VOGONS


Reply 60 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-11-28, 23:32:
What's your digital meter? You shouldn't get a big difference between the analog and digital meters. […]
Show full quote
PC@LIVE wrote on 2021-11-28, 20:59:

Today I tried to rearrange the ideas about the failure of the 6ABX2V, I tried to understand what works and where the failure is, i.e. in which line, according to the measurements made with three different testers, I concluded that the measurements of the analog meter actually show what is in short, with the digital one there is a less reliable indication of short, for comparison I used the Atlas which measures resistances up to 0.01OHm, but some readings can put the doubt (especially if the trace is long?).

What's your digital meter? You shouldn't get a big difference between the analog and digital meters.

I don't think your Atlas can measure shorts, I think the resistance it shows is the ESR, which is the effective resistance at a high frequency. At a high enough frequency an ideal capacitor should measure as a short, but at DC it should (eventually) measure as open circuit. The problem here seems to be a DC short. If you've got some resistors around would you mind using your analog and digital meters to measure their resistance and make sure that your readings match the value given by the colour bands? That will tell you if the meters are working ok. Then use the meters to recheck the resistance from pin 2 of the regulator to Gnd (use a Gnd pin on the power connector).

Some power lines on a motherboard can be low resistance (my KA7 measures around 5 ohms on the CPU regulator output), so low doesn't necessarily mean there's a problem. At the start of this I think you said that a POST card showed the 3.3V supply as ok. We know from measurements that you took before that the 1.5V regulator has 5V going to it, and that the resistance from the 1.5V line to Gnd was about 0.1ohm (or less). That means the regulator would have to put out about 15A to maintain a 1.5V output, which is much too high. So there's a definite problem somewhere between that regulator and Ground. I would concentrate on finding the cause of that problem before worrying about things that might not be a problem (like the 3.3V supply).

If you're uncomfortable with having to keep removing SMD components until you find the one causing the problem then maybe the best thing is to go back to checking to see if any component is getting hot. The simple way (although not as accurate as a thermal camera) is just applying power to the board, then start putting your finger on various components until you find one that's hot. Hopefully something will heat up enough for you to find it before the 1.5V regulator overheats and the over-temperature protection cuts in. If you can't find anything, then it's back to removing components until you find the problem.

As a digital meter I use a small one (see photo), it is quite full of functions, something is missing, but for that I have another bigger one which is pretty similar.
Unfortunately I have a broken one with the capacitance meter, it has a stupid fault, the ignition switch is broken, I bought a new one but it has contacts on one side backwards, I have looked for one that is the same but there seems to be none made like this, those who sell are not good, I keep it aside because maybe I should modify the circuit to adapt it, and get that tester back working.
But you know I'm not a technician and I don't know how the components work, but I can apply a control technique precise enough to tell me if it's okay or not. I also use a comparison technique, that is by measuring a line (+) I note the measured values ​​OHmx1, if I find strange values ​​there must be a problem, if instead the values ​​are lower there is usually something like mosfet or faulty diode, of usually these are shorted, the capacitors also thought, but now I understand that the resistance increases.
That was why I thought the + 3.3V had a problem, but if you tell me 1.2OHm can be fine, OK.
Yes, the regulator is connected to PIN3 directly at + 5V, in fact the measured value is + 4.97V, and the POST CARD has the + 3.3V led on, a sign that the circuit is working, then I don't know if the PCI slot voltage arrives direct from the + 3.3V (orange wire), for this I will do a check on the PCI pins.
Anyway okay I concentrate on the 1085 circuit, I try to remove the other SMDs and see if I find a faulty one, I hope it is one of those in the vicinity, maybe I also try to power up the MB (reassembling the one I removed first), and see if there is anything hot enough.

Apart from that, I worked in parallel on other MBs, one in particular I thought I could easily retrieve it, it is an ECS P4M890T-M2 with Core 2 Duo 6320, it had a swollen capacitor near the chipset, The ESR was 27.6OHm, I changed it with another one that was the same but of a different brand, the removed one was an OST, I changed it with a Rubycon.
Unfortunately, starting it goes into protection, I think I understand what it is, but for the moment I will keep it aside, I will return to take care of this and other MBs at a later time.

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 61 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member

Today I performed further checks on the 6ABX2V, this time I focus my attention on the 1085 and its 3 PINs, I notice that at each PIN there is at least one capacitor (SMD) and that one part goes to the PIN and the other goes to the ground (black wire ).
Looking for info on operating voltages, I find here a useful info: https://www.edaboard.com/threads/step-by-step … inboard.161775/
In practice it is just like that, PIN3 goes directly to + 5V (red wire), and PIN2 goes to + 3.3V (orange wire), PIN1 I don't know if it should have voltage (or I don't understand), in any case it is so if it is so at PIN2 there should be about 3.3V, while the voltage detected is only 0.30V, and therefore insufficient for operation.
Previously I removed the EC24 electrolytic capacitor, after removal the values ​​are these 103.0uF and ESR = 0.84OHm, all ok then.
I think that even the other SMDs, if removed do not result in short, I would also have the doubt of how to check and replace them, that is, do I measure the OHm in x1 trying to exchange the tips? And what values ​​could I consider normal?
The only thing that does not seem normal to me are the values ​​measured between PIN1 and PIN2, on the diode test I read 022 (also by inverting the leads), but that value is the same when measuring the SMDs between PIN 1 and ground, so this makes me think that the short is between PIN1 and PIN2 of 1085.
To verify that this is the case I try to raise the PIN1, and redo the measurements, if it is short I would measure something like a few OHm.
If even this test does not give results, I would try to remove the SMDs connected to the PIN2, but I hope it is not necessary, we will see if the 1085 is broken, I think I would be close enough to solve, the only thing is that I should wait for the spare to arrive .
I add some detail photos of the area where the EC24 is located, you can see the tracks under the capacitor, and the connections with the 443BX chipset, I hope they will 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 62 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie

The best thing for understanding the 1085 is the datasheet: https://www.analog.com/media/en/technical-doc … ts/108345fh.pdf
Page 2 shows what the 3 pins are: pin 1 is Adjust, pin 2 is Vout, pin 3 is Vin. The output voltage is adjustable, and for the link you gave the output has been set to 3.3V. But it can be set to anything, as long as Vout is at least 1V less than Vin.

