VOGONS


First post, by feipoa

User metadata
Rank l33t++
Rank
l33t++
stuvize wrote:

Thank you for asking this question I would like to start a new thread on how to mod this cpu but since im a newb I don't think I have permission. Basically remove voltage regulator solder where the center leg went to the innermost leg then it can be powered by the mobo regulators set them for 4 V. The 5 V+ fan leg and 5 V- leg must be removed or insulated from contact other leg must be left, then run 5 V from psu or fan plug on mobo to the fan, if you do not mod the fan it will run too slow at 4 V and cause it to run multiplier of 1. I did this over a year ago on a POD that had almost every pin on it bent so it was a good victim, after I straitened all the pins of course and made sure it worked haven't had any problems running at 100mhz even with all the latencies set at the lowest and it is crazy fast in the m919 at 40/100mhz. I am sorry this is a little off topic as it gets into modding

Could you please provide some photos or a diagram sketch. I tried to follow your sentance structure, but I became confussed, particularly "basically remove voltage regulator solder where the center leg went to the innermost leg". How does the POD's voltage regulator's centre pin go to the innermost leg"? I do not see these two pins soldered together.

Also, "The 5 V+ fan leg and 5 V- leg must be removed or insulated from contact other leg must be left". I do not follow this sentance at all.

EDIT:
Could you confirm that your POD is working with L1 write-back enabled and not write-through enabled? I found that my m919 version 3.4 boards were actually running with L1 in write-through mode. Try running chkcpu.exe and running the Quake_SW version 1.06 timedemo.

Plan your life wisely, you'll be dead before you know it.

Reply 1 of 89, by Artex

User metadata
Rank l33t
Rank
l33t

This would be great if we can get these POD 83s working at 100Mhz consistently with WB enabled. I've tried at least 5 of these so far, and I cannot get any working reliably at 100Mhz.

My Retro B:\ytes YouTube Channel & Retro Collection
LihnlZ.jpg

Reply 2 of 89, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Attached is an image of how the POD's VRM is connected. Perhaps you could explain what you did in terms of this diagram?

POD83.jpg
Filename
POD83.jpg
File size
45.74 KiB
Views
4348 views
File license
Fair use/fair dealing exception

Plan your life wisely, you'll be dead before you know it.

Reply 3 of 89, by stuvize

User metadata
Rank Member
Rank
Member

Sorry for the poor description I was trying to condense the subject, as for the voltage regulator the pin closest to the edge of the cpu is the ground the center is the 3.3 V and the next is the 5 V in. I found this info from someone else who found it buried on Nikos website, the 5 V in must be soldered to the 3.3 V pin or at least were it was I removed the voltage regulator from the cpu and bridged these two points with solder. I am not a expert on circuitry so some of this is over my head. What I meant by the fan was that it must be powered externally and the ground and 5 V pin that powered it from the cpu must be removed or insulated otherwise when the fan is powered externally the 5 V will backfeed into the 3.3 V circuit, the signal pin from the fan to the cpu must be left. I ran cpu check sadly it is running in write through mode but it still runs circles around the 100mhz ibm 586 I had even when it was running 40/120mhz and features turned on and it would crash eventually at 40mhz bus from games like quake or doom, I have never had the POD crash or freeze at 40/100mhz. I could not get 50mhz bus out of the POD even at 5 V with multiplier of 1 (fan removed) but I did not know that it was running in write through as it was so much faster than AM or CX 586 this may be linked to m919 issue with POD as I have read other posts with people running POD on via chipsets at 60/60mhz or even 66/66mhz. I hope my crude drawing helps clear things up too. what commands are used to get quake show fps in console?

Attachments

  • 100_1392.JPG
    Filename
    100_1392.JPG
    File size
    584.32 KiB
    Views
    4313 views
    File license
    Fair use/fair dealing exception

Reply 4 of 89, by Anonymous Coward

User metadata
Rank l33t
Rank
l33t

To me it really makes a lot of sense to remove the VRM, because really...who is going to run a POD83 on a 5V system anyway?

Now what I'd like to know if there are any other ways to change the multiplier on these things...1.5X, 2.0X and 3.0X settings would be quite useful.

Last edited by Anonymous Coward on 2015-10-24, 07:40. Edited 1 time in total.

"Will the highways on the internets become more few?" -Gee Dubya
V'Ger XT|Upgraded AT|Ultimate 386|Super VL/EISA 486|SMP VL/EISA Pentium

Reply 5 of 89, by feipoa

User metadata
Rank l33t++
Rank
l33t++

OK. This is very simple.

