Page 1 of 6

SPZ - The Last Dog

Posted: Wed Nov 29, 2006 10:32 pm
by Jim
Have not started work on this one yet; but the parts are accumulating. The Mobo is an FIC KA-6100 Slot 1 AT board. It has FSB settings for 100MHz, 103MHz, 112MHz, and 133MHz. (Not counting settings below 100MHz.) It also has 2 other undocumented settings. The Chipset is VIA w/ a VT82C691 Northbridge and a VT82C596 Southbridge. People more knowledgeable than me will know what model that is, but I think it's an MVP3 Apollo Pro. This board also has 3PCI Slots and an AGP Slot . One rather unusual feature is instead of the usual one twin onboard USB port, there are two of them.

Image

The caps on this board are in "Excellent" condition.

Image

Image

The board came w/ a P3 550 w/ 512k on die cache; and I suspect that "as was" it would have given a K6-3+ on a Super 7 a pretty good run for its money.

Image

But this board may have the potential to be my "Ultimate AT", so I decided to upgrade. The "Slot T" slocket, finally arrived today. The "Slot T" has jumpers which allow you to set the processor voltage at : 1.30v, to 2.05v in .05v increments. It also has jumpers which allow you to set the FSB at 66MHz, 100MHz, or 133MHz.

Image

The processor that I want to use arrived some time ago. Here we see it mounted in the "Slot T".

Image

Here are a couple of closeups that show what kind of processor it is. (a Tualatin 1.4 GHZ P3 S - SL6BY).

Image

Image

And I found a local store that has a stock of Thermaltake "Slim Volcano 10 +"s, so I picked up one of them too.

Image

The SDRAM is 3x 256Meg sticks of Hynix PC133 CL2 that I got when Sylvan 2626 and I got together to buy some.

Image

I also picked up another Vantec 7015 that will get the chop to make the chip coolers for the board. (Still have another one on the way for the SP2 board.)

Image

Still on the way is an IBM 300w AT power supply, the two 8500 Radeons, and the Rev 105 ASUS P5A-B.

Not sure how well this one is going to work. Never played w/ fixed multiplier processors before, not sure if the board will be stable at the 133 FSB, when running the multiplier that Tualatin is going to throw at it; and also not sure if the Tualatin will run at 100 FSB if that is what the board needs. A lot of maybes here.

Posted: Wed Nov 29, 2006 11:47 pm
by KGB
:shock: You DOG YOU!! Couldn't resist the Tualatin could you? ;)

I'm still saddened you are still using AT boards (yes, yes, I realize you have an affliction to those dinosaurs, but get with the times man :P ). I mentioned in a previous post about using a p3B-f. Pretty good board, overclocks my 1.2 to 1.6 very well, but can't get it stable. It might be the ram, (microns all different specs, arg), cooling, powerleap slocket, power supply. A whole host of things.

Good luck, but you could have gotten a 440BX or i820 (they're cheap), but no matter, I await the updates anxiously. I might have to get my digital camera to post my rig/mods soon too.

Posted: Thu Nov 30, 2006 7:37 am
by Jim
@ KGB : Sorry you are saddened; but it is impossible to build "the best possible ATX". Before you had accumulated half your parts, (never mind started building the thing), it would no longer be the "best possible", or even close to it. You would have to start w/ one of those new Intel processors, which I can't afford, etc. etc.

@ Nohr : KGB is right. I wasn't thinking. Could you please move this to the "Other Systems" section?

EDIT : Thanks Nohr!

Posted: Thu Nov 30, 2006 7:56 am
by tazwegion
Hmmmm... interesting, I can't make out from the picture what make those capacitors are :(

I shall await (with baited breath) the outcome of this endeavour :D

Posted: Thu Nov 30, 2006 9:38 pm
by Jim
News : My research has given me conflicting results. A review of the FIC KA-6100 indicates that the board only supports voltages as low as 2.0v. Another source, (relating to using this type of upgrade on ASUS boards), points out that a potential problem is that the "Slot T" does not have an integral voltage regulator, and what the voltage jumpers are setting, is a signal to the onboard regulator as to what voltage should be set.

The source, (Speaking of voltage regulators used by ASUS), went on to say some regulators went down to 1.3v while others only go down to 1.8v. This same source goes on to say that if the onboard regulator is incapable of supplying the voltage specified, the voltage will default to 1.8v; but it was talking about a different type of voltage regulator used by ASUS. It also gave instructions on how to identify the votage regulator chip.

