VOGONS


Test and troubleshoot PC@LIVE motherboards

Topic actions

Reply 20 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-11-17, 19:11:

Thanks for your valuable suggestions, I'll try to do other tests, the difficult part of the SMDs is to remove them without damaging the tracks. Today I removed the capacitor that seemed to be shorted, in reality by testing it it was found to be working, I did not do further checks, since once removed it is not shorted.
Removing it you can see the tracks below, the - goes to ground, while the + goes to the PIN2 of the AMS 1085CM, two SMDs are connected to the same track, and both seem to be shorted, being more problematic for me to remove them, I check the + and the - of the capacitor removed, and again surprise it is shorted.

Yeah, the problem will be figuring which component is causing the short. Simple example, imagine there are 4 capacitors between Vout and Gnd, but one has failed short (C3 in this example):

__________________________ Vout __________________________
| | | | | | |
=C1 =C2 |C3 =C4 =C2 |C3 =C4
| | | | | | |
-------------------------- GND --------------------------

If you measure the resistance across any of the capacitors it will show as a short because C3 has failed. If you removed C1 and tested it you'd find that it was ok. If you then checked the pads where it had been fitted, the board would still show a short because the actual problem (C3) was still there. You'd only find the actual fault when you removed C3, and all of a sudden the short would disappear.

So I check the track of the - and the PINs of the power supply collector (20 PIN), surprisingly I find the + 3.3V shorted, doing further investigations, I find why, in JP1 there are bridges between the PINs 1-2 3-4 and 5-6, basically the - of the capacitor is connected to + 3.3V, this of course should be normal, since bridges have always been like this.
A further check performed with the diode test of the digital tester, shows a reading of about 20 between PIN 1 and 2 of the AMS 1085CM, even by inverting the leads, I don't know if it is an anomalous value, but most likely there is a fault, unfortunately I don't have another one with which to check or compare.

Hmm, that's odd. Best I can tell is that the copper area by JP1 should be 3.3V, supplying pins B113, B117 & B121 on the slot 1 connector. The copper area by the capacitor you've removed should be Gnd the supplies some pins on the A side of the connector. So the layout looks right, but they definitely shouldn't be shorted together. Are they actually short, i.e. they measure close to 0 ohms (certainly <2ohms)?

20 ohms between Vadj and Vout of the 1085 sounds a bit on the low side, but that'd depend on the values of the resistors that actually control what the output voltage should be. Can you take a close up photo of around the 1085? What's the resistance from Vadj (pin 1) to GND?

Reply 21 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-11-17, 20:34:
Yeah, the problem will be figuring which component is causing the short. Simple example, imagine there are 4 capacitors between […]
Show full quote
PC@LIVE wrote on 2021-11-17, 19:11:

Thanks for your valuable suggestions, I'll try to do other tests, the difficult part of the SMDs is to remove them without damaging the tracks. Today I removed the capacitor that seemed to be shorted, in reality by testing it it was found to be working, I did not do further checks, since once removed it is not shorted.
Removing it you can see the tracks below, the - goes to ground, while the + goes to the PIN2 of the AMS 1085CM, two SMDs are connected to the same track, and both seem to be shorted, being more problematic for me to remove them, I check the + and the - of the capacitor removed, and again surprise it is shorted.

Yeah, the problem will be figuring which component is causing the short. Simple example, imagine there are 4 capacitors between Vout and Gnd, but one has failed short (C3 in this example):

__________________________ Vout __________________________
| | | | | | |
=C1 =C2 |C3 =C4 =C2 |C3 =C4
| | | | | | |
-------------------------- GND --------------------------

If you measure the resistance across any of the capacitors it will show as a short because C3 has failed. If you removed C1 and tested it you'd find that it was ok. If you then checked the pads where it had been fitted, the board would still show a short because the actual problem (C3) was still there. You'd only find the actual fault when you removed C3, and all of a sudden the short would disappear.

So I check the track of the - and the PINs of the power supply collector (20 PIN), surprisingly I find the + 3.3V shorted, doing further investigations, I find why, in JP1 there are bridges between the PINs 1-2 3-4 and 5-6, basically the - of the capacitor is connected to + 3.3V, this of course should be normal, since bridges have always been like this.
A further check performed with the diode test of the digital tester, shows a reading of about 20 between PIN 1 and 2 of the AMS 1085CM, even by inverting the leads, I don't know if it is an anomalous value, but most likely there is a fault, unfortunately I don't have another one with which to check or compare.

Hmm, that's odd. Best I can tell is that the copper area by JP1 should be 3.3V, supplying pins B113, B117 & B121 on the slot 1 connector. The copper area by the capacitor you've removed should be Gnd the supplies some pins on the A side of the connector. So the layout looks right, but they definitely shouldn't be shorted together. Are they actually short, i.e. they measure close to 0 ohms (certainly <2ohms)?

20 ohms between Vadj and Vout of the 1085 sounds a bit on the low side, but that'd depend on the values of the resistors that actually control what the output voltage should be. Can you take a close up photo of around the 1085? What's the resistance from Vadj (pin 1) to GND?

