> HOWEVER, after looking at the spec sheets for the .0086 that somebody
>> else posted (I never saw these until a couple minutes ago) -- At higher
>> currents, the QVE11223.0086 will maintain a lower VCE(SAT) than the
>> OPB804. At 10mA (just over the WMS current) - the .0086 has a VCE(SAT)
>> of 0.8V max. This is the point not mentioned by the Fairchild tech. The
>> OPB doesn't list the VCE(SAT) at this voltage but extrapolation would put
>> it pretty high. You'd still think this would be lower than the 6 volts
>> expected by the comparator, though.
>>
>> So, based on what I saw with this spec sheet - the guy I talked to at
>> Fairchild apparently didn't have as much of a clue as I hoped.
>>
>> -- Ed .. still shaking his head at how badly designed the williams boards
>> were... Thank God these guys don't design military or life support
>> hardware!
>>
>>
>>
>>> Please let me know where I am wrong.
>>>
>>> Jim Knight
>>>
>>>
>>> "GPE" <See_my_website_for_email@cox.net> wrote in message
>>> news:XskBe.16235$fV.3416@okepread06...
>>>>
>>>> The OPB804's are not rated 0.5mA but are rated far higher. The worst
>>>> case current of 30mA isn't even approached in 'normal' operation
>>>> (assuming no coil shorts to switch matrix, etc). Within the TZ clock,
>>>> these are operated at (12volts / 1200 ohms) = 10mA. The 1200 ohm value
>>>> is the switch matrix's LM339 input pullup resistor to 12 volts.
>>>> Actually, the current will be less than 10mA as you need to subtract
>>>> switch column diode voltage drop (D5 on clock board) and the switch
>>>> driver ULN2803 transistor array's VOL from the 12 volts. Actual
>>>> current flow is probably closer to 8mA. So, both OPB804 and QVE11233
>>>> can handle the current just fine.
>>>>
>>>> Both QVE11233 and OPB804 are rated at the exact same VCE(sat) of 0.4
>>>> volts - plenty low enough for the LM339 comparator. Both are spec'd at
>>>> this voltage with a diode forward current of 20mA and a transistor
>>>> collector of 250uA. Both have nearly identical current transfer ratio's
>>>> (and both are pretty bad). Electrically, they are quite
>>>> interchangeable.
>>>>
>>>> The main differences between the standard QVE11233 and the OPB804 is
>>>> the slit width (0.125 versus 0.150), the apertures for the QVE11233 are
>>>> 0.05 and 0.01 for the emitter & detector whereas the OPB804 is wide
>>>> open on both emitter and detector. And a few other minor package
>>>> dimensions that don't matter. The aperture's is probably what makes
>>>> the biggest difference here. The OPB804 has a wider gap AND a wide open
>>>> detector and therefore picks up interference IR from many directions.
>>>> There are also a few absolute max ratings but none that make any
>>>> difference in this application.
>>>>
>>>> The difference with the QT parts as ordered by Williams, so I was told
>>>> by tech support at Fairchild (who owns QT Opto), between the QVE11233
>>>> and the QVE11233.0086 is *only* the receiver sensitivity. And the
>>>> sensitivity adjustment was performed by adjusting the sensor's aperture
>>>> size. This is used to reduce IR interference from other sources...
>>>> i.e. lamps. I would guess that the reason an OPB804 doesn't work
>>>> properly is due to excessive IR interferrence. This excessive IR