Page 12 gives an example circuit and shows how to calculate what the output voltage is: Vout=Vref*(1+R2/R1)+Iadj*R2. Iadj is nearly 0, so can be ignored. On your board R95 (marked 39X) is 24.9ohms and R94 (marked 10A) is 124ohms. Those are equivalent to R2 and R1 in the datasheet, and Vref is 1.25V (also on page 12 of the datasheet). Putting those numbers in then Vout=1.25*(1+24.9/124) = 1.5V. Pin 1 (Adjust) should be 1.25V less then Pin 2 (Vout). So Pin 1 should be about 0.25V. But that will only happen when Pin 2 is at 1.5V. I'd expect the resistance from pin 1 to pin 2 to be around 20 ohms.

When you measure from Pin 1 to Ground and Pin 1 to Pin 2, then if Pin 2 is shorted to Ground, then they are the same thing. So it's not surprising you get the same reading. I don't think you want to use the diode test for that, resistance would be more useful. I'd expect the resistance to be around 20 ohms (24.9 in parallel with 124).

So the Vout pin of that 1085 should be 1.5V. Back at the start of this you removed EC21 and then checked between the + and - pads on the board and found there was a short. If that's actually short (what does your meter say?) then that means there is a short from 1095 pin 2 (Vout) to Ground.

I think we need to distinguish between short and low resistance. If you measure a few ohms, then it's not short. Remember that V=I*R. So I=V/R. So even if you measure about 1.25 ohms between 3.3V and Gnd, then that means the current will be about 2.6Amps. So fine (although it does seem oddly low resistance, but that is another problem, I think). If you had instead measured it as e.g. 0.2ohms, then I=3.3/0.2 = 16.5Amps. Not fine. The buzzer on your digital meter will probably make a sound in both cases. So it tells you if there is a connection, but not necessarily if it's actually a short.

On my (cheap) digital multimeter then if I select the 200ohm range and hold the probes together then it shows the resistance as about 0.8 ohms. That's the resistance of the leads/probe connectors/probe tips/internal meter. If I then use the probes to measure across some component and the meter reads 1.0ohm, then I can subtract the meter's resistance (0.8ohm), which tells me the component has a resistance of 0.2ohm.

The connection you measure between 3.3V and the 1085 pin 2 will be because of the short from pin 2 to ground. 3.3V -> Gnd -> Pin 2.

For testing removed capacitors, put the meter in 200ohm, then measure across the component. After a couple of seconds it should read open circuit. If it's ok then solder it back in. For testing diodes you can use the diode test mode.

I hadn't suggested removing the 1085 most because that might be hard. The main tab is probably soldered to the board and unsoldering those isn't fun, particularly because of how close the Slot connector is.

Let's go back to the start and measure the resistance from 1085 Pin 2 to Ground. Put your digital meter on 200ohm range. Touch the probes together and make a note of what the meter shows. Then measure across the pads of EC21. Subtract the probe resistance, and that will tell you what the resistance from Pin 2 to Ground is.

Reply 63 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-11-30, 21:41:
The best thing for understanding the 1085 is the datasheet: https://www.analog.com/media/en/technical-doc … ts/108345fh.pdf Page […]
Show full quote

The best thing for understanding the 1085 is the datasheet: https://www.analog.com/media/en/technical-doc … ts/108345fh.pdf
Page 2 shows what the 3 pins are: pin 1 is Adjust, pin 2 is Vout, pin 3 is Vin. The output voltage is adjustable, and for the link you gave the output has been set to 3.3V. But it can be set to anything, as long as Vout is at least 1V less than Vin.

Page 12 gives an example circuit and shows how to calculate what the output voltage is: Vout=Vref*(1+R2/R1)+Iadj*R2. Iadj is nearly 0, so can be ignored. On your board R95 (marked 39X) is 24.9ohms and R94 (marked 10A) is 124ohms. Those are equivalent to R2 and R1 in the datasheet, and Vref is 1.25V (also on page 12 of the datasheet). Putting those numbers in then Vout=1.25*(1+24.9/124) = 1.5V. Pin 1 (Adjust) should be 1.25V less then Pin 2 (Vout). So Pin 1 should be about 0.25V. But that will only happen when Pin 2 is at 1.5V. I'd expect the resistance from pin 1 to pin 2 to be around 20 ohms.

When you measure from Pin 1 to Ground and Pin 1 to Pin 2, then if Pin 2 is shorted to Ground, then they are the same thing. So it's not surprising you get the same reading. I don't think you want to use the diode test for that, resistance would be more useful. I'd expect the resistance to be around 20 ohms (24.9 in parallel with 124).

So the Vout pin of that 1085 should be 1.5V. Back at the start of this you removed EC21 and then checked between the + and - pads on the board and found there was a short. If that's actually short (what does your meter say?) then that means there is a short from 1095 pin 2 (Vout) to Ground.

I think we need to distinguish between short and low resistance. If you measure a few ohms, then it's not short. Remember that V=I*R. So I=V/R. So even if you measure about 1.25 ohms between 3.3V and Gnd, then that means the current will be about 2.6Amps. So fine (although it does seem oddly low resistance, but that is another problem, I think). If you had instead measured it as e.g. 0.2ohms, then I=3.3/0.2 = 16.5Amps. Not fine. The buzzer on your digital meter will probably make a sound in both cases. So it tells you if there is a connection, but not necessarily if it's actually a short.

On my (cheap) digital multimeter then if I select the 200ohm range and hold the probes together then it shows the resistance as about 0.8 ohms. That's the resistance of the leads/probe connectors/probe tips/internal meter. If I then use the probes to measure across some component and the meter reads 1.0ohm, then I can subtract the meter's resistance (0.8ohm), which tells me the component has a resistance of 0.2ohm.

The connection you measure between 3.3V and the 1085 pin 2 will be because of the short from pin 2 to ground. 3.3V -> Gnd -> Pin 2.

For testing removed capacitors, put the meter in 200ohm, then measure across the component. After a couple of seconds it should read open circuit. If it's ok then solder it back in. For testing diodes you can use the diode test mode.

I hadn't suggested removing the 1085 most because that might be hard. The main tab is probably soldered to the board and unsoldering those isn't fun, particularly because of how close the Slot connector is.

Let's go back to the start and measure the resistance from 1085 Pin 2 to Ground. Put your digital meter on 200ohm range. Touch the probes together and make a note of what the meter shows. Then measure across the pads of EC21. Subtract the probe resistance, and that will tell you what the resistance from Pin 2 to Ground is.