Thanks again. I try to explain better, I understand your example on the short, which disappears only by removing the faulty component.
And that's what I should be doing by removing the SMDs.
Then removed the EC21 capacitor, which I found good (at least it seems), but for more precision I have to test it with an atlas peak ESR70, in order to see if the capacitance and the ESR are normal, the short has not disappeared, and therefore it must be one of the SMDs or the 1085, because I don't see anything else connected to the PIN2.
The other check I performed is from the - of EC21 to the socket for the ATX connector, the - is connected to the black wires, and this is fine, but it is strangely connected to the + 3.3V as I said through the bridges on the PINs of JP1, in the photo I had put before you can see them (both front and back).
But it is possible that there are other + 3.3V tracks going to slot1, they are probably under the socket, and they are not easily visible.
I can still try to find out if the PINs of Slot1 are connected to the + 3.3V of the socket, so maybe I will use a SLOT1-SKT370 adapter that has the numbered pins, in order to easily find which ones they are.
I put the photos of 1085, I don't know if they are better, try to see if they can fit, to zoom in more I should use a lens.

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 22 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-11-17, 21:46:
Thanks again. I try to explain better, I understand your example on the short, which disappears only by removing the faulty comp […]
Show full quote

Thanks again. I try to explain better, I understand your example on the short, which disappears only by removing the faulty component.
And that's what I should be doing by removing the SMDs.
Then removed the EC21 capacitor, which I found good (at least it seems), but for more precision I have to test it with an atlas peak ESR70, in order to see if the capacitance and the ESR are normal, the short has not disappeared, and therefore it must be one of the SMDs or the 1085, because I don't see anything else connected to the PIN2.
The other check I performed is from the - of EC21 to the socket for the ATX connector, the - is connected to the black wires, and this is fine, but it is strangely connected to the + 3.3V as I said through the bridges on the PINs of JP1, in the photo I had put before you can see them (both front and back).
But it is possible that there are other + 3.3V tracks going to slot1, they are probably under the socket, and they are not easily visible.
I can still try to find out if the PINs of Slot1 are connected to the + 3.3V of the socket, so maybe I will use a SLOT1-SKT370 adapter that has the numbered pins, in order to easily find which ones they are.

I think I understand what you've measured, but I don't think the 3.3V to GND short is caused by those JP1 jumpers. I think they're supposed to connect the 3.3V supply to part of the slot connector. So if there is a short between 3.3V and GND then the problem is somewhere else. What's the resistance you measure between 3.3V and GND?

SplitPowerCopper.jpg
Filename
SplitPowerCopper.jpg
File size
63.8 KiB
Views
238 views
File license
Fair use/fair dealing exception

I put the photos of 1085, I don't know if they are better, try to see if they can fit, to zoom in more I should use a lens.

Ok, I think I can see the 2 resistors that set the output voltage of the 1085. They should both go to pin 1 (Vadj), one should go to pin 2 and the other should go to GND. Can you read off the numbers on these? If we know the value then we can work out what the voltage should be.

1085_Resistors.jpg
Filename
1085_Resistors.jpg
File size
77.41 KiB
Views
238 views
File license
Fair use/fair dealing exception

Reply 23 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-11-17, 22:22:
I think I understand what you've measured, but I don't think the 3.3V to GND short is caused by those JP1 jumpers. I think they […]
Show full quote
PC@LIVE wrote on 2021-11-17, 21:46:
Thanks again. I try to explain better, I understand your example on the short, which disappears only by removing the faulty comp […]
Show full quote

Thanks again. I try to explain better, I understand your example on the short, which disappears only by removing the faulty component.
And that's what I should be doing by removing the SMDs.
Then removed the EC21 capacitor, which I found good (at least it seems), but for more precision I have to test it with an atlas peak ESR70, in order to see if the capacitance and the ESR are normal, the short has not disappeared, and therefore it must be one of the SMDs or the 1085, because I don't see anything else connected to the PIN2.
The other check I performed is from the - of EC21 to the socket for the ATX connector, the - is connected to the black wires, and this is fine, but it is strangely connected to the + 3.3V as I said through the bridges on the PINs of JP1, in the photo I had put before you can see them (both front and back).
But it is possible that there are other + 3.3V tracks going to slot1, they are probably under the socket, and they are not easily visible.
I can still try to find out if the PINs of Slot1 are connected to the + 3.3V of the socket, so maybe I will use a SLOT1-SKT370 adapter that has the numbered pins, in order to easily find which ones they are.

I think I understand what you've measured, but I don't think the 3.3V to GND short is caused by those JP1 jumpers. I think they're supposed to connect the 3.3V supply to part of the slot connector. So if there is a short between 3.3V and GND then the problem is somewhere else. What's the resistance you measure between 3.3V and GND?
SplitPowerCopper.jpg

I put the photos of 1085, I don't know if they are better, try to see if they can fit, to zoom in more I should use a lens.