>>>> interferrence is forcing the OPB804 into the linear zone where it's
>>>> conducting but not going into saturation - forcing the output voltage
>>>> going into the LM339 to be at some mid level rather than a distinct
>>>> 'on' or 'off' voltage.
>>>>
>>>> The thing that makes the QVE11233.0086 so hard to replace with a newer
>>>> and more sensitive opto is the lead spacing. Most opto's are made with
>>>> 0.32" and 0.22" lead spacings. The QVE11233 has a 0.3" lead spacing.
>>>>
>>>> -- Ed
>>>>
>>>>
>>>>
>>>> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
>>>> news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
>>>>> Nicolas,
>>>>>
>>>>> You cannot assume the OPB804's will work in any WMS application. The
>>>>> OPB804s are rated at 0.5mA of current in the detector side. WMS tries
>>>>> to pump 12.0mA into there. WMS actually had Quality Technologies (QT)
>>>>> do a special fabrication which was stamped .0086. If you were to
>>>>> change your OPB804s to QVE1223.0086 you problem will go away!
>>>>>
>>>>> Jim Knight - Who is painfully aware of the differences between OPBs
>>>>> and QVEs.
>>>>>
>>>>> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
>>>>> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
>>>>>> Ok, before I get hate mail, let me tell you that I know this subject
>>>>>> has
>>>>>> been discussed to death before. But I just can't find a good
>>>>>> explanation to
>>>>>> my problem. Please help me!
>>>>>>
>>>>>> My clock's been broken for the longest and I need it fixed now. I
>>>>>> made my
>>>>>> own minutes board (with the extra diodes) to replace the old burnt
>>>>>> board.
>>>>>> Replaced all the optos with OPB804 (I know it's not supposed to be a
>>>>>> good
>>>>>> replacement part). These actually work wonderfully as the hour optos
>>>>>> but
>>>>>> just not as the minutes. Well actually that's not completely true,
>>>>>> they do
>>>>>> work when I run the clock test and trigger the optos with a piece of
>>>>>> paper,
>>>>>> I can see all 8 working well. Once I put the clock back together and
>>>>>> run the
>>>>>> clock test (doing forward), the 15 and 30 minute optos keep on
>>>>>> flickering
>>>>>> all the time for no reason. So what's the deal here? Why aren't the
>>>>>> hour
>>>>>> optos doing the same? I thought it might have been RF interference
>>>>>> from the
>>>>>> motor so I tried a ceramic cap on it, no luck. Other problem, I don't
>>>>>> quite
>>>>>> understand the thing with this sensitivity problem, I can't even find
>>>>>> a
>>>>>> friggin datasheet with graphs about this component (anybody has it?).
>>>>>> Why is
>>>>>> it that the .0086 sensitivity is required? How will the circuit
>>>>>> differ if I
>>>>>> use the OPB804 (electrically)? What I figure is that the
>>>>>> phototransistor
>>>>>> will trigger easier than with lower sensitivity, but that doesn't
>>>>>> quite make
>>>>>> sense since my OPB804 is triggerring all the time for no reason (I
>>>>>> thought
>>>>>> it had a lesser sensitivity...). I heard those 339s die all the time,
>>>>>> could
>>>>>> this possibly be the case?
>>>>>>
>>>>>> Any comment or answer is highly appreciated!
>>>>>> Thanks
>>>>>> Nic
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
Anonymous
July 14, 2005 3:50:39 AM