Having identified my voltage regulator chip, I then did a search for info on the part number : SC1164CS 9842 D82517. The closest match I could find was SC1164CSW which is a voltage regulator made by Semtech which is capable of providing from 2.0v to 3.5v in 100 millivolt increments; and 1.3v to 2.05v in 50 millivolt increments.

The point of all this is the same source on ASUS boards also said that 1.8v would kill a Tualatin. So the question is : Is my regulator capable of going down to 1.3v as per the specs on the SC1164CSW, or is it only capable of going down to 2.0v as per the board review. If it is the latter, then I need a Powerleap Slocket w/ built in voltage regulator. I just checked E-Bay, there is one older looking model in Australia going for about $5.00; and one new one in the U.S. going for $69.00 on a buy now due to end in 3 hours.

Help anyone? KachiWachi?

Posted: Thu Nov 30, 2006 10:24 pm
by KGB
Jim, I do believe you stumbled onto the same perdicament I had found myself back in 2002. My P3C-L provides voltages < 1.5 however, when used with a Powerleap PL-ipt v.2, the BIOS only allows me to set 2.0V (like what an old Pentium2 would use) and above. To go around this, the powerleap unit powers tjhe CPU from the regulator you mentioned in your post. Even my "newer" P3B-F has this issue, allowing only; setting the Slot 1 voltage 2V or greater.

About 2 years ago, the P3C-L stopped functioning with the Celeron 1.2 and Powerleap, and now, I am regalated to using a Slot 1 P3's from here on end. (Board caps seem okay, it just does not work any longer), and now the Celery is resurrected with the P3B-F (lucky s.o.b. ain't I)?

Be vary Jim, there are two Powerleap models: the newer SlotWonder (which uses the motherboard's own voltage regulator) and the older PL-iP3T v.2 which I use. $5 is good price, include shipping, it shouldn't set you back no more that $15 CDN.

Ohh, and I concur with your article that 1.8V does kill Tualatin's. As I type this, I am viewing my 1.3 Celery which got fried on my Soyo TISU doing 1729MHz (133 FSB)@ 1.75 V with water cooling. Funny story really, I realize the system is throttlling during CS 1.6, I quickly open the case and see the ZIF Socket melting off.. :(

The 1.3 now boots for a 20 seconds, and the award bios does 2 continuous beeps High-Low, which indicated overheating/dying cpu)

Boy this went on longer than I had hoped. Good luck with the auction.

Posted: Thu Nov 30, 2006 11:19 pm
by Jim
@ KGB : Not sure what the story is on voltage. The "Slot T" has jumpers which you can set for 1.45V which sends a signal to the onboard voltage regulator requesting 1.45v. If my onboard voltage regulator is capable of going down to 1.3v, that should be all that is required. Catch is I am not sure that it is capable of going down to 1.3v. Therein lies the rub. The part number I found the data sheet for was not an exact match. SC1164CS vs SC1164CSW . Don't want to wreck either the board or the processor. The former cause they are hard to find; and the latter cause they are expensive!

Posted: Fri Dec 01, 2006 6:38 am
by KachiWachi
I couldn't find what the "W" stands for. :(

The bigger question is whether the BIOS you have can set the CPU voltage that low...or not.

For that, you'd have to disassemble the thing and check the code.

Posted: Fri Dec 01, 2006 8:10 am
by DasMan2
KGB wrote:Jim, I do believe you stumbled onto the same perdicament I had found myself back in 2002. My P3C-L provides voltages < 1.5 however, when used with a Powerleap PL-ipt v.2, the BIOS only allows me to set 2.0V (like what an old Pentium2 would use) and above. To go around this, the powerleap unit powers tjhe CPU from the regulator you mentioned in your post. Even my "newer" P3B-F has this issue, allowing only; setting the Slot 1 voltage 2V or greater.

About 2 years ago, the P3C-L stopped functioning with the Celeron 1.2 and Powerleap, and now, I am regalated to using a Slot 1 P3's from here on end. (Board caps seem okay, it just does not work any longer), and now the Celery is resurrected with the P3B-F (lucky s.o.b. ain't I)?

Be vary Jim, there are two Powerleap models: the newer SlotWonder (which uses the motherboard's own voltage regulator) and the older PL-iP3T v.2 which I use. $5 is good price, include shipping, it shouldn't set you back no more that $15 CDN.
Jim make sure that should you do try for the PL-iP3T v.2 that you get the special power plug adaptor to provide proper power source to the PL-iP3T unit.
KGB do you use that power plug adaptor and can you post a complete picture of it ... thanks .