Ok, I think I can see the 2 resistors that set the output voltage of the 1085. They should both go to pin 1 (Vadj), one should go to pin 2 and the other should go to GND. Can you read off the numbers on these? If we know the value then we can work out what the voltage should be.
1085_Resistors.jpg

I have to say thank you, in fact the zone you marked as + 3.3V has at least a couple of tracks going to Slot1, for the moment I couldn't see which PINs of Slot1 are connected, actually maybe it would be easier to spot them from below , but unfortunately that short makes things more difficult.
I'll try to better analyze the + 3.3V area, maybe there is some capacitor or other fault causing the short.
I measured with diode test tester and the display shows 001, so it is as connected directly, this had made me think that it could be the jumpers of JP1.
Anyway about the + 3.3V, according to what you can see from the post card, the circuit is working, because the LED is on, even if I think it refers to that of the PCI slot.
Regarding the two SMD resistors, they are indicated in the MB as R94 and R95, the capacitor instead is BC35. In R95 it seems to be written X6E, while in R94 I read 10A, I don't know if they are strange values, usually they should have 3 digits, I hope this helps you to understand how many OHms there are.

Today the six MB that I had ordered have arrived, but for the moment I keep them in part, first I would like to finish the work on this 6ABX2V.

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 24 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-11-18, 20:01:

I have to say thank you, in fact the zone you marked as + 3.3V has at least a couple of tracks going to Slot1, for the moment I couldn't see which PINs of Slot1 are connected, actually maybe it would be easier to spot them from below , but unfortunately that short makes things more difficult.

I think it should connect to B113, B117 & B121 (using the pinout from here: http://ps-2.kev009.com/eprmhtml/eprma/f706.htm )

I measured with diode test tester and the display shows 001, so it is as connected directly, this had made me think that it could be the jumpers of JP1.
Anyway about the + 3.3V, according to what you can see from the post card, the circuit is working, because the LED is on, even if I think it refers to that of the PCI slot.

I think the way the diode tester works is measuring what voltage is required for current to start flowing (threshold voltage). If there's a resistive load then current will start flowing even at very low voltages.

So I think it's better to measure the actual resistance. Given the post card says 3.3V is ok, and the power supply didn't cut out, then I suspect the 3.3V supply is actually ok.

Regarding the two SMD resistors, they are indicated in the MB as R94 and R95, the capacitor instead is BC35. In R95 it seems to be written X6E, while in R94 I read 10A, I don't know if they are strange values, usually they should have 3 digits, I hope this helps you to understand how many OHms there are.

I think those are EIA-96 markings: https://www.hobby-hour.com/electronics/smdcalc.php
They should be two numbers (resistor value) and a letter (power of 10 multiplier), so I think the X6E might be upside down and actually be 39X. 39X is 24.9 ohms and 10A is 124 ohms.
Plugging those into 1.25*(1+R2/R1), with R2 as 24.9 and R1 as 124 gives 1.25*(1+24.9/124)=1.5V

R2 should go from pin 1 to Gnd, R1 should go from pin 2 to pin 1. So, when you measured 20 ohms from pin 1 to pin 2, it should be ~124 ohms. But the 20 ohms reading is because of the short. Pin 2 is currently shorted to ground, so those two resistors are effectively in parallel with each other. And 24.9 || 124 = (24.9*124 / (24.9+124)) = 20.7 ohms.

So all that confirms that there is a short from Vout (pin 2) to Gnd, and that Vout should be 1.5V when it's working. Find that short and it may fix the board. I think the 1085 has an internal thermal shutdown, so it might actually be ok.

[edit: I was using R1 and R2 because they're what's used in the 1085 datasheet. In this case, R95 = 39X = 24.9ohm = R2 and R94 = 10A = 124ohm = R1]

Reply 25 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-11-18, 20:53:
I think it should connect to B113, B117 & B121 (using the pinout from here: http://ps-2.kev009.com/eprmhtml/eprma/f706.htm ) […]
Show full quote
PC@LIVE wrote on 2021-11-18, 20:01:

I have to say thank you, in fact the zone you marked as + 3.3V has at least a couple of tracks going to Slot1, for the moment I couldn't see which PINs of Slot1 are connected, actually maybe it would be easier to spot them from below , but unfortunately that short makes things more difficult.

I think it should connect to B113, B117 & B121 (using the pinout from here: http://ps-2.kev009.com/eprmhtml/eprma/f706.htm )

I measured with diode test tester and the display shows 001, so it is as connected directly, this had made me think that it could be the jumpers of JP1.
Anyway about the + 3.3V, according to what you can see from the post card, the circuit is working, because the LED is on, even if I think it refers to that of the PCI slot.

I think the way the diode tester works is measuring what voltage is required for current to start flowing (threshold voltage). If there's a resistive load then current will start flowing even at very low voltages.

So I think it's better to measure the actual resistance. Given the post card says 3.3V is ok, and the power supply didn't cut out, then I suspect the 3.3V supply is actually ok.