Archived from groups: rec.games.pinball (More info?)

That would be entirely a matter of opinion
I'm sure.

(Not mine, though.)

--
Fred
TX
CARGPB#8
******************


"Nwojedi" <nwojedi@hotmail.com> wrote
>
> if you don't want to pay pbliz board price. why not buy from
> rottendog. His are around $75 and better design.
Anonymous
July 14, 2005 4:50:50 AM

Archived from groups: rec.games.pinball (More info?)

Will send out a url tomorow to group(it is on my office machine), and also
send you a direct pdf copy.

Jim Knight

"GPE" <See_my_website_for_email@cox.net> wrote in message
news:iplBe.16249$fV.2877@okepread06...
> If you have a softcopy of the spec that's printable - that would be great.
> GPE.at.cox.net
>
> -- Ed.
>
>
> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
> news:sp-dnQBoX_8EfUjfRVn-tw@comcast.com...
>> ;-)
>>
>> Jim Knight
>>
>> PS Do you want me to send you the spec for the QVE11223.0086? It took me
>> about 1 year to get it and at least 10 service calls.
>>
>> Optek has sent me specially built OPB804 type prototypes that meet the
>> WMS spec, but, without an order of x00k of them, we are out of luck.
>>
>> All of my newly layed out boards with optos (if any) will have
>> pre-drivers. I agree, what new grad at WMS designed these into ALL of
>> their optos.
>>
>> Jim Knight
>>
>>
>> "GPE" <See_my_website_for_email@cox.net> wrote in message
>> news:a5lBe.16244$fV.11459@okepread06...
>>>
>>> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
>>> news:E9ednZX5gb6WREjfRVn-1w@comcast.com...
>>>> Ed,
>>>> Been on the phone with Fairchild engineers for long periods of time as
>>>> well. Here is the data sheet:
>>>>
>>>> http://www.optekinc.com/pdf/OPB804.pdf
>>>>
>>>> On page 2. VCEsat of 0.4V is only guaranteed for phototransistor
>>>> current of 0.25mA. The WMS currents are on the order of 12mA. I have
>>>> measured the VCEsat in WMS games at over 2V.
>>>>
>>>
>>> I've also been on the phone to the Fairchild guys. Took about ten of
>>> them guys before one finally had a clue about the .0086
>>>
>>> Take a look at the QVE11223 at Fairchild:
>>> http://www.fairchildsemi.com/ds/QV/QVE11233.pdf
>>> On page 2. VCEsat of 0.4V is also only guaranteed for a phototransistor
>>> current of 0.25mA. hmmm, this looks quite familiar, eh? And it will
>>> work just fine with a phototransistor current of upto 30mA.
>>>
>>> The tech said the electrical specs for the two (QVE11233 vs .0086) would
>>> be identical. Only the aperture sizes changed.
>>>
>>> HOWEVER, after looking at the spec sheets for the .0086 that somebody
>>> else posted (I never saw these until a couple minutes ago) -- At higher
>>> currents, the QVE11223.0086 will maintain a lower VCE(SAT) than the
>>> OPB804. At 10mA (just over the WMS current) - the .0086 has a VCE(SAT)
>>> of 0.8V max. This is the point not mentioned by the Fairchild tech.
>>> The OPB doesn't list the VCE(SAT) at this voltage but extrapolation
>>> would put it pretty high. You'd still think this would be lower than
>>> the 6 volts expected by the comparator, though.
>>>
>>> So, based on what I saw with this spec sheet - the guy I talked to at
>>> Fairchild apparently didn't have as much of a clue as I hoped.
>>>
>>> -- Ed .. still shaking his head at how badly designed the williams
>>> boards were... Thank God these guys don't design military or life
>>> support hardware!
>>>
>>>
>>>
>>>> Please let me know where I am wrong.
>>>>
>>>> Jim Knight
>>>>
>>>>
>>>> "GPE" <See_my_website_for_email@cox.net> wrote in message
>>>> news:XskBe.16235$fV.3416@okepread06...
>>>>>
>>>>> The OPB804's are not rated 0.5mA but are rated far higher. The worst
>>>>> case current of 30mA isn't even approached in 'normal' operation
>>>>> (assuming no coil shorts to switch matrix, etc). Within the TZ clock,
>>>>> these are operated at (12volts / 1200 ohms) = 10mA. The 1200 ohm
>>>>> value is the switch matrix's LM339 input pullup resistor to 12 volts.
>>>>> Actually, the current will be less than 10mA as you need to subtract
>>>>> switch column diode voltage drop (D5 on clock board) and the switch
>>>>> driver ULN2803 transistor array's VOL from the 12 volts. Actual
>>>>> current flow is probably closer to 8mA. So, both OPB804 and QVE11233
>>>>> can handle the current just fine.
>>>>>
>>>>> Both QVE11233 and OPB804 are rated at the exact same VCE(sat) of 0.4
>>>>> volts - plenty low enough for the LM339 comparator. Both are spec'd
>>>>> at this voltage with a diode forward current of 20mA and a transistor
>>>>> collector of 250uA. Both have nearly identical current transfer
