biessea wrote on 2022-09-06, 20:07:

Anyway in the inage you ask me yo measure pin 11 on u10 to pin 15 on simm sockets. (not pin 10 to pin 15). Pin 10 on the chip u10 is GROUND.

Probably you made a little mistake on the scheme you gave me before.

Yes, I'm sorry, I was off by one. The correct pin numbers would have been 11 to 18, not 10 to 17.

biessea wrote on 2022-09-06, 20:07:

Anyway measurement are successful, I read about 33ohm (only in the pin 4 I can read about 34ohm but I think is correct so.)

Anything between 32 and 38 is acceptable, so the traces are fine.

biessea wrote on 2022-09-06, 20:07:

The strange thing is that I can confirm measurement only in the first FOUR Simm socket near the resistances, the four socket upper give me 0.0.

Is four simm socket gone?

Doesn't have to be so. The chip named U10 is used as amplifier to send the address signal to the SIMMs. Possibly the upper four sockets use a different amplifier. Consider that a single SIMM might have 9 chips, all connected to the address lines. To operate four SIMMs, you thus need to drive the address signal into 36 chips. That's a lot of (electrical) work to do. It makes a lot of sense to use two amplifier chips, to distribute the load when 72 chips are installed (36 in bank 0 and 36 in bank 1). I don't think you have to check address traces for the second bank. I told you to test the address traces for bank 0, because U10 is so close to the battery, so traces connected to it might have taken damage.

biessea wrote on 2022-09-06, 20:12:

I am sorry, but for the diode test, where I put the black lead you said me on 5v?? Where is this 5v pin?

+5V is the positive power supply pin for most chips on your board. You can find it at pin 20 of U10 for example.

biessea wrote on 2022-09-06, 20:40:

Infact, when I populate the four lower slot system arrive to bip-bip-bip-bip-bip-bip-bip... Etcetera after about thirty seconds from my power on.
When I populate the four upper simm slot the pc speaker soon gave me Three low tone beep (ram problem). Probably the upper slot are gone, or the chip u10 is fault with addressing the upper four simm slot. Am I right?

No, you are not right. In these old boards, you must fill bank 0 with memory before you fill bank 1. And if you use differently sized modules, you usually have to put the smaller modules into bank 0 and the bigger modules into bank 1. As the allowed memory configurations were usually listed in a table in a mainboard, the feature of modern chipset to allow any kind of module to be installed into any slot is called "table-free memory configuration". Your board is not table-free, and the behaviour you observed is expected.

biessea wrote on 2022-09-06, 20:40:

Anyway I put the VGA on the ISA slot and I have no image at all. I tried to switch the video card in ALL ISA slot. Always the same result. No video signal.

This confirms my suspicion that something is wrong on the board that prevents the computer from getting far enough into the boot process to initialize the VGA card. I couldn't be sure, though.

Hey, I take a look at the jumper configuration table, and I saw that cache jumper was wrong based in my configuration of 64k of cache. It was all wrong. I followed the table and put the jumper correctly now. The NPU configuration was set bad too... Like someone messed with the jumpers in past.

The NPU has to be set SYNCHRONOUS or ASYNCRONOUS or DISABLED?

After checked the pins in the jumper, tried removing the NPU and setting the NPU pins to disable, the error is the same.

PC speaker beeps short forever until I shut down the power supply.

Now I have to give up? Something in the board is wrong and the PC speaker tells me that?

Please dont' tell me I have to give up, I really really wanted this board to work.

Another question:

Jp1 (the jumper for external battery) I leave opened and connected the positive pole of the battery to pin 1. It's correct?

Cause I found in a table jumper configuration that they write that jp1 has to be closed for external battery.

Some people told me to take the jumper off and leave all open and power the first pin through a normal cr2032.

mkarcher wrote on 2022-09-06, 20:52:

No, you are not right. In these old boards, you must fill bank 0 with memory before you fill bank 1. And if you use differently […]
Show full quote

biessea wrote on 2022-09-06, 20:40:

Infact, when I populate the four lower slot system arrive to bip-bip-bip-bip-bip-bip-bip... Etcetera after about thirty seconds from my power on.
When I populate the four upper simm slot the pc speaker soon gave me Three low tone beep (ram problem). Probably the upper slot are gone, or the chip u10 is fault with addressing the upper four simm slot. Am I right?

