VOGONS


Identify motherboard Slot 1 SIS chipset

Topic actions

Reply 40 of 64, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-05-16, 21:37:
PC@LIVE wrote on 2021-05-16, 21:06:
Ok I checked. From a quick check the PIN A21 of the PCI and the PIN6 of the RAM are not connected. But the PINs are not very acc […]
Show full quote
weedeewee wrote on 2021-05-16, 20:31:

maybe check between pin 6 of sdram and pin A21 of pci to see if connected .

Ok I checked.
From a quick check the PIN A21 of the PCI and the PIN6 of the RAM are not connected.
But the PINs are not very accessible, I'll do another check from the back of the motherboard as soon as possible.
For the moment I have not done the PIN checks of the RAM, and I have to finish those of the PCI.

I don't know if it applies to this motherboard, but I know at least one (Abit KA7) where the RAM voltage supply is switched between the board's common 3.3V regulator (comes via the ATX 12V) and a separate 3.3V supply from the ATX 5VSB (for standby mode support). So for that board the RAM Vdd isn't directly connected to other 3.3V pins.

Still, all these 3.3V pins have to go somewhere.

Ok thanks for the info.
I had time today to do some checking. regarding the RAM I have identified where the + 3.3V PINs go, I didn't have time to check them all, those of the final part of the DIMM (PIN 73 84 157 168) go to PIN3 of Q4.
At this point I think it is useless to check the rest of the PINs, also because the RAM will be powered by the Q4 PIN3 (+ 3.47V measured).
Another check I did is the one on the other mosfets, in practice four are powered directly from the power supply, the current comes directly from the + 5V (red wire), and these are Q8 PIN2, Q22 PIN2, Q4 PIN2, Q7 PIN3.
The Q6 PIN2 is directly connected to the Q8 PIN3, the current measured between these PINs is about 1.85V.
At this point I miss to check the BXX side of the PCI slot, currently I have not yet identified where the + 3.3V current can come from, and I think it is essential for the functioning of the PC, but I'm not sure about this.

AMD 286-16 287-10 4MB HD 45MB VGA 256KB
AMD 386DX-40 Intel 387 8MB HD 81MB VGA 256KB
Cyrix 486DLC-40 IIT387-40 8MB VGA 512KB
AMD 5X86-133 16MB VGA VLB CL5428 2MB and many others
AMD K62+ 550 SOYO 5EMA+ and many others
AST Pentium Pro 200 MHz L2 256KB

Reply 41 of 64, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie

I just went back through and had a look at the measurements you took before. To summarize:
Corner:

Q4 - NFET  G1=7.8V  D2=5V    S3=3.47V

Q6 - NFET G1=6.3V D2=1.8V S3=0V
Q8 - NFET G1=5.6V D2=5V S3=1.9V

Below SlotA:

Q22 - NFET G1=6.6V  D2=5V    S3=3.3V
Q7 - LDO
Q5 - LDO

I couldn't work out from the previous post what the voltages on those LDO regulators was, but they should be 1-AdjustFeedback, 2-Vout, 3-Vin.

So as you've already identified, there are two plausible 3.3V sources, one is pin 3 of Q4 and the other is pin 3 of Q22. It's possible that one feeds the RAM and the other feeds the PCI and other things (chipset?). So at the moment nothing looks particularly wrong with those outputs. Q6&8 look like the usual upper-lower type regulator, probably switch from U6. All the G1 voltages are likely to be switching, so the meter will give an average. They're not stuck high or low, so whatever's controlling them is probably working.

If you could check the LDO voltages we can check they look reasonable as well. Although perhaps they're only used on an AT supply.

None of which explains why the 3.3V doesn't light up on your card. Do you know if it works in other motherboards?

One possible thing: Q4 pin 3 might be reading a bit high if it's supposed to be 3.3V. That could happen if there's excessive ripple on the line. I know my cheap meter doesn't average noisy signals very well, which it handy as it means it gives different voltage readings depending on which way around the probes are. So could you try measuring Q4 pin 3 with black probe on Ground (should be the same as you measured last time), and then again with the Red probe on Ground (should read around -3.3V)?

If (ignoring the +ve/-ve bit) they give different numbers, then it could be noise on the line, which might be a problem with capacitors.

Reply 42 of 64, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-05-17, 22:49:
I just went back through and had a look at the measurements you took before. To summarize: Corner: […]
Show full quote

I just went back through and had a look at the measurements you took before. To summarize:
Corner:

Q4 - NFET  G1=7.8V  D2=5V    S3=3.47V

Q6 - NFET G1=6.3V D2=1.8V S3=0V
Q8 - NFET G1=5.6V D2=5V S3=1.9V

Below SlotA:

Q22 - NFET G1=6.6V  D2=5V    S3=3.3V
Q7 - LDO
Q5 - LDO

I couldn't work out from the previous post what the voltages on those LDO regulators was, but they should be 1-AdjustFeedback, 2-Vout, 3-Vin.

So as you've already identified, there are two plausible 3.3V sources, one is pin 3 of Q4 and the other is pin 3 of Q22. It's possible that one feeds the RAM and the other feeds the PCI and other things (chipset?). So at the moment nothing looks particularly wrong with those outputs. Q6&8 look like the usual upper-lower type regulator, probably switch from U6. All the G1 voltages are likely to be switching, so the meter will give an average. They're not stuck high or low, so whatever's controlling them is probably working.

If you could check the LDO voltages we can check they look reasonable as well. Although perhaps they're only used on an AT supply.

None of which explains why the 3.3V doesn't light up on your card. Do you know if it works in other motherboards?

One possible thing: Q4 pin 3 might be reading a bit high if it's supposed to be 3.3V. That could happen if there's excessive ripple on the line. I know my cheap meter doesn't average noisy signals very well, which it handy as it means it gives different voltage readings depending on which way around the probes are. So could you try measuring Q4 pin 3 with black probe on Ground (should be the same as you measured last time), and then again with the Red probe on Ground (should read around -3.3V)?

If (ignoring the +ve/-ve bit) they give different numbers, then it could be noise on the line, which might be a problem with capacitors.