>>>>> ratio's (and both are pretty bad). Electrically, they are quite
>>>>> interchangeable.
>>>>>
>>>>> The main differences between the standard QVE11233 and the OPB804 is
>>>>> the slit width (0.125 versus 0.150), the apertures for the QVE11233
>>>>> are 0.05 and 0.01 for the emitter & detector whereas the OPB804 is
>>>>> wide open on both emitter and detector. And a few other minor package
>>>>> dimensions that don't matter. The aperture's is probably what makes
>>>>> the biggest difference here. The OPB804 has a wider gap AND a wide
>>>>> open detector and therefore picks up interference IR from many
>>>>> directions. There are also a few absolute max ratings but none that
>>>>> make any difference in this application.
>>>>>
>>>>> The difference with the QT parts as ordered by Williams, so I was told
>>>>> by tech support at Fairchild (who owns QT Opto), between the QVE11233
>>>>> and the QVE11233.0086 is *only* the receiver sensitivity. And the
>>>>> sensitivity adjustment was performed by adjusting the sensor's
>>>>> aperture size. This is used to reduce IR interference from other
>>>>> sources... i.e. lamps. I would guess that the reason an OPB804
>>>>> doesn't work properly is due to excessive IR interferrence. This
>>>>> excessive IR interferrence is forcing the OPB804 into the linear zone
>>>>> where it's conducting but not going into saturation - forcing the
>>>>> output voltage going into the LM339 to be at some mid level rather
>>>>> than a distinct 'on' or 'off' voltage.
>>>>>
>>>>> The thing that makes the QVE11233.0086 so hard to replace with a newer
>>>>> and more sensitive opto is the lead spacing. Most opto's are made
>>>>> with 0.32" and 0.22" lead spacings. The QVE11233 has a 0.3" lead
>>>>> spacing.
>>>>>
>>>>> -- Ed
>>>>>
>>>>>
>>>>>
>>>>> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
>>>>> news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
>>>>>> Nicolas,
>>>>>>
>>>>>> You cannot assume the OPB804's will work in any WMS application.
>>>>>> The OPB804s are rated at 0.5mA of current in the detector side. WMS
>>>>>> tries to pump 12.0mA into there. WMS actually had Quality
>>>>>> Technologies (QT) do a special fabrication which was stamped .0086.
>>>>>> If you were to change your OPB804s to QVE1223.0086 you problem will
>>>>>> go away!
>>>>>>
>>>>>> Jim Knight - Who is painfully aware of the differences between OPBs
>>>>>> and QVEs.
>>>>>>
>>>>>> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
>>>>>> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
>>>>>>> Ok, before I get hate mail, let me tell you that I know this subject
>>>>>>> has
>>>>>>> been discussed to death before. But I just can't find a good
>>>>>>> explanation to
>>>>>>> my problem. Please help me!
>>>>>>>
>>>>>>> My clock's been broken for the longest and I need it fixed now. I
>>>>>>> made my
>>>>>>> own minutes board (with the extra diodes) to replace the old burnt
>>>>>>> board.
>>>>>>> Replaced all the optos with OPB804 (I know it's not supposed to be a
>>>>>>> good
>>>>>>> replacement part). These actually work wonderfully as the hour optos
>>>>>>> but
>>>>>>> just not as the minutes. Well actually that's not completely true,
>>>>>>> they do
>>>>>>> work when I run the clock test and trigger the optos with a piece of
>>>>>>> paper,
>>>>>>> I can see all 8 working well. Once I put the clock back together and
>>>>>>> run the
>>>>>>> clock test (doing forward), the 15 and 30 minute optos keep on
>>>>>>> flickering
>>>>>>> all the time for no reason. So what's the deal here? Why aren't the
>>>>>>> hour
>>>>>>> optos doing the same? I thought it might have been RF interference
>>>>>>> from the
>>>>>>> motor so I tried a ceramic cap on it, no luck. Other problem, I
>>>>>>> don't quite
>>>>>>> understand the thing with this sensitivity problem, I can't even
>>>>>>> find a
>>>>>>> friggin datasheet with graphs about this component (anybody has
>>>>>>> it?). Why is
>>>>>>> it that the .0086 sensitivity is required? How will the circuit
>>>>>>> differ if I
>>>>>>> use the OPB804 (electrically)? What I figure is that the
>>>>>>> phototransistor
>>>>>>> will trigger easier than with lower sensitivity, but that doesn't
>>>>>>> quite make
>>>>>>> sense since my OPB804 is triggerring all the time for no reason (I
>>>>>>> thought
>>>>>>> it had a lesser sensitivity...). I heard those 339s die all the
>>>>>>> time, could
>>>>>>> this possibly be the case?
>>>>>>>
>>>>>>> Any comment or answer is highly appreciated!
>>>>>>> Thanks
>>>>>>> Nic
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
Anonymous
July 15, 2005 1:33:23 AM

