VOGONS


Reply 25 of 48, by rasz_pl

User metadata
Rank l33t
Rank
l33t

you could measure Vcore with oscilloscope(or even multimeter, wont show noise, but will show voltage drop) if you dont want to recap without being sure

Open Source AT&T Globalyst/NCR/FIC 486-GAC-2 proprietary Cache Module reproduction

Reply 26 of 48, by donhonk

User metadata
Rank Newbie
Rank
Newbie
rasz_pl wrote:

you could measure Vcore with oscilloscope(or even multimeter, wont show noise, but will show voltage drop) if you dont want to recap without being sure

Gotcha, Ive got a Multimeter. Would that basically involve probing caps near the cpu / atx power in while its powered on? Im new to this stuff so apologies if its a dumb question

Reply 27 of 48, by rasz_pl

User metadata
Rank l33t
Rank
l33t

it would involve soldering a little taps so you can permanently connect multimeter/scope and run computer at the same time - then you can look at the voltage while running at different frequencies / running heavy programs / starting windows etc

Open Source AT&T Globalyst/NCR/FIC 486-GAC-2 proprietary Cache Module reproduction

Reply 28 of 48, by donhonk

User metadata
Rank Newbie
Rank
Newbie
rasz_pl wrote:

it would involve soldering a little taps so you can permanently connect multimeter/scope and run computer at the same time - then you can look at the voltage while running at different frequencies / running heavy programs / starting windows etc

Gotcha, appreciate the help! I will be getting a proper desoldering gun soon, would like to revive this board if I can.

Reply 30 of 48, by donhonk

User metadata
Rank Newbie
Rank
Newbie
RoyBatty wrote:

RAM settings are finicky with the VIA chipsets, try turning off CPU-DRAM back to back transaction in the advanced chipset menu. My TUV4X will not boot with this setting enabled.

Gotcha! I can give that a go tomorrow

Reply 31 of 48, by donhonk

User metadata
Rank Newbie
Rank
Newbie
RoyBatty wrote:

RAM settings are finicky with the VIA chipsets, try turning off CPU-DRAM back to back transaction in the advanced chipset menu. My TUV4X will not boot with this setting enabled.

HT91rs7m.jpg

Hm, I still cant quite reach Windows. Anything you'd change about these settings?

Reply 32 of 48, by dottoss

User metadata
Rank Member
Rank
Member
donhonk wrote:
Rather frustratingly, this machine simply wont run stably with the processor running at 1400mhz. (The system frequency of 133) […]
Show full quote

Rather frustratingly, this machine simply wont run stably with the processor running at 1400mhz. (The system frequency of 133) It will get to bios and then the keyboard stops responding after 1 keystroke. Kinda defeats the purpose of the CPU I got.

BiUj0mGm.jpg

Ive even taken everything out just to eliminate variables. (And swapped to an NVIDIA vanta just to check) Windows 98 is installing as we speak, but any ideas?

EDIT:
nh1kF96m.jpg
Got windows installed just fine while it was running at 840mhz, totally stable. Again, kinda defeats the purpose if I cant actually run it at its intended speed?

Just popping in to say that my CUV4X behaves exactly the same, i just got my CPU from Korea too. It's fine up to about 120MHz FSB and after that it will behave like yours, unstable and keyboards stops working after one to two strokes. Bios settings has been tweaked, like the previous mention DRAM option.

Reply 33 of 48, by donhonk

User metadata
Rank Newbie
Rank
Newbie
dottoss wrote:

Just popping in to say that my CUV4X behaves exactly the same, i just got my CPU from Korea too. It's fine up to about 120MHz FSB and after that it will behave like yours, unstable and keyboards stops working after one to two strokes. Bios settings has been tweaked, like the previous mention DRAM option.

Hmm, interesting. Doesnt bode well for this combo it seems, even though the ebay page does list it as compatible. A similar Capacitor issue perhaps? (no idea) It does leave me in a weird spot with this machine. It seems I've got several options.

A. Attempt to recap the motherboard in hopes it will fix the issue. I dont think I have the technical know-how to diagnose the issue like previously mentioned. No oscilloscope, but I do have the know-how / equipment to replace the caps.

B. Buy another supposed compatible Motherboard to use with the Korea modded Tualatin.

C. Go with a non modded Tualatin in a supported motherboard.

Reply 34 of 48, by rasz_pl

User metadata
Rank l33t
Rank
l33t

have you tried bumping vcore to 1.6? its still safe and would give buffer against any vcore drop/fluctuations

Open Source AT&T Globalyst/NCR/FIC 486-GAC-2 proprietary Cache Module reproduction

Reply 35 of 48, by dottoss

User metadata
Rank Member
Rank
Member
rasz_pl wrote:

have you tried bumping vcore to 1.6? its still safe and would give buffer against any vcore drop/fluctuations

I've tested all between 1.45 to 1.70.

Also, I noticed that the OP has motherboard rev 1.03 which is the same as mine. Perhaps coincidence, perhaps not.

Reply 36 of 48, by donhonk

User metadata
Rank Newbie
Rank
Newbie
dottoss wrote:
rasz_pl wrote:

have you tried bumping vcore to 1.6? its still safe and would give buffer against any vcore drop/fluctuations

I've tested all between 1.45 to 1.70.

Also, I noticed that the OP has motherboard rev 1.03 which is the same as mine. Perhaps coincidence, perhaps not.

Yep, just tried raising vcore as well, no luck. I think I will attempt a recap, but failing that I might go for a native Tualatin build unless others have a board they really like for a modded Tualatin.

Reply 37 of 48, by rasz_pl

User metadata
Rank l33t
Rank
l33t

That would most likely exclude caps as the source of problems. Personally I wouldnt even touch VIA board with a stick, and rather take my chances overclocking 440BX.

Open Source AT&T Globalyst/NCR/FIC 486-GAC-2 proprietary Cache Module reproduction