NostalgicAslinger wrote on 2023-09-21, 17:01:

The P3B-F is produced for FSB133 marketing on the BX. So no 3,30 default VIO settings.

VIO is fed through +5V rail, that's why you can increase it. Hardware monitoring only shows PSU rails voltage and CPU voltage.

shevalier wrote on 2023-09-21, 14:47:

When there are unexplained problems with a video card, they usually start testing it. But from personal experience, it’s worth s […]
Show full quote

When there are unexplained problems with a video card, they usually start testing it.
But from personal experience, it’s worth starting with testing the RAM and CPU.
When RAM is buggy, there are all sorts of glitch. Including the video card, because the video driver is located in system memory and is executed by the CPU.
Although, may be unlucky with the motherboard too

I wish I had faulty CPU and\or RAM but that other P3B-F is dead as a doornail sadly. It has other issues that I'm not able to find with my knowledge and above all with my equiment (I'm no Adrian Black).

mrfusion92 wrote on 2023-09-22, 17:04:

NostalgicAslinger wrote on 2023-09-21, 17:01:

This means, that the bios still says ~3,44V in the Hadware Monitor without any VIO Jumper? Same as position 1-2?
3,50V is only interesting for FSB133 on the i440BX, but for FSB 66/100 is 3,30V more than enough.

Are you a wizard or something? Because that's exactly the value it says. And there is no difference between no jumper and 1-2 position.

I use the same mainboard, a P3B-F Rev. 1.04 with a real Slot 1 SL4KL PIII 1GHz FSB100, and want to go down to 3,3V VIO, because this is more than enough, if you do not drive FSB133 on the i440BX and should give a longer life to the components.

I have build this system in 2016 and it runs fine without any problems with a GeForce4 Ti 4200 AGP and 2x Voodoo2 12MB SLI since this time. Only one 256MB SDRAM Memory from Samsung has said goodbye during this time, all other components are still the same since 2016.

So there is no chance to go down to 3,30V VIO with a P3B-F without any hardware mods.

mrfusion92 wrote on 2023-09-21, 06:20:

I have also checked the AGP voltage, I don't have the oscilloscope but apparently, at least in my motherboard, there is no difference between VIO jumper in 1-2 position or no jumper at all. The reading was always ~3.5V. Instead in 2-3 position was >3.6V as the manual reports.

Thanks for the information. I will then continue to leave in jumper position 1-2. I have never dared without jumper, because does not want to scrap my great working P3B-F board.

mrfusion92 wrote on 2023-09-22, 17:04:

Because that's exactly the value it says. And there is no difference between no jumper and 1-2 position.

It was in P3B-F that Asus decided not to solder the second resistor, removing the ability to set exactly 3.3V.
It is on this board that position 1-2 is the position of “storing the jumper from loss”
No R1411, no additional adjustment step.
There is an P3BF circuit online, but it is of such quality that it would be better not to have it.
https://vinafix.com/threads/shematic-boardvie … -desktop.19882/

I noticed only now that the Matrox G400 works but it is at 1x while both card and motherboard should be agp 2x capable. In the bios there isn't any setting regarding AGP rate.

Is there a way to understand why this happens?

yes, by disabling onboard VIO generating circuit by lifting one leg of the coil and wiring 3.3V straight from ATX connector

Just an update - there's a way to replicate this issue on Asus boards. I've tested a P2B 1.04, P2B-S, and a P3B-F. All exhibit the same issue. On the P2B 1.04 (and the P3B-F, but I didn't test it), I tried shifting the VIO jumper to all three positions (1-2, 2-3, OFF), all with the same result... If you update Quake 3 to 1.32C (that's the 1.32 point release and then the 1.23c update from 2006 where you just overrwite quake3.exe from the archive), then set the resolution in the game to 1024x768, set both the texture and color depth to 32-bit, set to trilinear and max the sliders, then run timedemo 1 from the console and play demo001 (not four, but demo001). This is an excellent test that will crash your card if the AGP is out of spec. Your system can be otherwise stable, but in this it will crash.