Archived from groups: rec.games.pinball (More info?)

Has anybody tried using Fairchild H22B2's in place of these QVE11233.0086's?
Close to same size and good specs. Rated VCEsat of 1.5 volts with an IC of
12mA -- more than enough to pull the LM339 input below the 5 volt threshold.

I'll probably be buying a set of these with my next Fairchild parts buy.
Just gotta find a guinee pig machine to try them on...

-- Ed


"Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
news:D J6dnUGMGquxckjfRVn-3g@comcast.com...
> Will send out a url tomorow to group(it is on my office machine), and also
> send you a direct pdf copy.
>
> Jim Knight
>
> "GPE" <See_my_website_for_email@cox.net> wrote in message
> news:iplBe.16249$fV.2877@okepread06...
>> If you have a softcopy of the spec that's printable - that would be
>> great.
>> GPE.at.cox.net
>>
>> -- Ed.
>>
>>
>> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
>> news:sp-dnQBoX_8EfUjfRVn-tw@comcast.com...
>>> ;-)
>>>
>>> Jim Knight
>>>
>>> PS Do you want me to send you the spec for the QVE11223.0086? It took
>>> me about 1 year to get it and at least 10 service calls.
>>>
>>> Optek has sent me specially built OPB804 type prototypes that meet the
>>> WMS spec, but, without an order of x00k of them, we are out of luck.
>>>
>>> All of my newly layed out boards with optos (if any) will have
>>> pre-drivers. I agree, what new grad at WMS designed these into ALL of
>>> their optos.
>>>
>>> Jim Knight
>>>
>>>
>>> "GPE" <See_my_website_for_email@cox.net> wrote in message
>>> news:a5lBe.16244$fV.11459@okepread06...
>>>>
>>>> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
>>>> news:E9ednZX5gb6WREjfRVn-1w@comcast.com...
>>>>> Ed,
>>>>> Been on the phone with Fairchild engineers for long periods of time
>>>>> as well. Here is the data sheet:
>>>>>
>>>>> http://www.optekinc.com/pdf/OPB804.pdf
>>>>>
>>>>> On page 2. VCEsat of 0.4V is only guaranteed for phototransistor
>>>>> current of 0.25mA. The WMS currents are on the order of 12mA. I have
>>>>> measured the VCEsat in WMS games at over 2V.
>>>>>
>>>>
>>>> I've also been on the phone to the Fairchild guys. Took about ten of
>>>> them guys before one finally had a clue about the .0086
>>>>
>>>> Take a look at the QVE11223 at Fairchild:
>>>> http://www.fairchildsemi.com/ds/QV/QVE11233.pdf
>>>> On page 2. VCEsat of 0.4V is also only guaranteed for a phototransistor
>>>> current of 0.25mA. hmmm, this looks quite familiar, eh? And it will
>>>> work just fine with a phototransistor current of upto 30mA.
>>>>
>>>> The tech said the electrical specs for the two (QVE11233 vs .0086)
>>>> would be identical. Only the aperture sizes changed.
>>>>
>>>> HOWEVER, after looking at the spec sheets for the .0086 that somebody
>>>> else posted (I never saw these until a couple minutes ago) -- At higher
>>>> currents, the QVE11223.0086 will maintain a lower VCE(SAT) than the
>>>> OPB804. At 10mA (just over the WMS current) - the .0086 has a VCE(SAT)
>>>> of 0.8V max. This is the point not mentioned by the Fairchild tech.
>>>> The OPB doesn't list the VCE(SAT) at this voltage but extrapolation
>>>> would put it pretty high. You'd still think this would be lower than