Regarding the two SMD resistors, they are indicated in the MB as R94 and R95, the capacitor instead is BC35. In R95 it seems to be written X6E, while in R94 I read 10A, I don't know if they are strange values, usually they should have 3 digits, I hope this helps you to understand how many OHms there are.

I think those are EIA-96 markings: https://www.hobby-hour.com/electronics/smdcalc.php
They should be two numbers (resistor value) and a letter (power of 10 multiplier), so I think the X6E might be upside down and actually be 39X. 39X is 24.9 ohms and 10A is 124 ohms.
Plugging those into 1.25*(1+R2/R1), with R2 as 24.9 and R1 as 124 gives 1.25*(1+24.9/124)=1.5V

R2 should go from pin 1 to Gnd, R1 should go from pin 2 to pin 1. So, when you measured 20 ohms from pin 1 to pin 2, it should be ~124 ohms. But the 20 ohms reading is because of the short. Pin 2 is currently shorted to ground, so those two resistors are effectively in parallel with each other. And 24.9 || 124 = (24.9*124 / (24.9+124)) = 20.7 ohms.

So all that confirms that there is a short from Vout (pin 2) to Gnd, and that Vout should be 1.5V when it's working. Find that short and it may fix the board. I think the 1085 has an internal thermal shutdown, so it might actually be ok.

[edit: I was using R1 and R2 because they're what's used in the 1085 datasheet. In this case, R95 = 39X = 24.9ohm = R2 and R94 = 10A = 124ohm = R1]

Thanks my friend, your suggestions are very valuable.
Meanwhile I was checking the + 3.3V, I saw what the PINs are and they should actually be the ones you indicated, looking closer I also saw the third track connected to the B121, and obviously I found the other two B117 and B113.
But from what you suggest everything should be ok, if I wanted I would have an analog multimeter to measure the OHm, but I have to check the batteries because the values ​​measured by setting X1 = OHm do not seem right.
Yes, I think it's a 39X too, sometimes I confuse the numbers for letters, I think the values ​​you have indicated can be verified, however I thought if lifting the PIN1 of 1085 we could make some useful measurements?
However the short at this point could be the SMD BC39 of the PIN2 (near the TURBO66 jumper), unfortunately with the Atlas Peak I do not think it can be tested, but in case I would try to remove it and I would also remove the Turbo66 jumper to make room, I think it can be him because I measure with the diode test 001, while the other BC41 measures about 020.

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 26 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-11-18, 21:30:
Thanks my friend, your suggestions are very valuable. Meanwhile I was checking the + 3.3V, I saw what the PINs are and they shou […]
Show full quote

Thanks my friend, your suggestions are very valuable.
Meanwhile I was checking the + 3.3V, I saw what the PINs are and they should actually be the ones you indicated, looking closer I also saw the third track connected to the B121, and obviously I found the other two B117 and B113.
But from what you suggest everything should be ok, if I wanted I would have an analog multimeter to measure the OHm, but I have to check the batteries because the values ​​measured by setting X1 = OHm do not seem right.
Yes, I think it's a 39X too, sometimes I confuse the numbers for letters, I think the values ​​you have indicated can be verified, however I thought if lifting the PIN1 of 1085 we could make some useful measurements?
However the short at this point could be the SMD BC39 of the PIN2 (near the TURBO66 jumper), unfortunately with the Atlas Peak I do not think it can be tested, but in case I would try to remove it and I would also remove the Turbo66 jumper to make room, I think it can be him because I measure with the diode test 001, while the other BC41 measures about 020.

Shouldn't need to lift pin 1, I hope. Remove and check, as you say, BC39 and also the SMD capacitor near the 10A R94 resistor that I think is connected to pin 2 (BC35? I can't quite see because of that yellow plastic thing (temperature sensor?). BC41 is probably ok, it's connected to pin 1, which I don't think is where the problem is. It really would be very useful to have some way to measure resistance, analog or digital. I'm not sure how much information a diode test is going to give. Which Atlas do you have?

Reply 27 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-11-18, 22:42:
PC@LIVE wrote on 2021-11-18, 21:30:
Thanks my friend, your suggestions are very valuable. Meanwhile I was checking the + 3.3V, I saw what the PINs are and they shou […]
Show full quote

Thanks my friend, your suggestions are very valuable.
Meanwhile I was checking the + 3.3V, I saw what the PINs are and they should actually be the ones you indicated, looking closer I also saw the third track connected to the B121, and obviously I found the other two B117 and B113.
But from what you suggest everything should be ok, if I wanted I would have an analog multimeter to measure the OHm, but I have to check the batteries because the values ​​measured by setting X1 = OHm do not seem right.
Yes, I think it's a 39X too, sometimes I confuse the numbers for letters, I think the values ​​you have indicated can be verified, however I thought if lifting the PIN1 of 1085 we could make some useful measurements?
However the short at this point could be the SMD BC39 of the PIN2 (near the TURBO66 jumper), unfortunately with the Atlas Peak I do not think it can be tested, but in case I would try to remove it and I would also remove the Turbo66 jumper to make room, I think it can be him because I measure with the diode test 001, while the other BC41 measures about 020.