So far an FX5600 does pass this test, but I think it's due to its lower power draw.

rasz_pl wrote on 2023-10-04, 16:54:

yes, by disabling onboard VIO generating circuit by lifting one leg of the coil and wiring 3.3V straight from ATX connector

Can you show an example of this as well as the diode D33 on the P2B? I'd like to give this a shot on both boards. Thanks

mockingbird wrote on 2024-02-12, 03:34:

rasz_pl wrote on 2023-10-04, 16:54:

yes, by disabling onboard VIO generating circuit by lifting one leg of the coil and wiring 3.3V straight from ATX connector

Can you show an example of this as well as the diode D33 on the P2B? I'd like to give this a shot on both boards. Thanks

L16 is the source of 3V on Asus P2B. You can either lift leg closer to SLOT1 up or set VIO to 3.3V and let it be, then fit thick wire jumper in position of D33 just above L16 - this will link ATX 3V to motherboard 3V and provide all the 3V power any card would ever want.
Asus P2B boardview Re: Early revision ASUS P2B and Coppermine Pentium 3 its like a diagram, but for pcb inspection. Sadly I dont have P3B boardview/schematic. There is a P3BF.pdf containing a collection of pages from various Pentium2 board schematics, some BX98 P2B98 P2B98-LS with one sheet from P3BF https://vinafix.com/threads/shematic-boardvie … -desktop.19882/ not sure if its a hodge podge or if Asus reused diagrams just changing sections betweeen board models. Sheet 15 shows 3V circuit with U10 generating 3V_dual (upper right corner), but also JP11 and JP12 (just under L14) as a provision of linking ATX 3V with 3V_dual. Looks like in case of P3BF you can solder those two jumpers and thats it.
There might be potential problem if board maintains generating 3V_dual in standby mode? Maybe thats why there is provision for D34?

rasz_pl wrote on 2024-02-12, 08:57:

L16 is the source of 3V on Asus P2B. You can either lift leg closer to SLOT1 up or set VIO to 3.3V and let it be, then fit thick wire jumper in position of D33 just above L16 - this will link ATX 3V to motherboard 3V and provide all the 3V power any card would ever want.
Asus P2B boardview Re: Early revision ASUS P2B and Coppermine Pentium 3 its like a diagram, but for pcb inspection. Sadly I dont have P3B boardview/schematic. There is a P3BF.pdf containing a collection of pages from various Pentium2 board schematics, some BX98 P2B98 P2B98-LS with one sheet from P3BF https://vinafix.com/threads/shematic-boardvie … -desktop.19882/ not sure if its a hodge podge or if Asus reused diagrams just changing sections betweeen board models. Sheet 15 shows 3V circuit with U10 generating 3V_dual (upper right corner), but also JP11 and JP12 (just under L14) as a provision of linking ATX 3V with 3V_dual. Looks like in case of P3BF you can solder those two jumpers and thats it.
There might be potential problem if board maintains generating 3V_dual in standby mode? Maybe thats why there is provision for D34?

Excellent, now this board works with any card I throw at it!

I bridged D33 and lifted the leg of L16 (and bent it out at a sharp angle so that it would not make contact with the hole). The VIO jumper is on pins 2-3 (normal).

Quake3 1.32c (and 1.32 without the suffix) does not like the combination of DirectX 7 and the nvopengl.dll of Detonator 45.23, and that's why that was crashing. The 1999 Quake3 vanilla release installed without any patches runs DEMO001 just fine at 1024x768 with everything maxxed out, so that's that.

I also was able to find the P2B-S boardview in your link. Is there any provision for this mod on that board? If there is, it's not obvious. Would you have a look please (I have this board too)?

Since Quake crash wasnt caused by power issues maybe you dont need those mods after all.
P2B-s doesnt use 3.3V from ATX socket, all 3V pins are not connected. 3V is generated in upper right corner by smaller HIP chip and L13. Disconnect L13, connect to ATX socket pins 1 2.

rasz_pl wrote on 2024-02-12, 18:35:

Since Quake crash wasnt caused by power issues maybe you dont need those mods after all.
P2B-s doesnt use 3.3V from ATX socket, all 3V pins are not connected. 3V is generated in upper right corner by smaller HIP chip and L13. Disconnect L13, connect to ATX socket pins 1 2.

Thanks. For the P2B, I am almost certain it was required because I remember putting in an Fx5800 and that caused the system to become unstable. The P2B-S seems to have a better 3.3V circuit. I just tested it without the mod and the 5800 with Q3 and DEMO001 and it worked great.

In order to get anything definitive I'd have to undo the mod with the P2B and test again... If someone here asks in the future, I'm open to doing that.

Thanks again!