Ok thank you very much, thank you very much for your patience, but now thanks to your very detailed explanations, I understand what I was wrong.
For the 1085 everything is ok so, now I have to redo the measurements, this time with the tester on 200OHm, I was using the diode test and I was wrong.
Currently on the PIN1 I would have 0,10V, at least this was the value measured the last time I turned on the MB, this value should be 0.25V, however, if the PIN2 is 1.5V, and currently on the PIN2 there is 0.30V, the why it is 0.30V instead of 1.5V is to be found.
I will check the resistance between PIN1 and PIN2, if I remember correctly I had measured 22OHm, which could be less by subtracting the resistance of the probes, but I will double check these values.
Regarding EC21 and the resistance between the holes of the + and -, I have to double check and redo the measurements in OHm (selecting 200), I thought it was short because it sounded, but it may not be.
I will do the checks of the SMD capacitors using the technique you suggested, if they are OK I will go back to them.
I think by tomorrow I have done all the measurements, and I hope to find some anomaly, useful to solve this voltage problem.

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 64 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-11-30, 22:39:

Ok thank you very much, thank you very much for your patience, but now thanks to your very detailed explanations, I understand what I was wrong.
For the 1085 everything is ok so, now I have to redo the measurements, this time with the tester on 200OHm, I was using the diode test and I was wrong.

It's mostly a case of knowing a bit about how to use your tools and what the readings mean. For example, the 200ohm setting just means that the maximum resistance the meter can read in that mode is 200ohm. So if it gives no reading between two point then it doesn't necessarily mean that they aren't connected, just that it might be more than 200ohms. So you can then try increasing the range and see if you get a reading. But the higher the range, the lower the precision, so there's a trade off. When looking for shorts the resistance will be very close to 0, so you want to select the smallest range.

Currently on the PIN1 I would have 0,10V, at least this was the value measured the last time I turned on the MB, this value should be 0.25V, however, if the PIN2 is 1.5V, and currently on the PIN2 there is 0.30V, the why it is 0.30V instead of 1.5V is to be found.

So as Pin 2 (Vout) increases, Pin 1 (Adjust) will also increase, but not as much because of the resistors connecting them together. Vout should keep increasing until the difference between Vout and Adjust is 1.25V. Because of the values of those two resistors (R94 and R95), when Vout reaches 1.5V then Adjust will be at 0.25V, which is 1.25V difference. But there's a limit. As the Vout voltage increases then the current flowing through the regulator will increase, and there's a maximum current that the regulator can handle. If that maximum current is reached then Vout can't go any higher.

You've measured Vout as 0.3V and Adjust as 0.1V. That means that Vout should be increasing, but it isn't. There are at least two possible reasons for that. One is that the regulator is broken. Another possibility (the one I think is most likely) is that it's reached it's maximum current because there's a short/very low resistance connecting Vout to Ground. All the current is going through the short, which means Vout can't go any higher.

I think by tomorrow I have done all the measurements, and I hope to find some anomaly, useful to solve this voltage problem.

Remember that if you do measure a short then it can be difficult to find. We know there are several components that go between Vout and Ground. Any one of them could be the fault and there's no easy way to tell without removing them and then testing to see if the short has gone. There are some pieces of equipment that can find shorts in circuit (Toneohm) by measuring the very small changes in resistance as a probe gets nearer where a short is, but they're quite expensive.

Reply 65 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member

Today I made various measurements, summing up I did not find anything strange, and I think the situation is always the same, I would say that I have not made progress by finding something faulty, but perhaps it could be to look for in areas other than those analyzed.
These are the measurements made, let's start from 1085 and its 3PIN, I measured a resistance of 21.6ohm between PIN1 and PIN2 (22.2-0.6 crossing the leads), keeping in mind that the resistances are R94 = 10A = 124ohm and R95 = 39X = 24.9ohm, for these two it is not possible to measure them individually (they should be removed), so I measure the resistance between the PIN2 of the 1085 and the black wire - (ground), the value found is 1.5ohm (2.0-0.5), I think very similar to the 1.25ohm measured from + 3.3V to black wire - with the analog tester.
I measure the ohms between the holes of the + and - of the EC21 capacitor (always setting to 200ohm), the value found is always 1.5ohm (2.0-0.5), the value found which obviously is the same as that between PIN2 and - (ground) .
I do the check with the BC39 SMD capacitor. Even if the test leads of the tester are reversed to 200ohm, the display always shows the starting digit 1. and this should therefore be, if I understand correctly, normal, that is, working ???
I also measure between the PINs of Slot1 A1 and B1, also here I find a value of 1.5ohm (2.0-0.5), I follow the track from A1 and A3 and I arrive at the resistance R71, it seems written above 220, I don't know how many ohms they are, nor I measure between the two soldering points 14ohm (14.5-0.5), this R71 from the side towards Slot1 goes to the + 3.3V orange wire, obviously the other side goes to PIN A1 and A3 from a track on the back where it is connected the + of the EC20 capacitor, this has not been removed but if needed I could remove it and check it with the Atlas.
In the doubt that the 1085 had some internal problem, I raised the PIN1 leaving the other two connected, but redoing the measurement between PIN1 and PIN2 I find a different value, the value measured before is therefore the total resistance of R95 + R94, and therefore it should work.
In the back I removed a sticker and cleaned the area, I was hoping that by chance it could lead and create problems, but nothing has changed, I always have those values ​​between the PINs of slot1 in A1 and B1, at this point I would not know, if everything is normal I would try to resolder everything I removed, and attempt to boot, I don't think it can start, unless it was the A1 PIN that prevented it from starting (when it was below).

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 66 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-12-01, 12:30:
It's mostly a case of knowing a bit about how to use your tools and what the readings mean. For example, the 200ohm setting jus […]
Show full quote
PC@LIVE wrote on 2021-11-30, 22:39:

Ok thank you very much, thank you very much for your patience, but now thanks to your very detailed explanations, I understand what I was wrong.
For the 1085 everything is ok so, now I have to redo the measurements, this time with the tester on 200OHm, I was using the diode test and I was wrong.

It's mostly a case of knowing a bit about how to use your tools and what the readings mean. For example, the 200ohm setting just means that the maximum resistance the meter can read in that mode is 200ohm. So if it gives no reading between two point then it doesn't necessarily mean that they aren't connected, just that it might be more than 200ohms. So you can then try increasing the range and see if you get a reading. But the higher the range, the lower the precision, so there's a trade off. When looking for shorts the resistance will be very close to 0, so you want to select the smallest range.

Currently on the PIN1 I would have 0,10V, at least this was the value measured the last time I turned on the MB, this value should be 0.25V, however, if the PIN2 is 1.5V, and currently on the PIN2 there is 0.30V, the why it is 0.30V instead of 1.5V is to be found.

