Sign in with
Sign up | Sign in
Your question

MY Celeron @1600MHz!

Last response: in Overclocking
Share
a b à CPUs
a b K Overclocking
February 9, 2002 12:52:41 AM

OK, First things First, the reason my Tually wouldn't overclock is because there's something wrong with my IP3/T adapter, whereas NONE of the VID pins were connected to VSS (maybe a bad VSS connection on the PCB?), leaving it a 1.30v! So I simply wired in Vid3 to make it 1.70v. Problem is is overheats too quickly now! Tried 1.80v by hooking in Vid1 also, but that only resulted in it overheating even faster! So anyway, I have an ND-8 on it and it's still not cool enough, I think it's the junction between the heat spreader and the core that's not transfering heat fast enough. Although it does run for a short time at 1600MHz now, it still overheats at any overclocked speed, it just takes longer at slower speeds. Any thoughts?

What's the frequency, Kenneth?

More about : celeron 1600mhz

February 9, 2002 4:18:26 AM

Well, I think 1.3V->1.7V is probably too big a increase in one step.

Since the nominal value of the core voltage is 1.475V for the Tuallys, 1.3V is already low for it. Try setting it 1.5V by wiring VID3:0 as 1011

Too bad any of the popular HSFs wont fit on the IP3/T. Here is <A HREF="http://www.roberthancock.com/dell/plip3t/" target="_new">something</A> you might find a solution from. It says that the IP3/T adapter has not too good layer of thermal paste that needs to be set right. I suggest you wipe it off completely and put a better on all over the IHS. Overclocking the fan a bit also shouldnt be a problem else just replace the fan (I think it can be easily replaced) with a faster and less noisy one.

However, its pretty interesting Powerleap left these pins open, I thought they had a onboard jumper block for core voltage settings!

Keep posted, too bad I dont have any access to this Powerleap stuff here in India. Maybe I will ask a friend coming back home to get one with him!

girish

<font color=red>Nothing is fool-proof. Fools are Ingenious!</font color=red>
a b à CPUs
a b K Overclocking
February 9, 2002 4:48:08 AM

I got mine with Artic Silver III on it. The power problem could be on the CPU itself, I'll pull it to test the pin connections. Yes, I did get it to run at 1.50v, 110MHz FSB, and it did not overheat.

What's the frequency, Kenneth?
Related resources
a b à CPUs
a b K Overclocking
February 9, 2002 8:12:03 AM

OK, now I'm getting PISSED. I got the thing to run stable for 2 minutes at 1600MHz. So far here's what I've done:
Swapped out the Epox EP-BX3 for a Soyo SY-6BA+III, my benchmarks improved with that board, but it DID NOT like my 2 256MB Crucial PC133 DIMMS. In fact, sometimes it would read them as 384MB total. Using only one DIMM improved it, but to nought, the Epox board is more stable. Which sucks since the Soyo performs better, has a nicer clock generator, and more settings.
So I swapped back in the Epox board.
Now when I get it up to 124MHz FSB it starts to overload the 5v power connectors. If I unhook the other plug from its device I can end that problem.
At 1.70v everything works fine for a while. At 1.80v it overheats almost instantly. At 1.75v it overheats somewhat slower.
The voltage regulator on the iP3/T also overheats. In fact when I went so far as to put the plastic casing back on it, I could no longer overclock.
My power supply is also at the edge. To get more 5v power I started unhooking drives.
So what I have is a Tualatin that's perfectly capable of hitting 1600MHz IF I could keep it cool, a slotket that also overheats, and a motherboard that only supports 110, 124, and 133MHz FSB. Unless I want to find memory that's more stable in the Soyo board.
Hmmm, how much would a big Peltier and giant power supply run me?

What's the frequency, Kenneth?
February 9, 2002 2:25:54 PM

Well, it <i>does</i> seem to be he heat problem. AMD had its share and now intel is having theirs!

And its even more interesting when it is occuring to a smaller cooler 0.13u processor in its lower end speeds! Indications of Tualatin limited overclockability? Well, 1200->1600 is a pretty big jump, but then it just shows that <i>any</i> Tualatin core processor will not go much beyond 1.6 GHz, say 1.8 GHz would be the limit on it I guess.

And you do have a good choice for the boards, did you try it on a Asus P3B-F? I have found this board very good for memory combinations. I guess you could try using just the single DIMM and a better power supply.

Do you have a manual of this iP3/T adapter, in pdf or something? I guess I can make some tweaks with the onboard power regulator? I found a distributor selling Powerleap products in India, but they are so lazy that dont have any IP3/T on their site, I have mailed them for details today morning (IST) maybe I will have my iP3/T in a week or two!