1) short the 5V and 3.3V pins on the POD's voltage regulator
2) insulate the fan's 5V pin from making contact with the gold pad (heatsink side), for example, with a piece of electrical tape.
3) run a 5V wire from the PSU to the fan's 5V connector, perhaps by soldering it on

Any reason you removed the voltage regulator IC? Will it work with the VRM in place? Does leaving the VRM in place, with input/output shorted cause much voltage drop? Perhaps it is sufficient to snip these 2 leads.

You will get an extra 1-2 fps from DOS Quake if you run the POD in WB-mode. If you want to view the numeric benchmark result in Quake: Load Quake; select single player mode; once in the game, press this key: ` to bring down the command console; in the command console type: timedemo demo1; press enter; quickly press the ` key again to hide the command console; the benchmark should start running; wait for it to complete to view the fps result.

Plan your life wisely, you'll be dead before you know it.

Reply 6 of 89, by stuvize

User metadata
Rank Member
Rank
Member

I removed the VRM because it shattered cutting the leads and they had to be removed so my soldering iron could get in to bridge the 3.3 V and 5 V, the VRM probably could be left with the leads cut. Can POD L1 cache run in WB-mode on M919 or are you referring to other mobos with the extra 1-2 fps in dos quake? these are my results from quake VER. 1.06 POD at 83mhz 969 frames 52.2 seconds 18.6 fps POD at 100mhz 969 frames 43.3 seconds 22.4 fps. Changing the multiplier on this cpu may be possible as it shares the same core as the socket 4 pentiums and I remember reading a post about a group that modded the multiplier in some socket 4 pentiums back in the 90s.

Reply 7 of 89, by feipoa

User metadata
Rank l33t++
Rank
l33t++

stuvize, I was unable to get the m919 to function with L1 in write-back mode. I am refering to using a different motherboard for L1-WB when using the POD. These are my Quake 1.06 results,

POD83-WT = 19.7 fps
POD83-WB = 20.8 fps
POD100-WT = 23.6 fps
POD100-WB = 24.4 fps

Plan your life wisely, you'll be dead before you know it.

Reply 8 of 89, by HighTreason

User metadata
Rank Oldbie
Rank
Oldbie
Anonymous Coward wrote:

To me it really makes a lot of sense to remove the VRM anyway, because really...who is going to run a POD83 on a 5V system anyway?

* Clears throat *

Actuallym, i don't anymore as I have moved it to a PCI board, but I still have to run it at 5V as it won't boot with a lower voltage selected. The regulator is still installed and the CPU is not modified in any way.

Feipoa, the results posted are pretty much consistent with my own... I run yet another UM8886 motherbaord - the MB-4DUPM/E.

Never really realized until recently how lucky I was with my OverDrive, a large percentage apparently flat refuse to run at 100MHz under any circumstance, or so I heard. This is the only one I have ever owned so I can't say much for the majority of them. Certainly a nice CPU though when used the right way.

My Youtube - My Let's Plays - SoundCloud - My FTP (Drivers and more)

Reply 9 of 89, by feipoa

User metadata
Rank l33t++
Rank
l33t++

The benefit of leaving the VRM in place would be to put on a resistive trimmer on the POD's voltage set resistor, that way you can adjust the POD's voltage to whatever you want, without having to rely on the motherboard's VRM.

Plan your life wisely, you'll be dead before you know it.

Reply 10 of 89, by stuvize

User metadata
Rank Member
Rank
Member

So did you bridge the VRM pins to run off mobo VRM or is your on chip VRM modified? I noticed that the cpu runs much cooler with the on chip VRM is disabled even if you give it 5 V, in theory 5 V should be no problem for the core as it the same as socket 4 so there may be more OC potential for the POD but I have not run for extended periods at 5 V.

Reply 11 of 89, by feipoa

User metadata
Rank l33t++
Rank
l33t++

I haven't modified my POD yet. I am working with 386 hardware at present. The onboard VRM is probably the linear type, which is heating up the POD.

Plan your life wisely, you'll be dead before you know it.

Reply 12 of 89, by stuvize

User metadata
Rank Member
Rank
Member

Modified another POD did this one without removing the VRM, got the solder hot and bent the VRM pins up and bridged the circuit. I uploaded a pic of cpu and also a pic showing how the fan is externally powered. Oddly this cpu is a little faster cpu when overclocked cpu check says it runs at 101.5mhz and the other ran at 100.1mhz Quake score was also a little better 23.1fps vs. 22.4fps. This CPU also doesn't immediately freeze at 50mhz bus. feipoa what mobo are you testing your POD in? I wonder what the best mobo for the POD is?