Shouldn't need to lift pin 1, I hope. Remove and check, as you say, BC39 and also the SMD capacitor near the 10A R94 resistor that I think is connected to pin 2 (BC35? I can't quite see because of that yellow plastic thing (temperature sensor?). BC41 is probably ok, it's connected to pin 1, which I don't think is where the problem is. It really would be very useful to have some way to measure resistance, analog or digital. I'm not sure how much information a diode test is going to give. Which Atlas do you have?

Hello and thanks for the tips.
Therefore the condenser with the covered writing (it is the temperature sensor) is BC35.
The atlas Peak I have is an ESR70, for electrolytics it is very useful, some if visually perfect are broken, and often even on-board can make you understand if they do not go (but not always so after removal), it is It has been useful for several broken PSUs that I have repaired, but I have other broken ones with other problems, for the moment I don't want to dedicate some time to them, a couple I would like to fix in the future (also a Corsair HX750).
Returning to the MB, and the SMDs in short, I took a look at the notes written years ago when I started to repair it, I found one very useful (I think), in practice it is so tester on diode test, black lead to ground, lead red first on one side and then on the other of the SMD, if it sounds from both it is shorted.
Applying it to the three SMDs I would have these results,
BC35 I read 001 on both sides
BC39 I read 001 on both sides
BC41 I read 001 on one side and about 020 on the other.
So at this point I too believe that BC41 may be okay, while BC35 and BC39 must be removed and verified, if they confirm the short they should be replaced, the only thing I don't know if any one is fine or you have to find them the same, I still have some MB scrap from which I can take some if necessary.
For the analog Tester, the one I use is quite good, with the readings in OHmX1 you can see the shorts in the transistors well (for example), unfortunately I don't know if it is a battery problem or other, joining the two tips marks between 5 and 10 (even by adjusting the ADJ wheel), and therefore I do not understand if the value is good or not, tomorrow I try to change both batteries (two 1.5V and one 9V), they are probably the ones who show unreliable readings.

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 28 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-11-19, 20:33:

So at this point I too believe that BC41 may be okay, while BC35 and BC39 must be removed and verified

That sounds right to me. Hopefully one of them has failed, or else it means the short is somewhere else.

, if they confirm the short they should be replaced, the only thing I don't know if any one is fine or you have to find them the same, I still have some MB scrap from which I can take some if necessary.

You can probably use the ESR70 to measure the SMD capacitors. I'm guessing they'll be around 10uF, and the datasheet says it can measure down to 1uF. The exact value probably doesn't matter too much, so if you can take a capacitor from another board then that should be ok.

tomorrow I try to change both batteries (two 1.5V and one 9V), they are probably the ones who show unreliable readings.

Sounds possible, moving the needle all the way over to 0 will need the most power.

Reply 29 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-11-19, 21:28:
That sounds right to me. Hopefully one of them has failed, or else it means the short is somewhere else. […]
Show full quote
PC@LIVE wrote on 2021-11-19, 20:33:

So at this point I too believe that BC41 may be okay, while BC35 and BC39 must be removed and verified

That sounds right to me. Hopefully one of them has failed, or else it means the short is somewhere else.

, if they confirm the short they should be replaced, the only thing I don't know if any one is fine or you have to find them the same, I still have some MB scrap from which I can take some if necessary.

You can probably use the ESR70 to measure the SMD capacitors. I'm guessing they'll be around 10uF, and the datasheet says it can measure down to 1uF. The exact value probably doesn't matter too much, so if you can take a capacitor from another board then that should be ok.

tomorrow I try to change both batteries (two 1.5V and one 9V), they are probably the ones who show unreliable readings.

Sounds possible, moving the needle all the way over to 0 will need the most power.

I have performed further checks on the suspicious (faulty?) SMDs, to verify them I used the Atlas ESR70, I report the results, but I anticipate that the capacity has not been displayed, maybe it is normal for them or maybe you need to remove them to be able to view it.
The results are as follows:
BC39 ESR = 0.40OHm
BC35 ESR = 0.08OHm
BC41 ESR = 8.8OHm
BC40 ESR = 0.10OHm
From the measured values ​​I do not understand if they are actually short, but I would say that BC35 probably could be, BC39 could be too, while for BC41 the value could be influenced and be strange.
For comparison I also measured BC40, the measured value is not very reassuring, but since it looks fine, I don't know why I detect such a low value.
A further check I performed is the visual one, zooming in I look for traces of burns or any cracks, to see them they look fine, the only thing I notice is the color that seems darker than the others, I don't know if they are gray or dark purple, if they were purple compared to the others they are much darker, so it could be a sign of some overheating, and consequent probable failure.
Tomorrow I try to recover the analog tester, changing the batteries, today unfortunately I forgot to buy them.
As soon as I finish the detections, I move on to the removal of the suspicious SMDs, even if first I should recover the spare parts from other MBs, I hope not to be wrong by putting other different ones, I do not know not having written values ​​how to distinguish them (apart from the size), and if there is a difference between the various colors (eg purple and brown).
Unfortunately being a DIY hobbyist, what I know is from experience, the rest is done a lot by the instrumentation, unfortunately I don't have specific testers for each component, nor do I have an air soldering iron, to remove the SMDs I will do a little effort, but I hope not to irreparably ruin the slopes.