Ok thanks for the advice, some I don't think I understand them, but anyway as soon as I can I will make other measurements.
Today I measured the voltages of the two mosfets under the cpu, which I report here:
Q7 Pin 1 (+1.28V) | Pin2 (+2.50V) | Pin3 (+5,01V)
Q5 Pin1 (+0.27V) | Pin2 (+1.48V) | Pin3 (+3.47V)

Looking at the values ​​of the Q5 at PIN3 (+ 3.47V), the measured value is the same as that of the PIN3 of the Q4, so I think they are linked together, but I have to check, but I guess it is as I think.

Analyzing the measured values ​​with those of other motherboards for PIII, I realized that Q7 and Q5 are the equivalent of VR1 and VR2. Another thing that was useful to me is that almost all the voltages are quite similar, this makes me understand that the mosfets should all be working, and therefore the problem could be the lack of voltage (+ 3.3V) in the PCI PINs or other .
Another positive thing is that in two other three faulty motherboards I would have identified which mosfets could be replaced, among the various cards I have an ASUS P2BF, a SOYO BX, and other SIEMENS for PIII.
I will not deal with these at the moment, but I will do it as soon as I can.
I also checked the PINs of the PCI on the BXX side, PIN B25 is connected to B31 and B36 and to B41 and B43, it is also connected to PIN A21 and therefore also to PIN A27 A33 A39 and A45.
On the RAM I saw that there is also the voltage detected on the Q4 PIN3 (+ 3.47V).

AMD 286-16 287-10 4MB HD 45MB VGA 256KB
AMD 386DX-40 Intel 387 8MB HD 81MB VGA 256KB
Cyrix 486DLC-40 IIT387-40 8MB VGA 512KB
AMD 5X86-133 16MB VGA VLB CL5428 2MB and many others
AMD K62+ 550 SOYO 5EMA+ and many others
AST Pentium Pro 200 MHz L2 256KB

Reply 43 of 64, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-05-18, 19:19:

Ok thanks for the advice, some I don't think I understand them, but anyway as soon as I can I will make other measurements.

Thanks for the extra measurements, adding those to the table (and a few comments about what's there):
Corner:

Q4 - NFET  G1=7.8V  D2=5V    S3=3.47V - Generates 3.5V on pin 3 (probable RAM supply)

Q6 - NFET G1=6.3V D2=1.8V S3=0V - Lower FET, output 1.8V on pin 2
Q8 - NFET G1=5.6V D2=5V S3=1.9V - Upper FET, output 1.8V on pin 3

Below SlotA:

Q22 - NFET G1=6.6V   D2=5V       S3=3.3V - Outputs 3.3V pin 3
Q7 - LDO Adj1=1.28V Vout2=2.5V Vin3=5V - voltage regulator, 2.5V on pin 2
Q5 - LDO Adj1=0.27V Vout2=1.5V Vin3=3.47V - voltage regulator, 1.5V on pin 2 (Vin is probably from Q4 pin 3)

So I think that gets us all the supplies generated on the board. It's not important for the moment, but Q5 and Q7 are voltage regulators rather than FETs, so they burn more power, but there's less switching noise.

I also checked the PINs of the PCI on the BXX side, PIN B25 is connected to B31 and B36 and to B41 and B43, it is also connected to PIN A21 and therefore also to PIN A27 A33 A39 and A45.

On the RAM I saw that there is also the voltage detected on the Q4 PIN3 (+ 3.47V).

Ok, so all the 3.3V pins on the PCI slot are connected together, and we know from Weedeewee that A21 is what the card uses for detecting 3.3V. So it should be present. My guess is that Q22 pin 3 is supposed to be the PCI voltage and Q4 pin 3 is the RAM voltage. There's what looks like a transistor of some sort next to Q22 (D16). Could you check if any of the pins of D16 connect to Q22 pin 3, or to PCI A21? Also check the voltages on the pins of D16. I'm wondering if something is supposed to switch to connect the PCI slots to the 3.3V supply.

[actually , that D16 might just be switching the gate voltage for Q22. Still, worth checking anyway. Could you take a reasonably high resolution photo of the whole board?]

Reply 44 of 64, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-05-18, 20:08:
Thanks for the extra measurements, adding those to the table (and a few comments about what's there): Corner: […]
Show full quote
PC@LIVE wrote on 2021-05-18, 19:19:

Ok thanks for the advice, some I don't think I understand them, but anyway as soon as I can I will make other measurements.

Thanks for the extra measurements, adding those to the table (and a few comments about what's there):
Corner:

Q4 - NFET  G1=7.8V  D2=5V    S3=3.47V - Generates 3.5V on pin 3 (probable RAM supply)

Q6 - NFET G1=6.3V D2=1.8V S3=0V - Lower FET, output 1.8V on pin 2
Q8 - NFET G1=5.6V D2=5V S3=1.9V - Upper FET, output 1.8V on pin 3

Below SlotA:

Q22 - NFET G1=6.6V   D2=5V       S3=3.3V - Outputs 3.3V pin 3
Q7 - LDO Adj1=1.28V Vout2=2.5V Vin3=5V - voltage regulator, 2.5V on pin 2
Q5 - LDO Adj1=0.27V Vout2=1.5V Vin3=3.47V - voltage regulator, 1.5V on pin 2 (Vin is probably from Q4 pin 3)

So I think that gets us all the supplies generated on the board. It's not important for the moment, but Q5 and Q7 are voltage regulators rather than FETs, so they burn more power, but there's less switching noise.

I also checked the PINs of the PCI on the BXX side, PIN B25 is connected to B31 and B36 and to B41 and B43, it is also connected to PIN A21 and therefore also to PIN A27 A33 A39 and A45.

On the RAM I saw that there is also the voltage detected on the Q4 PIN3 (+ 3.47V).

Ok, so all the 3.3V pins on the PCI slot are connected together, and we know from Weedeewee that A21 is what the card uses for detecting 3.3V. So it should be present. My guess is that Q22 pin 3 is supposed to be the PCI voltage and Q4 pin 3 is the RAM voltage. There's what looks like a transistor of some sort next to Q22 (D16). Could you check if any of the pins of D16 connect to Q22 pin 3, or to PCI A21? Also check the voltages on the pins of D16. I'm wondering if something is supposed to switch to connect the PCI slots to the 3.3V supply.