No, you are not right. In these old boards, you must fill bank 0 with memory before you fill bank 1. And if you use differently sized modules, you usually have to put the smaller modules into bank 0 and the bigger modules into bank 1. As the allowed memory configurations were usually listed in a table in a mainboard, the feature of modern chipset to allow any kind of module to be installed into any slot is called "table-free memory configuration". Your board is not table-free, and the behaviour you observed is expected.

biessea wrote on 2022-09-06, 20:40:

Anyway I put the VGA on the ISA slot and I have no image at all. I tried to switch the video card in ALL ISA slot. Always the same result. No video signal.

This confirms my suspicion that something is wrong on the board that prevents the computer from getting far enough into the boot process to initialize the VGA card. I couldn't be sure, though.

Ehy my dear, you abandoned me?

Cmon there are something I can do more isn't it?

I didn't make the diode test at the moment, I am away from home four days I will do on sunday without doubts.

biessea wrote on 2022-09-06, 21:40:

The NPU has to be set SYNCHRONOUS or ASYNCRONOUS or DISABLED?

It is a good plan to first try without the NPU at all. For that case, "disabled" is correct. If the NPU and the CPU should operate at the same clock (40MHz for your board), "synchronous" is the right choice.

biessea wrote on 2022-09-06, 21:53:

PC speaker beeps short forever until I shut down the power supply.

This can happen if the keyboard or keyboard controller misbehaves badly. Did you ever try with the keyboard disconnected? You should get the dreaded "keyboard missing - press F1 to continue" message without a keyboard, no repeated beeps. Please do the following conductivity checks for the keyboard interface if disconnecting the keyboard doesn't help: The pin of the keyboard jack closest to the battery must be connected to pin 1 of the keyboard controller (U17). The pin in the center of the keyboard jack must be connected to pin 39 of the keyboard controller. Both of these pins of the keyboard jack must not be directly connected to ground (for example at pin 20 of the keyboard controller or pin 10 of U10) or +5V (for example at pin 40 of the keyboard connector or pin 20 of U10). Try carefully pulling the keyboard controller out of its socket, and reinsert it. Possibly you have a contact issue there.

biessea wrote on 2022-09-06, 22:05:

Jp1 (the jumper for external battery) I leave opened and connected the positive pole of the battery to pin 1. It's correct?

Cause I found in a table jumper configuration that they write that jp1 has to be closed for external battery.

Some people told me to take the jumper off and leave all open and power the first pin through a normal cr2032.

The idea is that the external battery is connected to a 4-pin plug you plug into J1. the two central pins are not connected to anything, and the two outer pins are connected to the positive and the negative pole of the battery. So "having something plugged into J1" is correct for use of an external battery, but calling it "closed" is misleading. Your way of connecting the battery and removing the jumper from J1 is correct. Connecting the two inner pins of J1 enables the on-board battery.

mkarcher wrote on 2022-09-07, 18:53:

This can happen if the keyboard or keyboard controller misbehaves badly. Did you ever try with the keyboard disconnected? You should get the dreaded "keyboard missing - press F1 to continue" message without a keyboard, no repeated beeps. Please do the following conductivity checks for the keyboard interface if disconnecting the keyboard doesn't help: The pin of the keyboard jack closest to the battery must be connected to pin 1 of the keyboard controller (U17). The pin in the center of the keyboard jack must be connected to pin 39 of the keyboard controller. Both of these pins of the keyboard jack must not be directly connected to ground (for example at pin 20 of the keyboard controller or pin 10 of U10) or +5V (for example at pin 40 of the keyboard connector or pin 20 of U10). Try carefully pulling the keyboard controller out of its socket, and reinsert it. Possibly you have a contact issue there.

