VOGONS

Search

Common searches

VRM very low voltage with MMX CPU on i430VX board

Topic actions

First post, by RetroSpector78

Posted on 
User metadata
Rank Member
Rank
Member

So I have this PT-2006 i430VX motherboard that came with a WinChip C6 200.

It appears that the jumpers on the board were configured for an Intel Pentium 200 MMX (see picture), using 2.8v core and 3.3v io.
However, the system did start with the WinChip but it was very unstable and even after setting the jumpers to 3.5v it either hanged or blue-screened on me.

So I decided to switch to an Intel Pentium 200 MMX, and set the jumpers to 2.8v core and 3.3v io (see picture)

It managed to boot once, but then it hanged in the windows98 desktop. Since then I haven't been able to boot it with the intel mmx. (no post, nothing on screen, no beeps).
The WinChip is still booting with this motherboard, but has serious stability issues.

Now on the board I saw 2 voltage regulators. I am assuming one for the IO and one or the cpu core.

I got the following measurements without a CPU :

  • The EZ1084AC always outputs 3.3v (I guess this is the IO voltage).
  • The EZ1083CT is setup as an adjustable voltage regulator, and depending on VR1 jumper outputs the correct cpu core voltage (2.5v, 2.8v, 2.9v, 3.3v, 3.5v)
  • Both are getting a clean 5V signal as input.

So these readings seem to be correct, albeit without load, so I decided to do the measurements again with the CPU:

With the intel MMX CPU installed :

  • The EZ1084AC always outputs 3.3v
  • The EZ1083CT outputs 1.4v
  • Both are getting a clean 5V signal as input.

1.4v is defintely too low, and would explain why the system does not start and the CPU doesn't "warm" up. However, why would it output only 1.4v

With the WinChip CPU installed :

  • The EZ1084AC always outputs 3.3v (I guess this is the IO voltage).
  • The EZ1083CT outputs 3.3v (regardless of the jumper settings)
  • Both are getting a clean 5V signal as input.

Strange that it always outputs 3.3v and not taking into account the VR1 jumper settings (2.5v, 2.8v, 2.9v, 3.3v, 3.5v)

2020-11-21 22.13.40 copy.jpg
Filename
2020-11-21 22.13.40 copy.jpg
File size
885.08 KiB
Views
487 views
File license
CC-BY-4.0

Any ideas ?

Reply 2 of 9, by evasive

Posted on 
User metadata
Rank Oldbie
Rank
Oldbie

Found datasheets for both of them.

Both have no voltage rating so are adjustable voltage.

Check the datasheets for the resistor values. You say the output drops to 1.4V, what happens on the input at that time? Is that still 5V? Oh and how clean is clean, did you check with an oscilloscope?

Attachments

Reply 3 of 9, by RetroSpector78

Posted on 
User metadata
Rank Member
Rank
Member
computerguy08 wrote on 2020-11-22, 11:05:

I would suspect maybe the EZ1083CT is bad (from a previous OC or something like that) ?

EDIT: Could also be bad capacitors on that rail.

Don't know much about the history of the system, but fact that the jumper settings on the board didn't match up with the installed CPU doesn't look great 😀

The WinChip C6 supports 1 of the 2 voltage ranges :

  • 3.52V (3.45V–3.6V)
  • 3.3V (3.135V–3.6V

Mine has 3.52 written on it so I guess it expects 3.45V - 3.6V. However with the WinChip installed its always getting 3.3v. Could that explain the instability (not enough voltage) ?

Will try to check the caps and the regulator. It is strange however that

  • It was able only able to boot the Intel MMX once and never again after that.
  • The output voltage on the winchip always being 3.3 while the intel it dropping to 1.4

Reply 4 of 9, by RetroSpector78

Posted on 
User metadata
Rank Member
Rank
Member
evasive wrote on 2020-11-22, 11:11:

Found datasheets for both of them.

Both have no voltage rating so are adjustable voltage.

Check the datasheets for the resistor values. You say the output drops to 1.4V, what happens on the input at that time? Is that still 5V? Oh and how clean is clean, did you check with an oscilloscope?