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 30 of 137, by weedeewee

User metadata
Rank Oldbie
Rank
Oldbie

Are you measuring the smds on the board ? because if you are, your results will be skewed and unreliable.
from your measurements, I'd say BC35, BC39 and BC40 are shorted, or have a short on the line they are decoupling, if you are measuring these on the board.

edit, I get you're trying to find the short...

edit2, no idea if anyone has suggested it already, but if you know what voltage is normally on that line, you could inject a slightly lower voltage and try to localize which component gets hot/warm.
best to do this with anything that can be removed from the board, removed from the board.

Right to repair is fundamental. You own it, you're allowed to fix it.
How To Ask Questions The Smart Way
Do not ask Why !

Reply 31 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
weedeewee wrote on 2021-11-20, 21:36:
Are you measuring the smds on the board ? because if you are, your results will be skewed and unreliable. from your measurements […]
Show full quote

Are you measuring the smds on the board ? because if you are, your results will be skewed and unreliable.
from your measurements, I'd say BC35, BC39 and BC40 are shorted, or have a short on the line they are decoupling, if you are measuring these on the board.

edit, I get you're trying to find the short...

edit2, no idea if anyone has suggested it already, but if you know what voltage is normally on that line, you could inject a slightly lower voltage and try to localize which component gets hot/warm.
best to do this with anything that can be removed from the board, removed from the board.

Thanks for the info.
Yes, in fact I am trying to understand which SMDs are short, and the measurements I have performed all on-board, the intention is to remove the short ones and see if they actually are.
If after removal they are faulty, I should replace them with recovery ones, and after I should check if anything changes compared to the measurements made so far, in short I have a voltage of only 0.30V on the PIN2 of 1085, while it should be higher than 1.5V , this is what seems most likely.
If you want to see more details and info, you can go to the previous page, in the initial part you will find other MB but further down you will find this 6ABX2V (Slot1).

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 32 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-11-20, 21:25:
I have performed further checks on the suspicious (faulty?) SMDs, to verify them I used the Atlas ESR70, I report the results, b […]
Show full quote

I have performed further checks on the suspicious (faulty?) SMDs, to verify them I used the Atlas ESR70, I report the results, but I anticipate that the capacity has not been displayed, maybe it is normal for them or maybe you need to remove them to be able to view it.
The results are as follows:
BC39 ESR = 0.40OHm
BC35 ESR = 0.08OHm
BC41 ESR = 8.8OHm
BC40 ESR = 0.10OHm

Ah, those are measuring the ESR (apparent resistance at a particular frequency, 100kHz I think), not the actual resistance, so doesn't really help locate the short. I meant to suggest using the ESR70 to measure the capacitance of any replacement capacitors you take from another board. The one that looks off to me is BC41. Ceramic capacitors should have a much lower ESR than that. But that's on pin 1 of the 1085, and as far as we can tell so far there's no short there.

I don't think you can identify which component is shorted whilst it's still in circuit. Weedeewee's suggestion of using a current source to see if anything warms up might work. I think you really need to remove BC35 and/or BC39. Remove one, then test to see if the short has gone. If it hasn't then remove the other.

Unfortunately being a DIY hobbyist, what I know is from experience, the rest is done a lot by the instrumentation, unfortunately I don't have specific testers for each component, nor do I have an air soldering iron, to remove the SMDs I will do a little effort, but I hope not to irreparably ruin the slopes.

It is possible to remove them with a normal soldering iron, if there's space. Put some solder on the tip, then try to get the tip across the top so the solder touches both ends of the component at the same time, then lift using some tweezers. A bit of flux on the component first can help the solder on the iron make contact with the component terminals. Give yourself time, don't rush and don't use force. If you've got a properly broken board with some SMD components then try it out on that first. Flux, a solder sucker and some copper braid can be useful for cleaning up the pads afterward. Avoid dragging the braid in case it snags and pulls a pad.

Reply 33 of 137, by weedeewee

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-11-20, 21:55:
Thanks for the info. Yes, in fact I am trying to understand which SMDs are short, and the measurements I have performed all on-b […]
Show full quote

Thanks for the info.
Yes, in fact I am trying to understand which SMDs are short, and the measurements I have performed all on-board, the intention is to remove the short ones and see if they actually are.
If after removal they are faulty, I should replace them with recovery ones, and after I should check if anything changes compared to the measurements made so far, in short I have a voltage of only 0.30V on the PIN2 of 1085, while it should be higher than 1.5V , this is what seems most likely.
If you want to see more details and info, you can go to the previous page, in the initial part you will find other MB but further down you will find this 6ABX2V (Slot1).