>>>> the 6 volts expected by the comparator, though.
>>>>
>>>> So, based on what I saw with this spec sheet - the guy I talked to at
>>>> Fairchild apparently didn't have as much of a clue as I hoped.
>>>>
>>>> -- Ed .. still shaking his head at how badly designed the williams
>>>> boards were... Thank God these guys don't design military or life
>>>> support hardware!
>>>>
>>>>
>>>>
>>>>> Please let me know where I am wrong.
>>>>>
>>>>> Jim Knight
>>>>>
>>>>>
>>>>> "GPE" <See_my_website_for_email@cox.net> wrote in message
>>>>> news:XskBe.16235$fV.3416@okepread06...
>>>>>>
>>>>>> The OPB804's are not rated 0.5mA but are rated far higher. The worst
>>>>>> case current of 30mA isn't even approached in 'normal' operation
>>>>>> (assuming no coil shorts to switch matrix, etc). Within the TZ
>>>>>> clock, these are operated at (12volts / 1200 ohms) = 10mA. The 1200
>>>>>> ohm value is the switch matrix's LM339 input pullup resistor to 12
>>>>>> volts. Actually, the current will be less than 10mA as you need to
>>>>>> subtract switch column diode voltage drop (D5 on clock board) and the
>>>>>> switch driver ULN2803 transistor array's VOL from the 12 volts.
>>>>>> Actual current flow is probably closer to 8mA. So, both OPB804 and
>>>>>> QVE11233 can handle the current just fine.
>>>>>>
>>>>>> Both QVE11233 and OPB804 are rated at the exact same VCE(sat) of 0.4
>>>>>> volts - plenty low enough for the LM339 comparator. Both are spec'd
>>>>>> at this voltage with a diode forward current of 20mA and a transistor
>>>>>> collector of 250uA. Both have nearly identical current transfer
>>>>>> ratio's (and both are pretty bad). Electrically, they are quite
>>>>>> interchangeable.
>>>>>>
>>>>>> The main differences between the standard QVE11233 and the OPB804 is
>>>>>> the slit width (0.125 versus 0.150), the apertures for the QVE11233
>>>>>> are 0.05 and 0.01 for the emitter & detector whereas the OPB804 is
>>>>>> wide open on both emitter and detector. And a few other minor
>>>>>> package dimensions that don't matter. The aperture's is probably
>>>>>> what makes the biggest difference here. The OPB804 has a wider gap
>>>>>> AND a wide open detector and therefore picks up interference IR from
>>>>>> many directions. There are also a few absolute max ratings but none
>>>>>> that make any difference in this application.
>>>>>>
>>>>>> The difference with the QT parts as ordered by Williams, so I was
>>>>>> told by tech support at Fairchild (who owns QT Opto), between the
>>>>>> QVE11233 and the QVE11233.0086 is *only* the receiver sensitivity.
>>>>>> And the sensitivity adjustment was performed by adjusting the
>>>>>> sensor's aperture size. This is used to reduce IR interference from
>>>>>> other sources... i.e. lamps. I would guess that the reason an OPB804
>>>>>> doesn't work properly is due to excessive IR interferrence. This
>>>>>> excessive IR interferrence is forcing the OPB804 into the linear zone
>>>>>> where it's conducting but not going into saturation - forcing the
>>>>>> output voltage going into the LM339 to be at some mid level rather
>>>>>> than a distinct 'on' or 'off' voltage.
>>>>>>