So as Pin 2 (Vout) increases, Pin 1 (Adjust) will also increase, but not as much because of the resistors connecting them together. Vout should keep increasing until the difference between Vout and Adjust is 1.25V. Because of the values of those two resistors (R94 and R95), when Vout reaches 1.5V then Adjust will be at 0.25V, which is 1.25V difference. But there's a limit. As the Vout voltage increases then the current flowing through the regulator will increase, and there's a maximum current that the regulator can handle. If that maximum current is reached then Vout can't go any higher.

You've measured Vout as 0.3V and Adjust as 0.1V. That means that Vout should be increasing, but it isn't. There are at least two possible reasons for that. One is that the regulator is broken. Another possibility (the one I think is most likely) is that it's reached it's maximum current because there's a short/very low resistance connecting Vout to Ground. All the current is going through the short, which means Vout can't go any higher.

I think by tomorrow I have done all the measurements, and I hope to find some anomaly, useful to solve this voltage problem.

Remember that if you do measure a short then it can be difficult to find. We know there are several components that go between Vout and Ground. Any one of them could be the fault and there's no easy way to tell without removing them and then testing to see if the short has gone. There are some pieces of equipment that can find shorts in circuit (Toneohm) by measuring the very small changes in resistance as a probe gets nearer where a short is, but they're quite expensive.

Well for the resistors I use the analog tester, not so much for the accuracy, but thanks to that I have often found diode mosfets in short or interrupted, for the 200ohm discourse it is like this but usually when the value is higher it makes a lower reading and then immediately goes to 1. in this way you understand that it goes further, and sometimes you have to go up again until you see a reading.
On the possibility that the 1085 is broken, from the measurements made this morning by lifting the PIN1, I think it is okay, so there remains only the hypothesis of the resistance in short or other, which connects Vout to ground.
Unfortunately I do not have and do not have the possibility to get the tool you suggest, I think I will have to remove and test everything that seems suspicious, hoping to have luck and find the fault, as long as it is in components that can be removed without special equipment

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 67 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-12-01, 19:09:

These are the measurements made, let's start from 1085 and its 3PIN, I measured a resistance of 21.6ohm between PIN1 and PIN2 (22.2-0.6 crossing the leads), keeping in mind that the resistances are R94 = 10A = 124ohm and R95 = 39X = 24.9ohm, for these two it is not possible to measure them individually (they should be removed), so I measure the resistance between the PIN2 of the 1085 and the black wire - (ground), the value found is 1.5ohm (2.0-0.5), I think very similar to the 1.25ohm measured from + 3.3V to black wire - with the analog tester.

Ok, so you've got 1.5 ohms from Vout to Ground. Not quite what I'd think of as a short, but it sounds low.

R94 goes between Pin 1 (adjust) and Pin 2 (Vout). R95 goes between Pin 1 (adjust) and Ground. But Pin 2 is pulled to Ground by the 1.5 ohms you've measured. So that means when you measure from Pin 1 to Pin, current can flow through R94, but it can also flow through R95 to Ground, then from Ground (through the 1.25 ohms) to Pin 2. Effectively they are in parallel with each other. 124 || (24.9+1.5) = (124*26.4)/(124+26.4) = 21.8 ohms. Which is pretty close to the 21.6 you've measured. So it looks like the two resistors are probably ok.

I measure the ohms between the holes of the + and - of the EC21 capacitor (always setting to 200ohm), the value found is always 1.5ohm (2.0-0.5), the value found which obviously is the same as that between PIN2 and - (ground) .

Good, always nice to be able to double check a measurement.

I do the check with the BC39 SMD capacitor. Even if the test leads of the tester are reversed to 200ohm, the display always shows the starting digit 1. and this should therefore be, if I understand correctly, normal, that is, working ???

That sounds right, it sounds like that's what your meter displays when the resistance is too high for that range. A capacitor won't (or shouldn't) pass DC current. To measure a resistance the meter tries to put a small DC current through the probes and then measures the voltage. So with that capacitor the meter can't pass the DC current, so it shows open circuit. So that one's ok. If you do the same test with a bigger capacitor (one of the electrolytics) then you might see the resistance start low then gradually increase as the current gradually charges the capacitor.

I also measure between the PINs of Slot1 A1 and B1, also here I find a value of 1.5ohm (2.0-0.5), I follow the track from A1 and A3 and I arrive at the resistance R71, it seems written above 220, I don't know how many ohms they are, nor I measure between the two soldering points 14ohm (14.5-0.5), this R71 from the side towards Slot1 goes to the + 3.3V orange wire, obviously the other side goes to PIN A1 and A3 from a track on the back where it is connected the + of the EC20 capacitor, this has not been removed but if needed I could remove it and check it with the Atlas.

So I think A1 is on the 1.5V supply, and B1 is connected to Ground on the motherboard, so that 1.5ohm is another confirmation of a very low resistance between Pin 2 and Ground.

Now I think it gets odd. R71 marked 220 is almost certainly a 22ohm resistor (first two numbers are the value, 3rd number is how many zeros, so 22 with zero zeroes -> 22ohms). But A1 and A3 should definitely be 1.5V, not connected to the 3.3V supply. Can you see the trace that goes from A1&3? I can't see it any of the photos.

In the back I removed a sticker and cleaned the area, I was hoping that by chance it could lead and create problems, but nothing has changed, I always have those values ​​between the PINs of slot1 in A1 and B1, at this point I would not know, if everything is normal I would try to resolder everything I removed, and attempt to boot, I don't think it can start, unless it was the A1 PIN that prevented it from starting (when it was below).

It's worth a try. Perhaps 1.5 ohm is ok, even if it feels a bit low. V=IR, so with V=1.5 and R=1.5ohm, then I=1 amp. Which the regulator should be able to drive. We don't really know what the resistance was at the start. Maybe it was much lower.

PC@LIVE wrote on 2021-12-01, 19:27:

Well for the resistors I use the analog tester, not so much for the accuracy, but thanks to that I have often found diode mosfets in short or interrupted, for the 200ohm discourse it is like this but usually when the value is higher it makes a lower reading and then immediately goes to 1. in this way you understand that it goes further, and sometimes you have to go up again until you see a reading.
On the possibility that the 1085 is broken, from the measurements made this morning by lifting the PIN1, I think it is okay, so there remains only the hypothesis of the resistance in short or other, which connects Vout to ground.
Unfortunately I do not have and do not have the possibility to get the tool you suggest, I think I will have to remove and test everything that seems suspicious, hoping to have luck and find the fault, as long as it is in components that can be removed without special equipment

I didn't really mean for you to get one, just that it is technically possible to test things in circuit, but it's only really worth it if you're using it every day for work.