[actually , that D16 might just be switching the gate voltage for Q22. Still, worth checking anyway. Could you take a reasonably high resolution photo of the whole board?]

Ok thanks for the info and suggestions, I hope they will be useful to fix the fault.
Ok I didn't know exactly Q5 and Q7 were voltage regulators, I thought they were mosfets, but thanks for the info.
Regarding the nearby transistor Q22 (signed D16), it says APL 4031S A0112, obviously looking at the code D16 it should be a diode, I don't know what type, I'll try to make the measurements you suggested, and if any PIN is connected to the Q22 PIN3.
I also tell you if this can be helpful, that
there is another APL 4031S A0112 near the Q4 signed D1, this maybe could have one of the PINs connected to the PIN3 of the Q4.
For the high resolution photo, I can do it but not immediately, if they are interested I can also do the back of the MB.

AMD 286-16 287-10 4MB HD 45MB VGA 256KB
AMD 386DX-40 Intel 387 8MB HD 81MB VGA 256KB
Cyrix 486DLC-40 IIT387-40 8MB VGA 512KB
AMD 5X86-133 16MB VGA VLB CL5428 2MB and many others
AMD K62+ 550 SOYO 5EMA+ and many others
AST Pentium Pro 200 MHz L2 256KB

Reply 45 of 64, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-05-18, 20:45:
Ok thanks for the info and suggestions, I hope they will be useful to fix the fault. Ok I didn't know exactly Q5 and Q7 were vol […]
Show full quote

Ok thanks for the info and suggestions, I hope they will be useful to fix the fault.
Ok I didn't know exactly Q5 and Q7 were voltage regulators, I thought they were mosfets, but thanks for the info.
Regarding the nearby transistor Q22 (signed D16), it says APL 4031S A0112, obviously looking at the code D16 it should be a diode, I don't know what type, I'll try to make the measurements you suggested, and if any PIN is connected to the Q22 PIN3.
I also tell you if this can be helpful, that
there is another APL 4031S A0112 near the Q4 signed D1, this maybe could have one of the PINs connected to the PIN3 of the Q4.
For the high resolution photo, I can do it but not immediately, if they are interested I can also do the back of the MB.

Yep, front and back would be nice to have.
I've found something else to measure... Chip U6, near Q6&8. That's the driver that switches Q6&8 to produce 1.8V. It looks like it's a HIP6004. If so, pin 12 should be a Power Good signal, and you should read as a +ve voltage when the output from Q6&8 is correct. Maybe the board uses that as a signal to switch on other bits of the board? I can't yet see anything that would act as the switch though.

Attachments

  • HIP6004.jpg
    Filename
    HIP6004.jpg
    File size
    87.47 KiB
    Views
    298 views
    File license
    Fair use/fair dealing exception

Reply 46 of 64, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-05-18, 21:33:
PC@LIVE wrote on 2021-05-18, 20:45:
Ok thanks for the info and suggestions, I hope they will be useful to fix the fault. Ok I didn't know exactly Q5 and Q7 were vol […]
Show full quote

Ok thanks for the info and suggestions, I hope they will be useful to fix the fault.
Ok I didn't know exactly Q5 and Q7 were voltage regulators, I thought they were mosfets, but thanks for the info.
Regarding the nearby transistor Q22 (signed D16), it says APL 4031S A0112, obviously looking at the code D16 it should be a diode, I don't know what type, I'll try to make the measurements you suggested, and if any PIN is connected to the Q22 PIN3.
I also tell you if this can be helpful, that
there is another APL 4031S A0112 near the Q4 signed D1, this maybe could have one of the PINs connected to the PIN3 of the Q4.
For the high resolution photo, I can do it but not immediately, if they are interested I can also do the back of the MB.

Yep, front and back would be nice to have.
I've found something else to measure... Chip U6, near Q6&8. That's the driver that switches Q6&8 to produce 1.8V. It looks like it's a HIP6004. If so, pin 12 should be a Power Good signal, and you should read as a +ve voltage when the output from Q6&8 is correct. Maybe the board uses that as a signal to switch on other bits of the board? I can't yet see anything that would act as the switch though.

Today I checked the two diodes D16 and D1, and I measured the voltages, both have the PIN3 connected to the PIN1 of the nearby Mosfet, then the D16 with the Q22, and the D1 with the Q4.
Having the PIN connected directly with the Mosfet, the measured voltage is the same, so for the D16 it is + 6.60V, while for the D1 it is + 7.83V.
Again for the two diodes, I verified that the second PIN (central) is connected directly to ground (black wire), so by measuring the voltage I have values ​​of 0V.
I measured the voltage of both PIN1 of the two diodes, in both there is a voltage of + 2.45V.
Therefore, summarizing the voltages in the PINs of the two diodes is the following:

D6 PIN (1) + 2.45V (2) 0V (3) + 6.60V
D1 PIN (1) + 2.45V (2) 0V (3) + 7.83V

For high resolution photos, I only have one question, because when I post them here they are automatically reduced, so how should I put them without changing the size?
Regarding the HIP6004 (U6 chip) I will do the PIN measurements, but it is a bit risky due to the proximity of the adjacent pins.