I thought the iP3/T would be a inexpensive upgrade for older boards especially the BX ones, but as it turns out, $169 is already pretty high (almost the cost of a Celeron 1.2G + HSF + decent i815E-B board), and along with the Peltier and big PS it would be a overclockers upgrade, will do it just for the thrill/fun/heck of it!

girish

<font color=red>Nothing is fool-proof. Fools are Ingenious!</font color=red>
a b à CPUs
a b K Overclocking
February 9, 2002 4:36:55 PM

For the typical user the iP3/T combo is a good way to upgrade an existing system without doing a lot of work. Other people seem to be having better luck that me overclocking them. I've been using the Epox board without memory problems, but I really like the features of the Soyo board, which was also a better performer. I think the combination might work better if I cooled the logic chip on the adapter.

What's the frequency, Kenneth?
February 9, 2002 10:27:05 PM

Maybe you need to stick it in the oven.

Remember that one?

:cool: <b><font color=blue>The Cisco Kid</font color=blue></b> :cool:
a b à CPUs
a b K Overclocking
February 10, 2002 12:22:02 AM

I'm burning it in right now at 1488. It runs under full load for about 10 minutes at 1.5v before it overheats, unsing an Evercool ND-8 that's barely warm when overheating occurs. But the converter also gets very warm under full load, which could be the root of the problem! I hate the idea of making up little heatsinks for the subcomponents of the adapter.

What's the frequency, Kenneth?
February 10, 2002 12:39:26 AM

Maybe you can rig some sort of heatsink around the converter. A water pipe or a few baffles.

:cool: <b><font color=blue>The Cisco Kid</font color=blue></b> :cool:
a b à CPUs
a b K Overclocking
February 10, 2002 2:28:56 AM

Yes, I could do that. Using JB Weld to affix it. I could cure it in the OVEN! But no, I don't think it's worth my effort. I'm just trying to put off getting an AMD system until the Thoroughbred arrives! Now that there is a platform for AMD that I find acceptable-the XP333.

What's the frequency, Kenneth?
a b à CPUs
a b K Overclocking
February 10, 2002 7:14:28 AM

UPDATE! I removed the IHS from my CPU, and my iP3/T died! No, I don't think the two are related, as I was having voltage problems before. I think I just got a bad iP3/T, tested it with another CPU and it still doesn't work!

What's the frequency, Kenneth?
February 11, 2002 6:03:39 AM

so you removed the IHS (so the CPU is out of warranty) and now the iP3/T died! do you think they will get you a relpacement?

I havent had any experience with this adapter before, but from the posts and your description I think I will have to have it myself to get some experience and possibly invent some mods. Theres still no reply from the lone Powerleap Authorised Distributor in India!

Anyway, the product <i>is</i> defective if it isnt supplying proper core voltage in the first place. if the VID pins arent connected anywhere, its a bad design since there must be some means to set the core voltage right although all current Tualatin variants take 1.475V, specs may change in future. A jumper block with all the processor specified settings should be fine.

The cooling is bad, almost all the components seem to be overheating, including the power regulator and the PCB itself. I guess you should manually replace these parts with better models or provide custom heatsinks for them.

The adapter may be good for normal use, and not too good for overclocking. It was probabely designed that way, looking at its behaviour in OCed state.

I saw some detailed photos of the adapter <A HREF="http://www.oc.com.tw/article/0112/readgoodarticle.asp?i..." target="_new">here</A>, and I am interested about some things...
1. What are the two jumper sets meant for? One I guess is for selecting teh FSB (66/100/133/Auto) and whats the other on the lower right just behing the inductor an dthe two capacitors for?
2. Can anybody tell me the chip numbers on the back of the card and the one on the right of the socket (that is readable, 19C7YKK/IYC16222A, will find what out this chip does.
3. The PGA370 pin socket looks normal, why do they say normal HSFs dont fit in them? Or I am having traces of the Neo-370 in my memory?
4. That plastic jacket covers all of the CPU card even on the back isnt it? Crash even removed that, so does that void the warranty? I think removing it will allow better cooling of the back of the card.
5. The tall capacitor in all the four corners of the socket tend to impede the air-flow and not allow better bigger heatsinks to be put on the CPU. But tese capacitors do need to be physically closer to the processor for better stability, if their rating is known, better and smaller capacitors could be found, which could be mounted on the back of the card (not a good thing to do considering most boards and cases wont allow the processor cartridge growing laterally!)

girish

<font color=red>Nothing is fool-proof. Fools are Ingenious!</font color=red>
a b à CPUs
a b K Overclocking
February 11, 2002 6:19:00 AM

1.) The other jumper set is for PIII/Celeron SMP modes. That should mean the original Celeron, since newer ones are not SMP capable.
2.) I already repacked the thing for return.
3.) Normall CPU coolers do indeed fit, I installed an Evercool ND-8 without problems.
4.) Removing the plastic did aid in stability, and overheating of that covered chip is what lead to failure. This is obviously a design problem which they should have addressed.
5.) Most boards do have room behind the adapter, a shorter capacitor would fit on the back quite easily without the plastic. The plastic is cliped on, not press fit like Intel's.

What's the frequency, Kenneth?
!