Posted: Sat Dec 02, 2006 12:02 am
by Jim
Further research has given me the impression that this is not the best board to base this machine on, as reports on the board are generally negative. There is however one exception which indicates that someone, (rather experienced judging by content), is or has successfully run a 1.3GHz Tualatin, (presumably Celeron), on one of these boards with a "Slot T". See :

http://www.ocforums.com/showthread.php?t=18161

Posted: Sat Dec 02, 2006 1:08 am
by Me
Jim, what do you think of this board?

http://cgi.ebay.com/GA-6VA7-SOCKET-370- ... dZViewItem

Posted: Sat Dec 02, 2006 7:13 am
by Jim
Answer : = DROOL!!! Not sure if I could run a 1.4GHz Tualatin in one of those though. Would have to research the voltage reulator, and amount of power the board can supply the CPU. You have definitely got my interest here. Thanks.

Posted: Sat Dec 02, 2006 10:30 pm
by KGB
@ Jim, the Giga-byte board seems reasonable. I'm sure you have looked to the matter yourself but, the link to the board is http://www.gigabyte.com.tw/Products/Mot ... e=GA-6VA7+. It might be worth your while to get it, although it seems to me that Tualatin chips are not supported but, can be easily used with a Powerleap socket 370 adapter -> http://www.powerleap.com/PL-NeoT.jsp. The apollo pro [A] chipset is pretty good and I think allows asynchronous memory timings, so it wouldn't hurt to overclock the bus to 105-110 if you so dare.

Also, on the website for your slot T adapter, the FIC KA-6110 is reported as working, and the model you said to have purchased KA-6100 is not listed, or mis-spelled, not included or tested yet

About the voltage question, I am not familiar with your upgradeware.com model, but most boards recognize the Slocket, but with my limited experience on this matter, it appears to give DEFAULT P2 voltage (even with jumpers on motherboard set to CERTAIN voltages. Also, most motherboard BIOS' do not know of the CPU ID string when installed with a SLot T or powerleap adapter. Both my motherboards report the Celery as a Pentium II @ 1200 8). A quick check with CPU-Z sets things right, proclaiming the cache, mmx, sse are all supported and used.

@Dasman, yes my Powerleap does require the molex adapter being used when powering the Slot 1 unit. I'll try and get a pic on this post when I have available time.

Posted: Sat Dec 02, 2006 11:42 pm
by Jim
@ KGB : The link I gave shows that a 1.3GHZ Celeron will run ok w/ a "Slot T" adapter on a KA-6100. That means that the voltage regulator on the KA-6100 can supply the correct voltage required by the 1.3 Celeron, and that the bus can supply sufficient power to operate, as the "Slot T" has no regulator, nor does it provide for powering the processor separately from the mobo bus. How well that equates to the requirements of a 1.4GHz P3S is another issue.

Regards the Gigabyte, my concern is again the voltage regulator and the amount of power the mobo bus can supply the processor. If the onboard mobo voltage regulator does not go down to 1.3V, then the powerleap 370 (Neo) adapter will not work because it does not have an integral regulator; but rather, like the "Slot T", it sends a signal to the onboard regulator telling it what voltage to set. Catch is that only works if the onboard regulator is capable of setting the voltage to the level signalled. Similarly, like the "Slot T", the powerleap 370 (Neo) adapter does not supply an independant power source for the processor; but rather relies on the mobo bus to supply the power required. Again, that only works if the mobo bus is up to supplying the amount of power required by the processor used. (1.4GHz P3S Tualatin).

Posted: Sun Dec 03, 2006 12:00 am
by KGB
@Jim
Basically what the Neo 370 adapter does is this: -->http://digilander.iol.it/grandecigno/Tua_BX_41.htm

What this person does is; modify the Tualatin pins on the cpu to work with an older non-tualatin supporting PC133 fsb supporting motherboard. (Fooling both memory subsystem and cpu's own id string) With no ill affects. After modding the pins, drop the cpu in a REGULAR slocket (Slot 1) or a regular FCPGA socket (BX, 820, 810e,815e, Apollo Pro/A) based motherboard. For slot 1 its MORE risky, because board manufacturer's knew there were P2 and p3 Katmai's around and needed 2V support, the socket 370 boards had celery and CuMine P3's in mind when designing < 1.5V sockets. Granted Tualatin appeared later, a work around was found. Hence, a voltage regulator is NECESSARY for OLDER SLot 1 motherboards, while it is NOT for newer socket 370 boards. (I hope it clears up a few things) You can google for "Tualatin Wire Mod" or derivatives of the term for other testimonials.