AMD 286-16 287-10 4MB HD 45MB VGA 256KB
AMD 386DX-40 Intel 387 8MB HD 81MB VGA 256KB
Cyrix 486DLC-40 IIT387-40 8MB VGA 512KB
AMD 5X86-133 16MB VGA VLB CL5428 2MB and many others
AMD K62+ 550 SOYO 5EMA+ and many others
AST Pentium Pro 200 MHz L2 256KB

Reply 47 of 64, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-05-19, 19:09:
Today I checked the two diodes D16 and D1, and I measured the voltages, both have the PIN3 connected to the PIN1 of the nearby M […]
Show full quote

Today I checked the two diodes D16 and D1, and I measured the voltages, both have the PIN3 connected to the PIN1 of the nearby Mosfet, then the D16 with the Q22, and the D1 with the Q4.
Having the PIN connected directly with the Mosfet, the measured voltage is the same, so for the D16 it is + 6.60V, while for the D1 it is + 7.83V.
Again for the two diodes, I verified that the second PIN (central) is connected directly to ground (black wire), so by measuring the voltage I have values ​​of 0V.
I measured the voltage of both PIN1 of the two diodes, in both there is a voltage of + 2.45V.
Therefore, summarizing the voltages in the PINs of the two diodes is the following:

D6 PIN (1) + 2.45V (2) 0V (3) + 6.60V
D1 PIN (1) + 2.45V (2) 0V (3) + 7.83V

I can't for the life of me find out what an APL 4031s is. The only 4031 device in a TO-92 case that I can find is a PNP transistor, and it's not that. It looks like they control the switching (and so the output voltage) for the FETs Q4 and Q22. My guess is that the output voltage from FET pin 3 (around 3.3V) is connected to the 4031s via a resistor divider to produce 2.45V when the output voltage is correct. If 4031s pin 1 goes above 2.45V (FET output voltage too high) then the 4031s switches to allow current to flow from pin 3 to pin 2, grounding the Gate of the FET, switching it off and causing the voltage on the FET output to drop. That drops the 4031s pin 1 beneath 2.45V and it stops conducting. Then the FET gate input gets pulled high to +12V via a resistor somewhere. Maybe.

That's mostly speculation, but the result would be that the 4031s is actively switching the FET off and on to keep the output voltage correct. And you've measured the output voltages of the FETs Q4 and Q22, and they both look ok.

So we're still left with the idea that pin 3 of one of those two FETs (I guess Q22) is supposed to connect to the 3.3V pins of the PCI slots. But the LED on the card doesn't light up. Ignoring the test card, could you double check if there's any voltage on pin A21 of the PCI slots?

For high resolution photos, I only have one question, because when I post them here they are automatically reduced, so how should I put them without changing the size?
Regarding the HIP6004 (U6 chip) I will do the PIN measurements, but it is a bit risky due to the proximity of the adjacent pins.

Yes, you'll need to be careful not to short PGOOD to GND. It probably won't break anything if it's only for a moment. It might be worth spending a bit of time tracing where the PCB trace goes (visual inspection and measuring resistance with the board off) in case there's an easier place to measure from. Also practice placing the probe on the pin with the board off.

I'm not sure how best to post the photos. Maybe (and there must be a better way than this) .zip them up?

Reply 48 of 64, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-05-20, 10:25:
I can't for the life of me find out what an APL 4031s is. The only 4031 device in a TO-92 case that I can find is a PNP transis […]
Show full quote
PC@LIVE wrote on 2021-05-19, 19:09:
Today I checked the two diodes D16 and D1, and I measured the voltages, both have the PIN3 connected to the PIN1 of the nearby M […]
Show full quote

Today I checked the two diodes D16 and D1, and I measured the voltages, both have the PIN3 connected to the PIN1 of the nearby Mosfet, then the D16 with the Q22, and the D1 with the Q4.
Having the PIN connected directly with the Mosfet, the measured voltage is the same, so for the D16 it is + 6.60V, while for the D1 it is + 7.83V.
Again for the two diodes, I verified that the second PIN (central) is connected directly to ground (black wire), so by measuring the voltage I have values ​​of 0V.
I measured the voltage of both PIN1 of the two diodes, in both there is a voltage of + 2.45V.
Therefore, summarizing the voltages in the PINs of the two diodes is the following:

D6 PIN (1) + 2.45V (2) 0V (3) + 6.60V
D1 PIN (1) + 2.45V (2) 0V (3) + 7.83V

I can't for the life of me find out what an APL 4031s is. The only 4031 device in a TO-92 case that I can find is a PNP transistor, and it's not that. It looks like they control the switching (and so the output voltage) for the FETs Q4 and Q22. My guess is that the output voltage from FET pin 3 (around 3.3V) is connected to the 4031s via a resistor divider to produce 2.45V when the output voltage is correct. If 4031s pin 1 goes above 2.45V (FET output voltage too high) then the 4031s switches to allow current to flow from pin 3 to pin 2, grounding the Gate of the FET, switching it off and causing the voltage on the FET output to drop. That drops the 4031s pin 1 beneath 2.45V and it stops conducting. Then the FET gate input gets pulled high to +12V via a resistor somewhere. Maybe.

That's mostly speculation, but the result would be that the 4031s is actively switching the FET off and on to keep the output voltage correct. And you've measured the output voltages of the FETs Q4 and Q22, and they both look ok.

So we're still left with the idea that pin 3 of one of those two FETs (I guess Q22) is supposed to connect to the 3.3V pins of the PCI slots. But the LED on the card doesn't light up. Ignoring the test card, could you double check if there's any voltage on pin A21 of the PCI slots?

For high resolution photos, I only have one question, because when I post them here they are automatically reduced, so how should I put them without changing the size?
Regarding the HIP6004 (U6 chip) I will do the PIN measurements, but it is a bit risky due to the proximity of the adjacent pins.

Yes, you'll need to be careful not to short PGOOD to GND. It probably won't break anything if it's only for a moment. It might be worth spending a bit of time tracing where the PCB trace goes (visual inspection and measuring resistance with the board off) in case there's an easier place to measure from. Also practice placing the probe on the pin with the board off.

I'm not sure how best to post the photos. Maybe (and there must be a better way than this) .zip them up?