>>>>>> The thing that makes the QVE11233.0086 so hard to replace with a
>>>>>> newer and more sensitive opto is the lead spacing. Most opto's are
>>>>>> made with 0.32" and 0.22" lead spacings. The QVE11233 has a 0.3"
>>>>>> lead spacing.
>>>>>>
>>>>>> -- Ed
>>>>>>
>>>>>>
>>>>>>
>>>>>> "Jim Knight" <rottendog@NOSPAM.comcast.net> wrote in message
>>>>>> news:HuCdnVkQKsFVwEjfRVn-qA@comcast.com...
>>>>>>> Nicolas,
>>>>>>>
>>>>>>> You cannot assume the OPB804's will work in any WMS application.
>>>>>>> The OPB804s are rated at 0.5mA of current in the detector side. WMS
>>>>>>> tries to pump 12.0mA into there. WMS actually had Quality
>>>>>>> Technologies (QT) do a special fabrication which was stamped .0086.
>>>>>>> If you were to change your OPB804s to QVE1223.0086 you problem will
>>>>>>> go away!
>>>>>>>
>>>>>>> Jim Knight - Who is painfully aware of the differences between OPBs
>>>>>>> and QVEs.
>>>>>>>
>>>>>>> "Nicolas Legare" <no_email_please@hotmail.com> wrote in message
>>>>>>> news:vHbBe.2886$p5.768@nnrp.ca.mci.com!nnrp1.uunet.ca...
>>>>>>>> Ok, before I get hate mail, let me tell you that I know this
>>>>>>>> subject has
>>>>>>>> been discussed to death before. But I just can't find a good
>>>>>>>> explanation to
>>>>>>>> my problem. Please help me!
>>>>>>>>
>>>>>>>> My clock's been broken for the longest and I need it fixed now. I
>>>>>>>> made my
>>>>>>>> own minutes board (with the extra diodes) to replace the old burnt
>>>>>>>> board.
>>>>>>>> Replaced all the optos with OPB804 (I know it's not supposed to be
>>>>>>>> a good
>>>>>>>> replacement part). These actually work wonderfully as the hour
>>>>>>>> optos but
>>>>>>>> just not as the minutes. Well actually that's not completely true,
>>>>>>>> they do
>>>>>>>> work when I run the clock test and trigger the optos with a piece
>>>>>>>> of paper,
>>>>>>>> I can see all 8 working well. Once I put the clock back together
>>>>>>>> and run the
>>>>>>>> clock test (doing forward), the 15 and 30 minute optos keep on
>>>>>>>> flickering
>>>>>>>> all the time for no reason. So what's the deal here? Why aren't the
>>>>>>>> hour
>>>>>>>> optos doing the same? I thought it might have been RF interference
>>>>>>>> from the
>>>>>>>> motor so I tried a ceramic cap on it, no luck. Other problem, I
>>>>>>>> don't quite
>>>>>>>> understand the thing with this sensitivity problem, I can't even
>>>>>>>> find a
>>>>>>>> friggin datasheet with graphs about this component (anybody has
>>>>>>>> it?). Why is
>>>>>>>> it that the .0086 sensitivity is required? How will the circuit
>>>>>>>> differ if I
>>>>>>>> use the OPB804 (electrically)? What I figure is that the
>>>>>>>> phototransistor
>>>>>>>> will trigger easier than with lower sensitivity, but that doesn't
>>>>>>>> quite make
>>>>>>>> sense since my OPB804 is triggerring all the time for no reason (I
>>>>>>>> thought
>>>>>>>> it had a lesser sensitivity...). I heard those 339s die all the
>>>>>>>> time, could
>>>>>>>> this possibly be the case?
>>>>>>>>
>>>>>>>> Any comment or answer is highly appreciated!
>>>>>>>> Thanks
>>>>>>>> Nic
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
July 15, 2005 2:27:50 AM

