Archived from groups: alt.comp.periphs.mainboard.asus (
More info?)
In article <4e988155.0405130331.3ac38149@posting.google.com>,
bravnborg@mobilixnet.dk (arne) wrote:
> Thank you for the help
>
> i found following information on the homepage you was referring to
>
> AP9918H
> NMOS, Low gate drive
> TO-252(D-PAK)
> BVdss (breakdown voltage) 20V
> Rds_on 0.014 ohm at Vgs 4.5V
> Rds_on 0.028 ohm at Vgs 2.5V
> Id = 45A max Pd = 48W
>
> APEC isn't providing a lot of information, but i hope i can replace it
> with FDD6690A from fairchild.
>
>
>
http://www.fairchildsemi.com/ds/FD%2FFDD6690A.pdf
If you download a MOSFET for DCDC converters brochure, then
perhaps another selection would be better. The thing that worries
me about the original device, is the low gate turn on voltage.
http://www.fairchildsemi.com/collateral/dc_dcselection.pdf
If you look at some of the other devices, they have entries for
Vgs = 2.5V and Vgs = 4.5V, but none for Vgs = 10V. Like the
original device, they are designed to turn on with what are termed
"logic voltages". Let us assume for the moment that this is
a series pass linear regulator, and the op amp controlling the
MOSFET runs from +5V. Then the op-amp cannot produce a 10 volt
Vgs to fully turn on an ordinary MOSFET. It is possible Asus is
using the characteristic low turn on voltage, to keep the
op-amp based series regulator, in proper feedback.
The problem with substituting something in an already working design
is, you get no chances for mistakes. In a real design environment,
you would be prototyping the power section of the motherboard
as a separate circuit on a test board. You could then insert the
substitute device and run an analysis on the results for stability
and correct operation. You really don't want any expensive devices
connected to the power circuit while you are testing it.
This is basically the same problem as Volt Modding. A Volt Modder
thinks that they have the correct component value for a mod, but
all it takes is one slip up, and the devices connected to the voltage
regulator can be damaged.
If I was attempting this job (and I did select replacement
MOSFETS for my P2B-S, but never used them), I would want a
data sheet for the A-Power device, so I could compare more of
the aspects of the device. Also, I would want to examine the
circuitry driving the MOSFETS. I think there is an 8 pin
device, which has just enough pins to control a PWM regulator
circuit. That uses two of the MOSFETS. The third MOSFET
could be control by a quad op amp. Tracing down the power
source for the circuit controlling the linear regulator,
will tell you what output voltage range it can produce.
Based on how close to the rails the op-amp can swing, and
the voltage the op-amp has to work with, helps determine
whether a high or a low gate voltage MOSFET will work.
For the PWM regulator, getting the datasheet and examining
the reference application circuit, will help determine
what aspects of the 9918 were used for that application.
That is only important if one of those MOSFETs is damaged.
With some of these foreign companies, you might have to
write an email to the company, to get them to send you
a datasheet. The A-Power web site is such a mess, I cannot
imagine them being able to operate an FTP site. It seems
no matter where in the world you go, there are electronics
manufacturers crying for good web site design.
Also notice that, on the Fairchild web site, A-Power
isn't even acknowledged as a competitor.
HTH,
Paul
Facebook
Twitter
Instagram