Well what that APL4031 is I don't know, according to the serigraphs being signed as D .. they would be diodes, however in the end I don't think you care much about what they are, and your suggestions about it, and how they should work, I think they are correct, and I also think that both the Q22 and the Q4 work, so in fact, as you suggest, you need to check if there are + 3.3V on the PCI PINs, I will try to do this check, but if I can I will also check other PCI PINs, in particular I had read that on the PIN14 there would be a voltage of + 3.3V, if I find voltage somewhere I report the PINs and the voltages found.
Regarding the HIP6004 I checked the voltage on the PIN you indicated to me, on the PIN12 I found + 4.70V, at the moment I have not measured other PINs.
I also performed other measurements, on the inductors (indicated by L ..), these are the ones measured and the relative voltages:
L1 0V | L2 0V | L3 + 4.98V | L4 + 2.50V | L5 + 3.47V | L6 + 4.93V | L7 + 1.77V
I don't know if these measurements are useful, but they confirm that the voltage is present, for those that measure 0V I have to check if they are connected to ground (black wire) but I think it will be so.
I would also have the measurements of the SMD transistors (???), I measured four Q1 Q9 Q10 Q18, the PINs are always three (2 is the central), these are the voltages found:
Q1 PIN (1) + 0.57V PIN (2) 0V PIN (3) + 0.04V
Q9 PIN (1) 0V PIN (2) + 4.78V PIN (3) 0V
Q10 PIN (1) + 3.92V PIN (2) + 4.34V PIN (3) + 4.69V
Q18 PIN (1) 0V PIN (2) + 2.50V PIN (3) 0V
Apart from the PIN1 of the Q9 and the Q18 which I don't know where they go (they have a certain extent by testing them with the radio tester), the other PINs where I read 0V go directly to the black wire of the power supply.

In case I will try to start another MB Slot1 (or similar) not working, to see if it works and how the + 3.3V of the PCI is connected.

AMD 286-16 287-10 4MB HD 45MB VGA 256KB
AMD 386DX-40 Intel 387 8MB HD 81MB VGA 256KB
Cyrix 486DLC-40 IIT387-40 8MB VGA 512KB
AMD 5X86-133 16MB VGA VLB CL5428 2MB and many others
AMD K62+ 550 SOYO 5EMA+ and many others
AST Pentium Pro 200 MHz L2 256KB

Reply 49 of 64, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-05-20, 20:47:

need to check if there are + 3.3V on the PCI PINs, I will try to do this check, but if I can I will also check other PCI PINs, in particular I had read that on the PIN14 there would be a voltage of + 3.3V, if I find voltage somewhere I report the PINs and the voltages found.

Feel free to measure other pins as well, but I'm particularly interested in A21 because so far everything seems to say 3.3V is ok, but the test card says not, so I just want to double check the voltage on that pin. It's possible that all this worrying about the voltage regulators may be a big red herring, and it's actually all fine, and that 41 error is caused by something else.

Regarding the HIP6004 I checked the voltage on the PIN you indicated to me, on the PIN12 I found + 4.70V, at the moment I have not measured other PINs.

That's fine, I just wanted to make sure that it wasn't holding anything in reset. I think it probably goes to SIS 5595 and maybe other places.

I also performed other measurements, on the inductors (indicated by L ..), these are the ones measured and the relative voltages:
L1 0V | L2 0V | L3 + 4.98V | L4 + 2.50V | L5 + 3.47V | L6 + 4.93V | L7 + 1.77V
I don't know if these measurements are useful, but they confirm that the voltage is present, for those that measure 0V I have to check if they are connected to ground (black wire) but I think it will be so.

Double check if they are actually Ground. The others look like +5V?, 2.5V from Q7 pin 2, 3.47V from Q4 pin 3 (possible RAM supply), +5VSB? and the 1.8V supply (Q6 pin 2 & Q8 pin 3)

I would also have the measurements of the SMD transistors (???), I measured four Q1 Q9 Q10 Q18, the PINs are always three (2 is […]
Show full quote

I would also have the measurements of the SMD transistors (???), I measured four Q1 Q9 Q10 Q18, the PINs are always three (2 is the central), these are the voltages found:
Q1 PIN (1) + 0.57V PIN (2) 0V PIN (3) + 0.04V
Q9 PIN (1) 0V PIN (2) + 4.78V PIN (3) 0V
Q10 PIN (1) + 3.92V PIN (2) + 4.34V PIN (3) + 4.69V
Q18 PIN (1) 0V PIN (2) + 2.50V PIN (3) 0V
Apart from the PIN1 of the Q9 and the Q18 which I don't know where they go (they have a certain extent by testing them with the radio tester), the other PINs where I read 0V go directly to the black wire of the power supply.

So they look like they can switch various things to ground and (for Q9) +5. Q9 and Q18 are probably FETs, Q18 switched from the 2.5V supply (maybe to signal to something that the 2.5V supply is on) and Q18 could be switched from the PGOOD signal. Q1 might be an NPN transistor with pin 1 as the Base. It might go to a resistor, then the other side of the resistor is the signal that's actually switching it on. And Q10 might be a PNP transistor.

Those board level photos would help to figure out where everything is located.

In case I will try to start another MB Slot1 (or similar) not working, to see if it works and how the + 3.3V of the PCI is connected.

Always much easier to have something else to compare with.

Reply 50 of 64, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-05-20, 22:49:
Feel free to measure other pins as well, but I'm particularly interested in A21 because so far everything seems to say 3.3V is o […]
Show full quote
PC@LIVE wrote on 2021-05-20, 20:47:

need to check if there are + 3.3V on the PCI PINs, I will try to do this check, but if I can I will also check other PCI PINs, in particular I had read that on the PIN14 there would be a voltage of + 3.3V, if I find voltage somewhere I report the PINs and the voltages found.

Feel free to measure other pins as well, but I'm particularly interested in A21 because so far everything seems to say 3.3V is ok, but the test card says not, so I just want to double check the voltage on that pin. It's possible that all this worrying about the voltage regulators may be a big red herring, and it's actually all fine, and that 41 error is caused by something else.

Regarding the HIP6004 I checked the voltage on the PIN you indicated to me, on the PIN12 I found + 4.70V, at the moment I have not measured other PINs.

That's fine, I just wanted to make sure that it wasn't holding anything in reset. I think it probably goes to SIS 5595 and maybe other places.

I also performed other measurements, on the inductors (indicated by L ..), these are the ones measured and the relative voltages:
L1 0V | L2 0V | L3 + 4.98V | L4 + 2.50V | L5 + 3.47V | L6 + 4.93V | L7 + 1.77V
I don't know if these measurements are useful, but they confirm that the voltage is present, for those that measure 0V I have to check if they are connected to ground (black wire) but I think it will be so.