Go with you plan of having a go at putting it back together and seeing if anything changed after you fixed pin A1. Make sure to measure some voltages when you power the board up.

Reply 68 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-12-01, 19:52:
Ok, so you've got 1.5 ohms from Vout to Ground. Not quite what I'd think of as a short, but it sounds low. […]
Show full quote
PC@LIVE wrote on 2021-12-01, 19:09:

These are the measurements made, let's start from 1085 and its 3PIN, I measured a resistance of 21.6ohm between PIN1 and PIN2 (22.2-0.6 crossing the leads), keeping in mind that the resistances are R94 = 10A = 124ohm and R95 = 39X = 24.9ohm, for these two it is not possible to measure them individually (they should be removed), so I measure the resistance between the PIN2 of the 1085 and the black wire - (ground), the value found is 1.5ohm (2.0-0.5), I think very similar to the 1.25ohm measured from + 3.3V to black wire - with the analog tester.

Ok, so you've got 1.5 ohms from Vout to Ground. Not quite what I'd think of as a short, but it sounds low.

R94 goes between Pin 1 (adjust) and Pin 2 (Vout). R95 goes between Pin 1 (adjust) and Ground. But Pin 2 is pulled to Ground by the 1.5 ohms you've measured. So that means when you measure from Pin 1 to Pin, current can flow through R94, but it can also flow through R95 to Ground, then from Ground (through the 1.25 ohms) to Pin 2. Effectively they are in parallel with each other. 124 || (24.9+1.5) = (124*26.4)/(124+26.4) = 21.8 ohms. Which is pretty close to the 21.6 you've measured. So it looks like the two resistors are probably ok.

I measure the ohms between the holes of the + and - of the EC21 capacitor (always setting to 200ohm), the value found is always 1.5ohm (2.0-0.5), the value found which obviously is the same as that between PIN2 and - (ground) .

Good, always nice to be able to double check a measurement.

I do the check with the BC39 SMD capacitor. Even if the test leads of the tester are reversed to 200ohm, the display always shows the starting digit 1. and this should therefore be, if I understand correctly, normal, that is, working ???

That sounds right, it sounds like that's what your meter displays when the resistance is too high for that range. A capacitor won't (or shouldn't) pass DC current. To measure a resistance the meter tries to put a small DC current through the probes and then measures the voltage. So with that capacitor the meter can't pass the DC current, so it shows open circuit. So that one's ok. If you do the same test with a bigger capacitor (one of the electrolytics) then you might see the resistance start low then gradually increase as the current gradually charges the capacitor.

I also measure between the PINs of Slot1 A1 and B1, also here I find a value of 1.5ohm (2.0-0.5), I follow the track from A1 and A3 and I arrive at the resistance R71, it seems written above 220, I don't know how many ohms they are, nor I measure between the two soldering points 14ohm (14.5-0.5), this R71 from the side towards Slot1 goes to the + 3.3V orange wire, obviously the other side goes to PIN A1 and A3 from a track on the back where it is connected the + of the EC20 capacitor, this has not been removed but if needed I could remove it and check it with the Atlas.

So I think A1 is on the 1.5V supply, and B1 is connected to Ground on the motherboard, so that 1.5ohm is another confirmation of a very low resistance between Pin 2 and Ground.

Now I think it gets odd. R71 marked 220 is almost certainly a 22ohm resistor (first two numbers are the value, 3rd number is how many zeros, so 22 with zero zeroes -> 22ohms). But A1 and A3 should definitely be 1.5V, not connected to the 3.3V supply. Can you see the trace that goes from A1&3? I can't see it any of the photos.

In the back I removed a sticker and cleaned the area, I was hoping that by chance it could lead and create problems, but nothing has changed, I always have those values ​​between the PINs of slot1 in A1 and B1, at this point I would not know, if everything is normal I would try to resolder everything I removed, and attempt to boot, I don't think it can start, unless it was the A1 PIN that prevented it from starting (when it was below).

It's worth a try. Perhaps 1.5 ohm is ok, even if it feels a bit low. V=IR, so with V=1.5 and R=1.5ohm, then I=1 amp. Which the regulator should be able to drive. We don't really know what the resistance was at the start. Maybe it was much lower.

PC@LIVE wrote on 2021-12-01, 19:27:

Well for the resistors I use the analog tester, not so much for the accuracy, but thanks to that I have often found diode mosfets in short or interrupted, for the 200ohm discourse it is like this but usually when the value is higher it makes a lower reading and then immediately goes to 1. in this way you understand that it goes further, and sometimes you have to go up again until you see a reading.
On the possibility that the 1085 is broken, from the measurements made this morning by lifting the PIN1, I think it is okay, so there remains only the hypothesis of the resistance in short or other, which connects Vout to ground.
Unfortunately I do not have and do not have the possibility to get the tool you suggest, I think I will have to remove and test everything that seems suspicious, hoping to have luck and find the fault, as long as it is in components that can be removed without special equipment

I didn't really mean for you to get one, just that it is technically possible to test things in circuit, but it's only really worth it if you're using it every day for work.

Go with you plan of having a go at putting it back together and seeing if anything changed after you fixed pin A1. Make sure to measure some voltages when you power the board up.

Yes, those 1.5ohm from the beginning seemed a bit low to me, with the analog tester I read 1.2 or 1.25ohm, in fact it will not be a short but the value that should be normal, would be about 3ohm or a little more. So for the R94 and R95 resistors I avoid removing them, near that point there would be BC35 and BC41, these would have to be tested and removed both? If so, I would try to remove them after the boot test, at least I avoid unplugging them unnecessarily in case the MB starts up.
Regarding the BC39 capacitor, then I trace it and then I trace everything including the electrolytic and the PIN1 of 1085, after I try to start and take note of the voltages, if I can see to measure also those of slot1 in the + 1.5V PINs.
The trace of slot1 that connects A1 and A3 is just below the slot, I can see it by eye, however I take a photo of the detail, it continues under the slot and comes out before EC20 (you can see because it is much thicker than the others), then it passes in the back and continues towards the chipset, then it comes back up and goes to pin + of EC20, then it passes over again and exits near the chipset and arrives at R71.
For the tool thanks for the tip, I didn't want to buy one, but if I had several MB to fix a day, I would think about it and look for one, it would certainly be very useful, in this regard after I got the Atlas, I thought about getting one another always from Atlas (I don't remember the model), it is used to check transistors and the like and has three probes.