Haven't checked it with an oscilloscope yet, but on multimeter it had a stable 5.1v. (even when the output was only 1.4v)

Reply 5 of 9, by computerguy08

Posted on 
User metadata
Rank Member
Rank
Member
RetroSpector78 wrote on 2020-11-22, 11:47:
Don't know much about the history of the system, but fact that the jumper settings on the board didn't match up with the install […]
Show full quote

Don't know much about the history of the system, but fact that the jumper settings on the board didn't match up with the installed CPU doesn't look great 😀

The WinChip C6 supports 1 of the 2 voltage ranges :

  • 3.52V (3.45V–3.6V)
  • 3.3V (3.135V–3.6V

Mine has 3.52 written on it so I guess it expects 3.45V - 3.6V. However with the WinChip installed its always getting 3.3v. Could that explain the instability (not enough voltage) ?

Will try to check the caps and the regulator. It is strange however that

  • It was able only able to boot the Intel MMX once and never again after that.
  • The output voltage on the winchip always being 3.3 while the intel it dropping to 1.4

We have this board in uh19 recorded as Pentium [3.3V]: http://www.win3x.org/uh19/motherboard/show/2442 (you can find the original manual here btw)
As a recap, uh19 is a project I'm working on, intended to be a replacement for th99.

Did a bit of search, it's the same story on a th99 clone: https://arvutimuuseum.ee/th99/m/E-H/34091.htm
The CPU support list is made of 3.3V CPUs mostly. Strange, because your PCB clearly has MMX voltages written on the silkscreen.

P.S: If you can take a photo of your board so we can document it in uh19, that would be nice 😁

Reply 6 of 9, by RetroSpector78

Posted on 
User metadata
Rank Member
Rank
Member
computerguy08 wrote on 2020-11-22, 12:00:
We have this board in uh19 recorded as Pentium [3.3V]: http://www.win3x.org/uh19/motherboard/show/2442 (you can find the origina […]
Show full quote

We have this board in uh19 recorded as Pentium [3.3V]: http://www.win3x.org/uh19/motherboard/show/2442 (you can find the original manual here btw)
As a recap, uh19 is a project I'm working on, intended to be a replacement for th99.

Did a bit of search, it's the same story on a th99 clone: https://arvutimuuseum.ee/th99/m/E-H/34091.htm
The CPU support list is made of 3.3V CPUs mostly. Strange, because your PCB clearly has MMX voltages written on the silkscreen.

P.S: If you can take a photo of your board so we can document it in uh19, that would be nice 😁

I will take some detailed pictures this afternoon. And also need to checkout the win3x uh19 project in more detail ... Deksor already mentioned it a couple of times to me .... Will see if I can make a contribution ... Planning on doing a video of about 30 motherboards I still need to test / verify. Might as well take some pictures / notes and send them to you.

Reply 7 of 9, by Deksor

Posted on 
User metadata
Rank l33t
Rank
l33t

Sure, do it 😁

And also if you can backup the BIOSes please do 😀 (this might be too much to ask considering the quantity of boards you have to get through).

Trying to identify old hardware ? Visit The retro web - Project's thread The Retro Web project - a stason.org/TH99 alternative

Reply 8 of 9, by quicknick

Posted on 
User metadata
Rank Oldbie
Rank
Oldbie
RetroSpector78 wrote on 2020-11-22, 10:46:

Strange that it always outputs 3.3v and not taking into account the VR1 jumper settings (2.5v, 2.8v, 2.9v, 3.3v, 3.5v)

This is the easiest to explain. Speaking off the top of my head, there are around 50 power (Vcc) pins on a S7 CPU. In the case of split power planes, those 50 are (almost) equally split between Vcc2 (core) and Vcc3 (I/O). Some S7 boards don't use jumpers for choosing single or split power planes, and instead rely on the (single-plane) CPU itself to short the two rails. So with the WinChip you're seeing the I/O voltage "spilling out" to the (most certainly defective) EZ1083's output, and the CPU is only powered by the I/O VRM, which is not good.

Hope you sort this one out and looking forward to see it featured on your great channel! 😀

Go to top of page Go to top of page
Back to General Old Hardware
Close window
Close window