Double check if they are actually Ground. The others look like +5V?, 2.5V from Q7 pin 2, 3.47V from Q4 pin 3 (possible RAM supply), +5VSB? and the 1.8V supply (Q6 pin 2 & Q8 pin 3)

I would also have the measurements of the SMD transistors (???), I measured four Q1 Q9 Q10 Q18, the PINs are always three (2 is […]
Show full quote

I would also have the measurements of the SMD transistors (???), I measured four Q1 Q9 Q10 Q18, the PINs are always three (2 is the central), these are the voltages found:
Q1 PIN (1) + 0.57V PIN (2) 0V PIN (3) + 0.04V
Q9 PIN (1) 0V PIN (2) + 4.78V PIN (3) 0V
Q10 PIN (1) + 3.92V PIN (2) + 4.34V PIN (3) + 4.69V
Q18 PIN (1) 0V PIN (2) + 2.50V PIN (3) 0V
Apart from the PIN1 of the Q9 and the Q18 which I don't know where they go (they have a certain extent by testing them with the radio tester), the other PINs where I read 0V go directly to the black wire of the power supply.

So they look like they can switch various things to ground and (for Q9) +5. Q9 and Q18 are probably FETs, Q18 switched from the 2.5V supply (maybe to signal to something that the 2.5V supply is on) and Q18 could be switched from the PGOOD signal. Q1 might be an NPN transistor with pin 1 as the Base. It might go to a resistor, then the other side of the resistor is the signal that's actually switching it on. And Q10 might be a PNP transistor.

Those board level photos would help to figure out where everything is located.

In case I will try to start another MB Slot1 (or similar) not working, to see if it works and how the + 3.3V of the PCI is connected.

Always much easier to have something else to compare with.

Ok, I made other measurements because I was interested in verifying if I could find anomalous values ​​(or no voltage).
The functioning of the mosfets according to the detected values ​​seems to be normal, but this allows me to check other motherboards and to identify possible faulty components, so it is possible that I can fix other motherboards.
For this motherboard initially I thought that the failure was due to + 3.3V, because in fact it is a fundamental voltage for operation, but from the detected voltages it seems that it is missing only in the PCI (while the RAM are OK), of course maybe I could try other RAMs to see if by chance one is detected correctly, but from experience I would say that the RAMs I used are OK, the problem is in some PIN (in which there is no VOLTAGE), lately I have repaired the RAM bank of a PC Chips by doing a bridge in a broken track.
For the PIN A21 I did the verification, and the result is that
I did not find any voltage (measured 0.00V).
I have measured other PINs of the PCI slot, obviously in the PINs where there should be the + 3.3V I have detected no voltage), for now I have measured the PINs from A1 to A41, I will also try to detect the other PINs from A41 onwards and from B1 onwards.
Regarding the transitor, thanks for the info, I read even if it is written in lowercase the following abbreviations printed on the SMDs, on Q1 and Q9 and Q18 it says 1AM, on Q10 it says 2AX (even if it is not very clear) I don't know if these acronyms can be useful.

Attachments

AMD 286-16 287-10 4MB HD 45MB VGA 256KB
AMD 386DX-40 Intel 387 8MB HD 81MB VGA 256KB
Cyrix 486DLC-40 IIT387-40 8MB VGA 512KB
AMD 5X86-133 16MB VGA VLB CL5428 2MB and many others
AMD K62+ 550 SOYO 5EMA+ and many others
AST Pentium Pro 200 MHz L2 256KB

Reply 51 of 64, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-05-22, 19:39:

For the PIN A21 I did the verification, and the result is that
I did not find any voltage (measured 0.00V).

Ok, so definitely no 3.3V on the PCI slots. And yet what should be the 3.3V pins are connected together, so it should have been possible to route 3.3V to them. It might be possible to visually trace where the pins connect if you can post those photos of the top and bottom of the board.

I have just read that apparently early 5V signal level PCI slots didn't always connect the 3.3V power pins. Maybe that's happening here.

Regarding the transitor, thanks for the info, I read even if it is written in lowercase the following abbreviations printed on the SMDs, on Q1 and Q9 and Q18 it says 1AM, on Q10 it says 2AX (even if it is not very clear) I don't know if these acronyms can be useful.

I've got a list from somewhere of the short markings on SMD components. According to that, 2A is a MMBT3906 PNP transistor and 1AM is MMBT3094 NPN transistor. Using your numbering, Pin 1 is Base, Pin 2 (middle pin) is Collector and Pin 3 is Emitter. So a voltage difference between 1 and 3 connects 2 and 3. So you can see that Q1 and Q10 are turned on, and Q9 and Q18 are turned off.

Based on its location, I'd guess the Q10 is to do with clear CMOS, or maybe external battery selection. It looks like its Base (pin 1) is connected through a resistor to that jumper, so unplugging the jumper would stop the transistor conducting. No idea about Q9, looks like it's switched by whatever connects to R214. You might be able to work out what they're for by tracing what the Base connects to (it'll go to a resistor, then on to whatever actually controls it). Those small 3 pin transistors are probably all to do with signalling rather than power.

So, still no good idea why there's no 3.3V on the PCI slots, or if that problem is what causes the error 41 on the test card.

Reply 52 of 64, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-05-22, 22:14:
Ok, so definitely no 3.3V on the PCI slots. And yet what should be the 3.3V pins are connected together, so it should have been […]
Show full quote
PC@LIVE wrote on 2021-05-22, 19:39:

For the PIN A21 I did the verification, and the result is that
I did not find any voltage (measured 0.00V).

Ok, so definitely no 3.3V on the PCI slots. And yet what should be the 3.3V pins are connected together, so it should have been possible to route 3.3V to them. It might be possible to visually trace where the pins connect if you can post those photos of the top and bottom of the board.

I have just read that apparently early 5V signal level PCI slots didn't always connect the 3.3V power pins. Maybe that's happening here.