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 69 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-12-01, 21:09:

The trace of slot1 that connects A1 and A3 is just below the slot, I can see it by eye, however I take a photo of the detail, it continues under the slot and comes out before EC20 (you can see because it is much thicker than the others), then it passes in the back and continues towards the chipset, then it comes back up and goes to pin + of EC20, then it passes over again and exits near the chipset and arrives at R71.

Thanks for the photo, I added it to the composite I've got. As you found, I can see where it goes from A3 to EC20, then heads toward the 443BX, but then it goes under the label that I think you've now removed. I'd expect it to carry on and come out on the right of the 443, and join up with the trace from EC24. I'm a bit surprised if it heads up to R71 as there's nothing up there that needs 1.5V.

Attachments

  • 6ABX2V_1V5_Trace.jpg
    Filename
    6ABX2V_1V5_Trace.jpg
    File size
    304.77 KiB
    Views
    188 views
    File license
    Fair use/fair dealing exception

Reply 70 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-12-01, 22:06:
PC@LIVE wrote on 2021-12-01, 21:09:

The trace of slot1 that connects A1 and A3 is just below the slot, I can see it by eye, however I take a photo of the detail, it continues under the slot and comes out before EC20 (you can see because it is much thicker than the others), then it passes in the back and continues towards the chipset, then it comes back up and goes to pin + of EC20, then it passes over again and exits near the chipset and arrives at R71.

Thanks for the photo, I added it to the composite I've got. As you found, I can see where it goes from A3 to EC20, then heads toward the 443BX, but then it goes under the label that I think you've now removed. I'd expect it to carry on and come out on the right of the 443, and join up with the trace from EC24. I'm a bit surprised if it heads up to R71 as there's nothing up there that needs 1.5V.

Yes ok thanks, but in fact double-checking better I must have made a mistake, I think it comes out near R77, at the moment I don't know exactly where, I'll try to check better and establish the exact point, but I can do it tomorrow.
So R71 I think it is a coincidence that it is connected to + 3.3V, and I don't know what it connects to.
The label behind the chipset has been removed, I had put a photo before, the trace is visible for the thickness and for the proximity to the slot1, nearby there is another one that is the same thick but I don't know what it connects.

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 71 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member

I found this info, I hope it will be useful, I copy this part:
Over-current protection Circuit With AIC 1569CS AH13 (protect CPU from accident short circuit), and a set poly-fuses (protect Keyboard and USB devices from accident short circuit), if system current is over-loaded, the system will break down to protect direct damage of the main board. Special features! Jumperless design! Wake on lan (WOL)! Modem ring on! Auto detection of CPU voltage, fan & temperature! Creative PCI sound Blaster AWE64D header ( Support PC/PCI spec.)! Windows 95/98 power off! Keyboard wake up! PS/2 mouse wake up

The rest can be found here:
https://docplayer.net/62457218-6abx2v-6azx2v- … r-s-manual.html

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 72 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-12-01, 22:20:

The label behind the chipset has been removed, I had put a photo before, the trace is visible for the thickness and for the proximity to the slot1, nearby there is another one that is the same thick but I don't know what it connects.

Oops, I missed that. Thanks. So I think we've got most of the 1.5V line now.

Starting from A3 on the Slot it connects to EC20, heads under the 443, then branches, with one branch going up toward R77. The other branch goes up through a via and probably heads SE to join the via back down to join the trace coming in from the right. That has a via that comes up under the balls of the 443 so probably connects to it there.

It carries on and looks to connect to R106 and BC44, then on to EC24, and it looks like BC42 and BC43. From there is goes toward the 1085 regulator, connecting to BC35, R94 and pin 2 of the 1085. Finally it carries on to BC39 and EC21.

So, EC20, R77, R106, BC44, EC24, BC42, BC43, BC35, R94, Pin 2 (and tab) of 1085, BC39 and EC21. Plus the 443BX and the Slot 1 connector.

That looks like most of it. The only major bit not found is the trace coming from B5 & B9 that should also connect to the 1.5 trace. I wouldn't be surprised it it connected to a SMD capacitor above the Slot.

There's also R78, C78 (North of the 443BX) and R105, C97 (East of the 443BX), which I think connect to the 1.5V through resistors R77 and R106. They provide a reference voltage for some of the Pentium signals. I looked up the 443BX pinout and found E16 & M23 - GTLRef and F17 & M24 - Vtt. Vtt is the 1.5V line, GTLRef is a reference voltage, which I expect will be Vtt/2.

Still, go with seeing if fixing Pin A1 has fixed the problem. If it hasn't, then start working through the list of components. I've filled in the drawing a bit more, including the GTLRef lines in yellow.

[edit: GTL ref is actually Vtt*2/3. R106 is 75 ohms (marked 750), R105 is 150 ohms (marked 18A, using EIA-96 marking). So the yellow lines should be at 1V when the 1.5V supply is working.]

Attachments

  • 6ABX2V_1V5_GTLRef.jpg
    Filename
    6ABX2V_1V5_GTLRef.jpg
    File size
    285.3 KiB
    Views
    166 views
    File license
    Fair use/fair dealing exception

Reply 73 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-12-02, 00:09:
Oops, I missed that. Thanks. So I think we've got most of the 1.5V line now. […]
Show full quote
PC@LIVE wrote on 2021-12-01, 22:20:

The label behind the chipset has been removed, I had put a photo before, the trace is visible for the thickness and for the proximity to the slot1, nearby there is another one that is the same thick but I don't know what it connects.

Oops, I missed that. Thanks. So I think we've got most of the 1.5V line now.

Starting from A3 on the Slot it connects to EC20, heads under the 443, then branches, with one branch going up toward R77. The other branch goes up through a via and probably heads SE to join the via back down to join the trace coming in from the right. That has a via that comes up under the balls of the 443 so probably connects to it there.

It carries on and looks to connect to R106 and BC44, then on to EC24, and it looks like BC42 and BC43. From there is goes toward the 1085 regulator, connecting to BC35, R94 and pin 2 of the 1085. Finally it carries on to BC39 and EC21.

So, EC20, R77, R106, BC44, EC24, BC42, BC43, BC35, R94, Pin 2 (and tab) of 1085, BC39 and EC21. Plus the 443BX and the Slot 1 connector.

That looks like most of it. The only major bit not found is the trace coming from B5 & B9 that should also connect to the 1.5 trace. I wouldn't be surprised it it connected to a SMD capacitor above the Slot.