Archived from groups: rec.games.pinball (More info?)

It is the 10mA minimum on current that is important.
Anonymous
July 15, 2005 2:35:14 AM

Archived from groups: rec.games.pinball (More info?)

I did sometime back.

No-go for that application, (And too short without
some type of rigging.

Still use them for other things and they
work fine.

--
Fred
TX
CARGPB#8
******************



"GPE" <See_my_website_for_email@cox.net> wrote in message news:NfFBe.17291$fV.8839@okepread06...
>
> Has anybody tried using Fairchild H22B2's in place of these QVE11233.0086's?
> Close to same size and good specs. Rated VCEsat of 1.5 volts with an IC of
> 12mA -- more than enough to pull the LM339 input below the 5 volt threshold.
>
> I'll probably be buying a set of these with my next Fairchild parts buy.
> Just gotta find a guinee pig machine to try them on...
>
> -- Ed
>
>
Anonymous
July 15, 2005 3:14:41 AM

Archived from groups: rec.games.pinball (More info?)

Bummer....

-- Ed

"Fred Kemper" <pbgarage@davlin.net> wrote in message
news:D b7au40255h@enews4.newsguy.com...
>I did sometime back.
>
> No-go for that application, (And too short without
> some type of rigging.
>
> Still use them for other things and they
> work fine.
>
> --
> Fred
> TX
> CARGPB#8
> ******************
>
>
>
> "GPE" <See_my_website_for_email@cox.net> wrote in message
> news:NfFBe.17291$fV.8839@okepread06...
>>
>> Has anybody tried using Fairchild H22B2's in place of these
>> QVE11233.0086's?
>> Close to same size and good specs. Rated VCEsat of 1.5 volts with an IC
>> of
>> 12mA -- more than enough to pull the LM339 input below the 5 volt
>> threshold.
>>
>> I'll probably be buying a set of these with my next Fairchild parts buy.
>> Just gotta find a guinee pig machine to try them on...
>>
>> -- Ed
>>
>>
>
>
July 15, 2005 3:51:42 AM

Archived from groups: rec.games.pinball (More info?)

Agreed. But I am not convinced that you can drive these things to
saturation as they only guarantee 5mA. The effect is the same, however:
Whether it is saturating or operating in linear mode, not enough
current equals not enough voltage drop.


The H22B2 does look good, though: in this application, it guarantees
14mA with a VCEsat of less than 1.5V. Accounting for a 5V that could
be as low as 4.7V, and even 500mV of slop in the LM339A, we should
still be 300mV on the side of righteousness.
Anonymous
July 15, 2005 5:20:57 AM

Archived from groups: rec.games.pinball (More info?)

"martin" <martin.reynolds@gmail.com> wrote in message
news:1121405270.683938.96080@g14g2000cwa.googlegroups.com...
> It is the 10mA minimum on current that is important.
>

That, in combination with the very low VCE(sat). As the current rises, the
VCE(sat) rises. The collector is tied directly to a comparator input. When
the comparator input drops below 5 volts, the opto is determined to be 'on'.
When above 5 volts, the opto is determined to be 'off'. But, VCE(sat) isn't
the only voltage drop to the comparator input. You must also add one switch
matrix blocking diode (1N4148) plus one blocking diode located on the TZ
board (D5) plus the VOL of the ULN2803 (also part of the switch matrix).
Voltage drop approx equal (0.7V x 2) + VCE(sat) + 1.1V = VCE(sat) plus 2.4.
VCE(sat) must be less than 5 volts, so VCE(sat) < 5 - 2.4 < 2.6 volts
maxium.
This doesn't account for sloppy wiring and connector drops so it's most
probably less than 2.6 volts. This also doesn't account for granularity of
the LM339's input...which I'm too lazy to lookup tonight.

Vast majority of optocouplers have no problem sinking 10mA or more. The
problem they have is a rapidly rising VCE(sat) as that current exceeds 1mA
or so.

Surprises me that the Fairchild H22B2 doesn't work. It's rated with a max
VCE(Sat) of 1.5 volts while sinking 14mA.

-- Ed
!