I guess you do not have a bench power supply? or a power supply you can regulate down to 1 or 1.5 volt.
It doesn't need to be high power, though, if the component you're looking for is a really dead short, it might not be detectable since the wires/traces of the board will have a similar resistance. and thus the faulty component will not dissipate more power and thus heat up and be detectable.

Right to repair is fundamental. You own it, you're allowed to fix it.
How To Ask Questions The Smart Way
Do not ask Why !

Reply 34 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-11-20, 22:06:
Ah, those are measuring the ESR (apparent resistance at a particular frequency, 100kHz I think), not the actual resistance, so d […]
Show full quote
PC@LIVE wrote on 2021-11-20, 21:25:
I have performed further checks on the suspicious (faulty?) SMDs, to verify them I used the Atlas ESR70, I report the results, b […]
Show full quote

I have performed further checks on the suspicious (faulty?) SMDs, to verify them I used the Atlas ESR70, I report the results, but I anticipate that the capacity has not been displayed, maybe it is normal for them or maybe you need to remove them to be able to view it.
The results are as follows:
BC39 ESR = 0.40OHm
BC35 ESR = 0.08OHm
BC41 ESR = 8.8OHm
BC40 ESR = 0.10OHm

Ah, those are measuring the ESR (apparent resistance at a particular frequency, 100kHz I think), not the actual resistance, so doesn't really help locate the short. I meant to suggest using the ESR70 to measure the capacitance of any replacement capacitors you take from another board. The one that looks off to me is BC41. Ceramic capacitors should have a much lower ESR than that. But that's on pin 1 of the 1085, and as far as we can tell so far there's no short there.

I don't think you can identify which component is shorted whilst it's still in circuit. Weedeewee's suggestion of using a current source to see if anything warms up might work. I think you really need to remove BC35 and/or BC39. Remove one, then test to see if the short has gone. If it hasn't then remove the other.

Unfortunately being a DIY hobbyist, what I know is from experience, the rest is done a lot by the instrumentation, unfortunately I don't have specific testers for each component, nor do I have an air soldering iron, to remove the SMDs I will do a little effort, but I hope not to irreparably ruin the slopes.

It is possible to remove them with a normal soldering iron, if there's space. Put some solder on the tip, then try to get the tip across the top so the solder touches both ends of the component at the same time, then lift using some tweezers. A bit of flux on the component first can help the solder on the iron make contact with the component terminals. Give yourself time, don't rush and don't use force. If you've got a properly broken board with some SMD components then try it out on that first. Flux, a solder sucker and some copper braid can be useful for cleaning up the pads afterward. Avoid dragging the braid in case it snags and pulls a pad.

Ok thanks, in fact the Atlas I used it only for electrolytic capacitors or other similar (polarized).
Regarding the suggestion to use a power source, to see what heats up, it's something I've never done, but I could see if a selectable voltage power supply can go (I don't remember if it's 1.5V), alternatively maybe you could use one from PC take current from + 3.3V insert a resistor or diodes to reduce the voltage to 1.5V.
Finding the power supply and getting the right voltage should I feel what it heats by touching it with my fingers ??? Or should I use a thermometer, unfortunately among the testers I don't have the function to measure the temperature.
I have removed SMD diodes in the past, but they are not easy to remove, the capacitors should be simpler because smaller, of course I will see how it goes by removing the spare ones first, unfortunately I miss both the flux and the copper braid, I hope they can be remove anyway.

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 35 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
weedeewee wrote on 2021-11-20, 22:16:
PC@LIVE wrote on 2021-11-20, 21:55:
Thanks for the info. Yes, in fact I am trying to understand which SMDs are short, and the measurements I have performed all on-b […]
Show full quote

Thanks for the info.
Yes, in fact I am trying to understand which SMDs are short, and the measurements I have performed all on-board, the intention is to remove the short ones and see if they actually are.
If after removal they are faulty, I should replace them with recovery ones, and after I should check if anything changes compared to the measurements made so far, in short I have a voltage of only 0.30V on the PIN2 of 1085, while it should be higher than 1.5V , this is what seems most likely.
If you want to see more details and info, you can go to the previous page, in the initial part you will find other MB but further down you will find this 6ABX2V (Slot1).

I guess you do not have a bench power supply? or a power supply you can regulate down to 1 or 1.5 volt.
It doesn't need to be high power, though, if the component you're looking for is a really dead short, it might not be detectable since the wires/traces of the board will have a similar resistance. and thus the faulty component will not dissipate more power and thus heat up and be detectable.

Thanks for the helpful tip.
By bench power supply do you mean a switching?
I remember having a power supply with which I powered a powerful CB, this is adjustable, I don't know if it gets so low as voltage, I think it goes up to about 20V more or less, anyway tomorrow I look for it and put a picture, maybe it could be useful adding a circuit that reduces the voltage via a 78XX.