Attachments

  • 100_1406 (Large).JPG
    Filename
    100_1406 (Large).JPG
    File size
    152.31 KiB
    Views
    4149 views
    File comment
    Fan powered by mobo fan header
    File license
    Fair use/fair dealing exception
  • 100_1396 (Large).JPG
    Filename
    100_1396 (Large).JPG
    File size
    92.74 KiB
    Views
    4149 views
    File comment
    Bridged CPU
    File license
    Fair use/fair dealing exception

Reply 13 of 89, by feipoa

User metadata
Rank l33t++
Rank
l33t++

What is that aluminum coloured heatsink on the POD for?

In my opinion, the Biostar MB-8433UUD is the best motherboard for the POD. It works with L1 write-back cache and is very stable. I've tested several different PCI-based socket 3 motherboards and most were pseudo stable, at best, with the POD.

Plan your life wisely, you'll be dead before you know it.

Reply 14 of 89, by stuvize

User metadata
Rank Member
Rank
Member

That's a Zalman vga ram heatsink I put on for a little extra cooling it barley fits but does help. The Biostar MB8433UUD is the same mobo you used in fastest 486 thread I have seen it around but always at a ridiculous price. I think I will go to amibay and post would gladly trade for my socket 4 deskpro xe560

Reply 15 of 89, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Luckily the MB-8433 is a rather common motherboard. I bought a stash of them before making public the qualities of this board.

Plan your life wisely, you'll be dead before you know it.

Reply 16 of 89, by feipoa

User metadata
Rank l33t++
Rank
l33t++

I went ahead and modified one of my POD83 chips which wasn't able to run reliably at 100 MHz. I cut the POD's VRM input lead, bridged the Vin-to-Vout, and ran a 5V line to the POD's fan. I then set my motherboard's voltage to 4 V and powered up. The system can now run well at 100 MHz, however when I measure the voltage going to the CPU it is 4.90 V. Why?

I beleive the problem comes from the maximum output current of my motherboard's voltage regulator module, which is 2 A. According to the Intel spec sheet, the POD at 83 MHz requires 2.8 A. Obviously more current would be required at 100 MHz. My motherboard's VRM has an on/off freature, as well as an overheat protection circuit. I suspect something is getting tripped so that Vin and Vout are getting shorted, leading to 4.9 V going to the POD.

How do we solve this problem?
1) I was thinking I could figure out which is the voltage set resistor on the POD's VRM and adjust it accordingly to get 3.6 - 4.0 V to the POD.
2) I was also thinking to simply use 2 standard diodes connected in series to create a 1.34 V drop and connect them from the input to the output on the POD itself, near the POD's VRM. However, I figure I'd want some 5 A diodes. Most of what I have is 15 mA to 1 A for diodes.
3) Replace the VRM on my motherboard.

Plan your life wisely, you'll be dead before you know it.

Reply 17 of 89, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Stuvize - would you be willing to measure the voltage going to your POD when you set your motherboard's VRM to 4 V?

Plan your life wisely, you'll be dead before you know it.

Reply 18 of 89, by stuvize

User metadata
Rank Member
Rank
Member

This is interesting I recently acquired a V3.4 M919 and it was fussy with the POD only one would always boot at 4 V and neither would boot with 5 V or FPM memory the V3.3 did not have these issues although it has a tiny resistor on the reverse side of the board bridging two pins on the CPU socket thought I had a picture I'll have to take the board out of the case and take a picture. Also got inspired and removed the heat sink from one of the PODs and the rest of the on chip circuit so I could put a proper cooler on it went to a bunch of trouble grinding the corner on the het sink so I could solder a lead to POD fan signal found out it was for nothing tried a 50/50mhz boot like I had done before by pulling the fan but the fan signal seems to be null with all the on chip VRM circuitry gone it will boot at 2.5x multiplier with or without the fan. The better cooler did make it semi stable at 50mhz bus but not enough to run a successful Quake demo. I need to purchase a multimeter any particular brand you recommend had a cheap one before it didn't last long.

Attachments

  • 100_1441 (Large).JPG
    Filename
    100_1441 (Large).JPG
    File size
    86.94 KiB
    Views
    3946 views
    File comment
    Wonder what those 3 tiny round gold contacts are for?
    File license
    Fair use/fair dealing exception