There's also R78, C78 (North of the 443BX) and R105, C97 (East of the 443BX), which I think connect to the 1.5V through resistors R77 and R106. They provide a reference voltage for some of the Pentium signals. I looked up the 443BX pinout and found E16 & M23 - GTLRef and F17 & M24 - Vtt. Vtt is the 1.5V line, GTLRef is a reference voltage, which I expect will be Vtt/2.

Still, go with seeing if fixing Pin A1 has fixed the problem. If it hasn't, then start working through the list of components. I've filled in the drawing a bit more, including the GTLRef lines in yellow.

[edit: GTL ref is actually Vtt*2/3. R106 is 75 ohms (marked 750), R105 is 150 ohms (marked 18A, using EIA-96 marking). So the yellow lines should be at 1V when the 1.5V supply is working.]

Today with the 6ABX2V I made a good progress, and immediately after a big regress.
I reassembled everything I had taken apart, electrolytic capacitors and the SMD, the SMD has a lot of solder but still does its job, apart from that I prepare everything to try to start the MB, with Pentium II 350 and 64MB RAM, I put a new battery, I connect the power supply and without keyboard and VGA I connect the pins to start.
The card starts up, but the POST CARD stays on - -, this does not mean that the card is not working, but that it could work by replacing either the RAM or the CPU, this happened in the ASRock S.462 that I had fixed lately.
I check with the card on the voltages, and something has mostly changed for the better, checking the 1085, finally the voltages are exactly (enough) as they were thought to be, practically like this:
PIN1 0.27V | PIN2 1.50V | PIN3 4.92V
i also measure Q3 = CEB6030AL and Q5 = CEB703AL here:
Q3 PIN1 4.75V | PIN2 4.83V | PIN3 1.75V
Q5 PIN1 5.78V | PIN2 1.75V | PIN3 0.02V
in Q5 the PIN1 rises to 5.78V, and this I think is normal, unfortunately the two pins connected together are only at 1.75V, while they should be about 2V more or less, this could be the voltage of the VCore, and therefore being a bit low may not start the PC.
At this point I think of changing the CPU to see if it starts, but in inserting only with the weight of the CPU, the PIN A1 that I had fixed breaks. What a mess!!!
Now there are two alternatives, either replace the entire SLOT1, or try to find a solution to make it fit enough to make contact and allow the CPU to be inserted.
For the moment I'm thinking what to do to fix it, unfortunately since the space is limited, I can't weld in the point where it broke, so, despite myself, I have to suspend work on this MB, waiting for a solution that solves this problem.

Having more MBs under repair I switch to a Siemens D1107 MB, with BX chipset, it has a couple of capacitors to replace and I had set up a track on the back (making a bridge), basically I remember it wouldn't start, so I reconnect two more electrolytic capacitors of recovery (SANYO 3300uF 6.3V), and recover a RAM of a 128MB Siemens PC, I put the PII 350, the new battery, I connect the power supply, without keyboard and VGA.
I'll tell the rest in a while.

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 74 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member

I have several BXs awaiting repair, I pull out a socket about 10 years ago or so, it's a Siemens D1107 ATX, BX chipset (without any sayings), has a couple of capacitors to change and others a little dented on top. , so I rummage through the recovery capacitors and I find two SANYO 3300uF 6.3V, visually they are the same and I weld them in place of the removed ones.
I prepare the card for start-up, without connecting the keyboard and VGA, boot and only one EF code is displayed on the POST card - at this point I go to check the voltages of the mosfets, and on the 1085 I inadvertently make contact between PIN2 and PIN3! !! The power supply is turned off !!! Normally some not-so-cute words about dogs and pigs would follow
Oh well now it went like this. I try to restart and it starts, then the POST CARD codes begin to scroll up to some beeps, I must have accidentally made a defibrillation (chargeaa freeaa), the card is back to life.
Connected keyboard and vga, an AGP fastware Hurricane with Trident 3D9750 chip and 4MB of RAM, on the screen I have the various screens of the BIOS and a couple of errors that can be corrected by the BIOS, corrected the errors I connect an adapter with a 256MB CF and DOS to the PC off (which I use on older PCs).
Everything ok made a couple of benches and tests, now I need to add some cards, because the MB does not have integrated audio and LAN, for the LAN I use a PCI with realtek chip, for the audio I have not decided what to use , if I wanted I could also put an ISA, but I have few left, since I have resurrected a few old PCs.
Later I realized that the original capacitors were 1500uF and therefore I mistakenly put two more capable capacitors (over double), I hope I don't have to remove them and they can be fine anyway.

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 75 of 137, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
PC@LIVE wrote on 2021-12-02, 19:32:

I check with the card on the voltages, and something has mostly changed for the better, checking the 1085, finally the voltages are exactly (enough) as they were thought to be, practically like this:
PIN1 0.27V | PIN2 1.50V | PIN3 4.92V

Yay!

i also measure Q3 = CEB6030AL and Q5 = CEB703AL here: Q3 PIN1 4.75V | PIN2 4.83V | PIN3 1.75V Q5 PIN1 5.78V | PIN2 1.75V | PIN3 […]
Show full quote

i also measure Q3 = CEB6030AL and Q5 = CEB703AL here:
Q3 PIN1 4.75V | PIN2 4.83V | PIN3 1.75V
Q5 PIN1 5.78V | PIN2 1.75V | PIN3 0.02V
in Q5 the PIN1 rises to 5.78V, and this I think is normal, unfortunately the two pins connected together are only at 1.75V, while they should be about 2V more or less, this could be the voltage of the VCore, and therefore being a bit low may not start the PC.

Hmm. Using a digital meter on a switching signal can give odd results. The Gate input (Pin 1) is quite a fast switching signal, so you can probably ignore those. Plus they may use 12V as the switching signal. 1.75V output might mean a problem with the VID signals from the CPU which tell the motherboard what voltage it should provide. You measured them as 2V before, so they were working.
VID pins are all at the other end of the Slot: VID[4:0] on A121, B119, A119, A120, B120.

At this point I think of changing the CPU to see if it starts, but in inserting only with the weight of the CPU, the PIN A1 that I had fixed breaks. What a mess!!!

Boo!

Now, that Pin A1 is only one of 4 pins that connect the 1.5V supply to the CPU. If they're connected on the CPU, then missing one pin probably won't matter.