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 36 of 137, by pentiumspeed

User metadata
Rank Oldbie
Rank
Oldbie

Borrow a lab adjustable power supply and thermal camera. Set it up for 1V and 3A or more on the shorted rail and image the rest of board for anything hot with this camera. This is important when you find correct power rail that is shorted.

This what I have set up at work for locating shorted component for the phone repair. I recently fixed a USB storage drive this way. This was a capacitor shorted next to the controller glowed. Not on the 5V or other rails.

Generally, resistance of the short determines how easily seen on thermal imager. Very low ohm resistance tend to trip the over current on the power supply so need to insert 1 ohm 5W resistor in series to get enough watts flowing properly.

Cheers,

Great Northern aka Canada.

Reply 37 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-11-20, 22:48:
Ok thanks, in fact the Atlas I used it only for electrolytic capacitors or other similar (polarized). Regarding the suggestion t […]
Show full quote

Ok thanks, in fact the Atlas I used it only for electrolytic capacitors or other similar (polarized).
Regarding the suggestion to use a power source, to see what heats up, it's something I've never done, but I could see if a selectable voltage power supply can go (I don't remember if it's 1.5V), alternatively maybe you could use one from PC take current from + 3.3V insert a resistor or diodes to reduce the voltage to 1.5V.
Finding the power supply and getting the right voltage should I feel what it heats by touching it with my fingers ??? Or should I use a thermometer, unfortunately among the testers I don't have the function to measure the temperature.
I have removed SMD diodes in the past, but they are not easy to remove, the capacitors should be simpler because smaller, of course I will see how it goes by removing the spare ones first, unfortunately I miss both the flux and the copper braid, I hope they can be remove anyway.

Decent PC PSU should just shut down very quickly when it detects a short, before anything has a chance to heat up. The problem with using resistors to reduce the voltage is, as Weedeewee implied, all the voltage will be dumped across the resistor, so the short itself won't heat up. Diodes might work, but there's still the problem that the short may be low enough resistance that it doesn't heat up any more than the PCB tracks going to it. I've limited knowledge about exactly how ceramic caps fail, but what I've read is that internal cracks can form or be present from when they're made. Without the ceramic the fill the gap, sparks can happen between the metal layers. That can melt the metal, which then flows to fill the void, creating a really good short that can handle fairly large currents. So they can fail with nothing being visibly wrong. But it is very rare that they fail.

I can't think of any way past this without removing at least one of those capacitors (or cutting PCB traces, which sounds like a bad idea). If you get lucky, it'll be the first one you remove. SMD diodes can be trickier because the common body styles mean you can't get an iron to both terminals at the same time. Most passives (resistors, capacitors, inductors) have terminals that completely cover the ends and go right up to the top, so you can heat both ends at the same time with the same iron.

Put a blob of solder on the iron (most electronic solder has a flux core to it, so that'll help), try to get the blob to touch both terminals at the same time, wait a couple of seconds (you may see the solder on the pads melt and flow), then gently try to lift the capacitor.

Reply 38 of 137, by weedeewee

User metadata
Rank Oldbie
Rank
Oldbie

If you're really in a jam and have a few 1.5V or 1.2V batteries laying aroung, you could use those to do short detection.
One thing about this method is the chance of batteries going on fire, starting to leak, generally overheating if left shorted for too long.
Though with a 1.5v battery... the chance, i think, is quite low.
There's only so much power the tiny battery can deliver.

Right to repair is fundamental. You own it, you're allowed to fix it.
How To Ask Questions The Smart Way
Do not ask Why !

Reply 39 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
weedeewee wrote on 2021-11-21, 10:50:
If you're really in a jam and have a few 1.5V or 1.2V batteries laying aroung, you could use those to do short detection. One t […]
Show full quote

If you're really in a jam and have a few 1.5V or 1.2V batteries laying aroung, you could use those to do short detection.
One thing about this method is the chance of batteries going on fire, starting to leak, generally overheating if left shorted for too long.
Though with a 1.5v battery... the chance, i think, is quite low.
There's only so much power the tiny battery can deliver.

That crossed my mind for a moment, but I think the internal resistance of a battery (I think AA cells are around 0.2 ohms) would mean there'd be nearly no voltage across the rest of the circuit, so almost all the power from the battery would just be used to heat up the battery. It's also really not that hard to remove one or two capacitors. It's almost certainly easier than trying to identify in circuit which one might be at fault.

Of course, it may be neither and the problem is somewhere else. What's going to use 1.5V on the board? I know (just read on a datasheet) that the CPU needs it for a termination voltage (slot pins A1, A3, B5, B9). What else? Maybe AGP, although I think this board looks like it's got a 3.3V slot?

Actually, may be worth having a close look at pins A1, A2 and A3 on the CPU slot. Datasheet shows 1&3 at Vtt (1.5V) and 2 is GND. Could be a physical short there if the pins are damaged or there's debris in the slot.