Regarding the transitor, thanks for the info, I read even if it is written in lowercase the following abbreviations printed on the SMDs, on Q1 and Q9 and Q18 it says 1AM, on Q10 it says 2AX (even if it is not very clear) I don't know if these acronyms can be useful.

I've got a list from somewhere of the short markings on SMD components. According to that, 2A is a MMBT3906 PNP transistor and 1AM is MMBT3094 NPN transistor. Using your numbering, Pin 1 is Base, Pin 2 (middle pin) is Collector and Pin 3 is Emitter. So a voltage difference between 1 and 3 connects 2 and 3. So you can see that Q1 and Q10 are turned on, and Q9 and Q18 are turned off.

Based on its location, I'd guess the Q10 is to do with clear CMOS, or maybe external battery selection. It looks like its Base (pin 1) is connected through a resistor to that jumper, so unplugging the jumper would stop the transistor conducting. No idea about Q9, looks like it's switched by whatever connects to R214. You might be able to work out what they're for by tracing what the Base connects to (it'll go to a resistor, then on to whatever actually controls it). Those small 3 pin transistors are probably all to do with signalling rather than power.

So, still no good idea why there's no 3.3V on the PCI slots, or if that problem is what causes the error 41 on the test card.

Ok thanks for the useful info.
Regarding the PCI slot and the absence of the + 3.3V, today I checked all the PINs of the PCI Slot both of side A and side B, I don't know if it will be useful, but maybe it allows to highlight any anomalous values, obviously I could write one for one the values ​​found, but before posting I would like to analyze them to understand if there is no voltage in other PINs (not necessarily from + 3.3V).
I also measured the ISA PIN slot by PIN, obviously there is no 3.3V here, but I also wanted to understand here if there is any voltage missing and in which PIN.
Regarding what you have read about the 3.3V of the PCI slot, it could actually be like this, but I wouldn't be sure of this, if it was a 486 PCI maybe it will be like that, being a Slot1 (even if AT), it seems possible that the 3.3V LED should be on and not off.
If instead it were as you say, the 3.3V LED off means that those PINs are not essential for operation, and therefore the fault is to be found elsewhere.
For the photos I try to put a couple and see if they are good, if I do other better ones or some details.
Regarding SMD transistors, the info you gave me are useful, I made those measurements to see if by chance they could be somehow connected, but from what you suggest they do not go to PCI, and therefore are useful as a reference in other bad motherboards.

So ultimately, assuming there is no voltage in those + 3.3V PINs, I should focus the search nearby, but at this point I don't know if the problem is due to the malfunctioning chipset.

Unfortunately I can't put high resolution photos here, because the file exceeds the maximum 5MB, even if compressed (zip).

Attachments

AMD 286-16 287-10 4MB HD 45MB VGA 256KB
AMD 386DX-40 Intel 387 8MB HD 81MB VGA 256KB
Cyrix 486DLC-40 IIT387-40 8MB VGA 512KB
AMD 5X86-133 16MB VGA VLB CL5428 2MB and many others
AMD K62+ 550 SOYO 5EMA+ and many others
AST Pentium Pro 200 MHz L2 256KB

Reply 53 of 64, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-05-23, 19:45:

Regarding what you have read about the 3.3V of the PCI slot, it could actually be like this, but I wouldn't be sure of this, if it was a 486 PCI maybe it will be like that, being a Slot1 (even if AT), it seems possible that the 3.3V LED should be on and not off.

That's true, so maybe it is more likely that 3.3V should be there.

Thanks for the photos. I notice that there are a couple of fuses, (F1 and F3 I think) near those 6 inductors. If you haven't already, check to see if they're still intact (measure the resistance across them, should be 0 ohms). Look for anything else marked F (e.g. where's F2?) and check it as well.

Unfortunately I can't put high resolution photos here, because the file exceeds the maximum 5MB, even if compressed (zip).

Ah. Maybe you can run them through an image editor first to scale them down / reduce the JPEG quality just enough until they're under 5MB?

Reply 54 of 64, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-05-23, 21:44:
That's true, so maybe it is more likely that 3.3V should be there. […]
Show full quote
PC@LIVE wrote on 2021-05-23, 19:45:

Regarding what you have read about the 3.3V of the PCI slot, it could actually be like this, but I wouldn't be sure of this, if it was a 486 PCI maybe it will be like that, being a Slot1 (even if AT), it seems possible that the 3.3V LED should be on and not off.

That's true, so maybe it is more likely that 3.3V should be there.

Thanks for the photos. I notice that there are a couple of fuses, (F1 and F3 I think) near those 6 inductors. If you haven't already, check to see if they're still intact (measure the resistance across them, should be 0 ohms). Look for anything else marked F (e.g. where's F2?) and check it as well.

Unfortunately I can't put high resolution photos here, because the file exceeds the maximum 5MB, even if compressed (zip).

Ah. Maybe you can run them through an image editor first to scale them down / reduce the JPEG quality just enough until they're under 5MB?

Ok I hope those photos can be useful, for those of more than 5MB I'll try to reduce them, but I don't know what the result will be.
For the fuses I checked those three marked with F1 F3 and F4, these are all intact (perfect), I don't think there is an F2, for the moment I don't see it, if I find it I check it.
For the PCI and the missing + 3.3V, I didn't have time to check the measurements made PIN by PIN, and I think I do the same check in another bad motherboard (with the + 3.3V LED on), that way I should see where the voltage could come from (from which mosfet).

AMD 286-16 287-10 4MB HD 45MB VGA 256KB
AMD 386DX-40 Intel 387 8MB HD 81MB VGA 256KB
Cyrix 486DLC-40 IIT387-40 8MB VGA 512KB
AMD 5X86-133 16MB VGA VLB CL5428 2MB and many others
AMD K62+ 550 SOYO 5EMA+ and many others
AST Pentium Pro 200 MHz L2 256KB

Reply 55 of 64, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-05-24, 20:31:

Ok I hope those photos can be useful, for those of more than 5MB I'll try to reduce them, but I don't know what the result will be.
For the fuses I checked those three marked with F1 F3 and F4, these are all intact (perfect), I don't think there is an F2, for the moment I don't see it, if I find it I check it.