Obviously at some point you have to decide whether this board goes in the parts pile, but if you wanted to carry on with it then I'd have a careful look at the pins at the other end of the slot, spray a bit of contact cleaner/IPA in the slot and try to give them a clean. Clean the Pentium edge connector. Then give it another go. If there's no change then start to work out where those VID lines go so you can make some measurements. My guess is that they end up connecting to some resistors near the AIC 1569CS that I think controls the switching, but I can't find a datasheet for it.

Reply 76 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
snufkin wrote on 2021-12-02, 22:02:
Yay! […]
Show full quote
PC@LIVE wrote on 2021-12-02, 19:32:

I check with the card on the voltages, and something has mostly changed for the better, checking the 1085, finally the voltages are exactly (enough) as they were thought to be, practically like this:
PIN1 0.27V | PIN2 1.50V | PIN3 4.92V

Yay!

i also measure Q3 = CEB6030AL and Q5 = CEB703AL here: Q3 PIN1 4.75V | PIN2 4.83V | PIN3 1.75V Q5 PIN1 5.78V | PIN2 1.75V | PIN3 […]
Show full quote

i also measure Q3 = CEB6030AL and Q5 = CEB703AL here:
Q3 PIN1 4.75V | PIN2 4.83V | PIN3 1.75V
Q5 PIN1 5.78V | PIN2 1.75V | PIN3 0.02V
in Q5 the PIN1 rises to 5.78V, and this I think is normal, unfortunately the two pins connected together are only at 1.75V, while they should be about 2V more or less, this could be the voltage of the VCore, and therefore being a bit low may not start the PC.

Hmm. Using a digital meter on a switching signal can give odd results. The Gate input (Pin 1) is quite a fast switching signal, so you can probably ignore those. Plus they may use 12V as the switching signal. 1.75V output might mean a problem with the VID signals from the CPU which tell the motherboard what voltage it should provide. You measured them as 2V before, so they were working.
VID pins are all at the other end of the Slot: VID[4:0] on A121, B119, A119, A120, B120.

At this point I think of changing the CPU to see if it starts, but in inserting only with the weight of the CPU, the PIN A1 that I had fixed breaks. What a mess!!!

Boo!

Now, that Pin A1 is only one of 4 pins that connect the 1.5V supply to the CPU. If they're connected on the CPU, then missing one pin probably won't matter.

Obviously at some point you have to decide whether this board goes in the parts pile, but if you wanted to carry on with it then I'd have a careful look at the pins at the other end of the slot, spray a bit of contact cleaner/IPA in the slot and try to give them a clean. Clean the Pentium edge connector. Then give it another go. If there's no change then start to work out where those VID lines go so you can make some measurements. My guess is that they end up connecting to some resistors near the AIC 1569CS that I think controls the switching, but I can't find a datasheet for it.

So the work continues on another BX, this time an ASUS P2B-F, the voltages all seem fine but despite this it does not start, the post card always shows - -, sometimes turning off and on it shows a code or two , but it is not starting at least for now.

With the 6ABX2V I did not give up, I tried to put a 450 MHz Pentium CPU (II or III?) And start, the PIN A1 if it were not there would not be a problem, because both the MB and the CPU have connections ( slopes) with the A3 PIN and perhaps 2 other B PINs?
The connection is visible on the back of the ASUS, the tracks are in the back, on the Lucky Star they are inside.
Starting I saw that the voltages of 1085 were back down, also it gets very hot, a sign that there is a short with the PIN B1 (I guess?), So I remove it and try again to push the pin A1 in with the screwdriver, when I can to let it enter the block with the usual resistance, and at that point I insert the CPU, remove the block of the resistance and reconnect the power supply.
I start the MB, and the voltages return more or less normal, I detect the voltages on the Q6 Q3 Q5, some go up a little while others go down a little, I measure on L10 (Coil around the ring) 1.97V, if this is the VCore it should be right.
So I think it's okay, at least for the voltages, but the failure to boot could be due to something else, not having a suitable programmer for the BIOS chip I would have to do a hot flash, an operation that I have never done and that is not really easy, but at least it could remove the doubt if the BIOS were by chance faulty.
However, there is still a defect (or failure?), Pressing the switch starts up, holding it down does not turn off, also I saw that the reset does not work, pressing it does not restart, maybe it could be this last problem that prevents the start?

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 77 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member

I come back for some reason, trying to start the P2B-F, sometimes it seems that it is ready to go, some POST codes are displayed in sequence, but it stops almost immediately, two or three at most pass (by eye it seems like this).
The BIOS is AWARD. The codes displayed are various, and mostly have no description in the POST CARD manual, I will report some if by chance they are useful.
00 07 | 00 05 | 00 A7 | 00 27 | 00 A1 | 00 87 | 80 A0 | 00 20 | 00 25 | 00 AF | etc...
It would seem that it is about to start, I thought about changing various RAM, but there is no difference, as a CPU I use a PII 300 that on another card works, so in this one it should go too, VGA I use an AGP with ATi rage pro chip turbo, all things that normally work in other PCs, but I don't know the reason for the failed start, I'll try the CMOS reset, but I don't think it can suddenly make the MB work again.
All in all it is a nice card, I have two other ASUS subsequent to this one (P3B-F) working, and the Siemens that I recovered days ago is very robust (in my opinion), this ASUS instead has control (or protection) chips, that maybe read strange values ​​and block the start, for example I do not think that the lack of the CPU temperature sensor is read as a problem and therefore does not start.

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 78 of 137, by weedeewee

User metadata
Rank Oldbie
Rank
Oldbie

FYI, the codes you see, the first two digits are one code, the second two digits are the previous code.
that's how that post card shows the port 80 post codes. current / previous.

Do you get the same codes on the post card when inserted into the pci slot ?

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 79 of 137, by PC@LIVE

User metadata
Rank Member
Rank
Member
weedeewee wrote on 2021-12-04, 13:21:

FYI, the codes you see, the first two digits are one code, the second two digits are the previous code.
that's how that post card shows the port 80 post codes. current / previous.

Do you get the same codes on the post card when inserted into the pci slot ?

Yes ok thank you very much, I knew what the two displays indicated, but I don't understand if they indicate something that does not work on the MB, because in the booklet in many of those codes there is nothing written.
Honestly I have not tried on the PCI slot, for convenience I use the ISA one, because it is better visible, the PCI one the card goes to the opposite side, and you have to go and see if it shows something.
Among other things, I noticed that a couple of LEDs change on the PCI (on and off), I hope it is normal, the + 3.3V for example turns off.

In any case I try to make some attempts with the POST CARD in the PCI, I'll see if it goes the same, but I don't think it changes much.

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