Well, it was worth a try.

Might be worth going back to those inductors. You measured:
L1 0V | L2 0V | L3 + 4.98V | L4 + 2.50V | L5 + 3.47V | L6 + 4.93V | L7 + 1.77V

That looks like L3 is ATX +5V, L4 is Q7 pin 2, L5 is Q4 pin 3, L6 is ATX 5V (maybe 5VSB?) and L7 is Q6 pin 2/Q8 pin 3.

That covers all the voltages we've seen on the board, except for the 3.3V from Q22 pin 3 and 1.5V Q5 pin 2. Did you check if L1 and L2 are connected to Ground? If not, maybe one of them should be the 3.3V from Q22 pin 3. See if there's 0 ohms resistance between one of the 3.3V pins on the PCI slot and either of L1 or L2.

Reply 56 of 64, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-05-24, 21:06:
Well, it was worth a try. […]
Show full quote
PC@LIVE wrote on 2021-05-24, 20:31:

Ok I hope those photos can be useful, for those of more than 5MB I'll try to reduce them, but I don't know what the result will be.
For the fuses I checked those three marked with F1 F3 and F4, these are all intact (perfect), I don't think there is an F2, for the moment I don't see it, if I find it I check it.

Well, it was worth a try.

Might be worth going back to those inductors. You measured:
L1 0V | L2 0V | L3 + 4.98V | L4 + 2.50V | L5 + 3.47V | L6 + 4.93V | L7 + 1.77V

That looks like L3 is ATX +5V, L4 is Q7 pin 2, L5 is Q4 pin 3, L6 is ATX 5V (maybe 5VSB?) and L7 is Q6 pin 2/Q8 pin 3.

That covers all the voltages we've seen on the board, except for the 3.3V from Q22 pin 3 and 1.5V Q5 pin 2. Did you check if L1 and L2 are connected to Ground? If not, maybe one of them should be the 3.3V from Q22 pin 3. See if there's 0 ohms resistance between one of the 3.3V pins on the PCI slot and either of L1 or L2.

Yes ok thanks.
You can do these checks tomorrow, and in fact it could be as you suggest.
Because it is actually strange that there is no voltage, those inductors L1 and L2 are placed between the ISA and PCI slots, near the CD-IN audio connectors.

AMD 286-16 287-10 4MB HD 45MB VGA 256KB
AMD 386DX-40 Intel 387 8MB HD 81MB VGA 256KB
Cyrix 486DLC-40 IIT387-40 8MB VGA 512KB
AMD 5X86-133 16MB VGA VLB CL5428 2MB and many others
AMD K62+ 550 SOYO 5EMA+ and many others
AST Pentium Pro 200 MHz L2 256KB

Reply 57 of 64, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-05-24, 21:38:

Yes ok thanks.
You can do these checks tomorrow, and in fact it could be as you suggest.
Because it is actually strange that there is no voltage, those inductors L1 and L2 are placed between the ISA and PCI slots, near the CD-IN audio connectors.

Ah, I can't count. I meant the 6 inductors by the floppy connector (I can't read what they're labelled as on the current photo). Won't hurt to check those by the audio port either (I guess they're to do with the speaker out).

Reply 58 of 64, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-05-24, 21:49:
PC@LIVE wrote on 2021-05-24, 21:38:

Yes ok thanks.
You can do these checks tomorrow, and in fact it could be as you suggest.
Because it is actually strange that there is no voltage, those inductors L1 and L2 are placed between the ISA and PCI slots, near the CD-IN audio connectors.

Ah, I can't count. I meant the 6 inductors by the floppy connector (I can't read what they're labelled as on the current photo). Won't hurt to check those by the audio port either (I guess they're to do with the speaker out).

Yes ok thanks.
I understand it is a group of inductors, they are signed as L8 L9 L10 L11 L14 L15, even if the order is different from what I wrote.
Looking at them I noticed that some seem to have a leak of something, like if they had melted, here could be the problem here, if this is the problem here I think I can check tomorrow, if they are interrupted I will try to recover similar ones from some scrap MB.

AMD 286-16 287-10 4MB HD 45MB VGA 256KB
AMD 386DX-40 Intel 387 8MB HD 81MB VGA 256KB
Cyrix 486DLC-40 IIT387-40 8MB VGA 512KB
AMD 5X86-133 16MB VGA VLB CL5428 2MB and many others
AMD K62+ 550 SOYO 5EMA+ and many others
AST Pentium Pro 200 MHz L2 256KB

Reply 59 of 64, by PC@LIVE

User metadata
Rank Member
Rank
Member

Today I did some checks on the inductors, and they are all ok, although some have signs of melting.
For completeness I must add that these L8 L9 L10 L11 L14 L15, go to the connection socket of the ATX FORM CARD, in which there are the PS2 port, the two USB and one IR.
I have detected the values ​​and they are these:
L8 + 4.99V | L10 + 4.99V | L11 + 0.02V | L9 + 0.02V | L15 + 4.89V | L14 + 4.70V
Of these, both L11 and L9 are connected directly to ground (black wire).
I also measured the voltage of three other inductors L12 L13 L16, the voltages present are these:
L12 + 4.65V | L13 + 4.65V | L16 + 4.99V
Regarding fuses F1 and F3 are those of the USB, while F4 is that of the keyboard, regarding the latter I measured the PIN of the connector, and there is voltage.
Looking at the voltages of the PCI PIN by PIN, I have voltages of 3.44V on PIN A6 A7 B7 B8, of 3.30V on PIN A15, of 3.47V on PIN A17 B44, of 3.52V on PIN A34, while in PIN where there should be the + 3.3V I have not found any voltage.

Attachments

AMD 286-16 287-10 4MB HD 45MB VGA 256KB
AMD 386DX-40 Intel 387 8MB HD 81MB VGA 256KB
Cyrix 486DLC-40 IIT387-40 8MB VGA 512KB
AMD 5X86-133 16MB VGA VLB CL5428 2MB and many others
AMD K62+ 550 SOYO 5EMA+ and many others
AST Pentium Pro 200 MHz L2 256KB