Perfect, as soon as I return home I will do the try. And the diode test too. I has always the hope to find the problem and make this motherboard power on regularly.

mkarcher wrote on 2022-09-07, 18:56:

The idea is that the external battery is connected to a 4-pin plug you plug into J1. the two central pins are not connected to anything, and the two outer pins are connected to the positive and the negative pole of the battery. So "having something plugged into J1" is correct for use of an external battery, but calling it "closed" is misleading. Your way of connecting the battery and removing the jumper from J1 is correct. Connecting the two inner pins of J1 enables the on-board battery.

Ok, so this is all ok with this jumper.

You explain me perfectly, I understand.

But tell me, why I can't connect the ground of the cr2032 to the pin 4? You told me at the beginning to connect the cr2032 to the ground of the previous barrel battery.

I'm just curious.

biessea wrote on 2022-09-07, 21:19:

But tell me, why I can't connect the ground of the cr2032 to the pin 4? You told me at the beginning to connect the cr2032 to the ground of the previous barrel battery.

You can connect the ground of the CR2032 to pin 4 of J1. That pin is directly connected to ground, just as the ground of the previous barrel battery is. Either way works, so no need to change anything you have done.

I didn't check the polarity of the J1, so I couldn't tell you which pin of J1 should be the positive side of the battery and which pin of J1 is connected to ground. I tried to tell you to check which pin of J1 is conncted to the ground pin of the barrel battery. You would have found that pin 4 is connected, and I expected that you use pin 4 after finding out that pin 4 is meant for ground.

mkarcher wrote on 2022-09-07, 18:53:

It is a good plan to first try without the NPU at all. For that case, "disabled" is correct. If the NPU and the CPU should opera […]
Show full quote

biessea wrote on 2022-09-06, 21:40:

The NPU has to be set SYNCHRONOUS or ASYNCRONOUS or DISABLED?

It is a good plan to first try without the NPU at all. For that case, "disabled" is correct. If the NPU and the CPU should operate at the same clock (40MHz for your board), "synchronous" is the right choice.

biessea wrote on 2022-09-06, 21:53:

PC speaker beeps short forever until I shut down the power supply.

This can happen if the keyboard or keyboard controller misbehaves badly. Did you ever try with the keyboard disconnected? You should get the dreaded "keyboard missing - press F1 to continue" message without a keyboard, no repeated beeps. Please do the following conductivity checks for the keyboard interface if disconnecting the keyboard doesn't help: The pin of the keyboard jack closest to the battery must be connected to pin 1 of the keyboard controller (U17). The pin in the center of the keyboard jack must be connected to pin 39 of the keyboard controller. Both of these pins of the keyboard jack must not be directly connected to ground (for example at pin 20 of the keyboard controller or pin 10 of U10) or +5V (for example at pin 40 of the keyboard connector or pin 20 of U10). Try carefully pulling the keyboard controller out of its socket, and reinsert it. Possibly you have a contact issue there.

Here I am.

I tested the keyboard controller, an all is fine, no short circuits. The pin of the keyboard jack closest to the battery is correctly connected to the keyboard controller pin 1 and the pin in the center is correctly connected to the pin 39. They are NOT connected to the ground or to the +5V, so it's ok. I pulled out the keyboard controller and I check the contacts and I put a contact spray revive then I put the keyboard controller inside its socket.

Anyway, tried don't connect the keyboard connector to the motherboard and the same situation occurs. After about 30 seconds the speaker short beep until I power of. So it seems that there is no difference if I put the keyobard connector or not.

Then I tried to have the chock to the diode, and here something isn't correct as you explained. The black pole of the multimeter put on the +5V and the red put on the SIMM pins that you have me told before, my multimeter for the pin 15 and 14 result in a ZERO (0) no value like the other pins like 12, 8...

I tried to put the red pole of the multimeter to ground (keyboard connector ground or p10 of u10) and they all gave me a reading in multimeter.

So now what we can say?

biessea wrote on 2022-09-11, 09:36:

Then I tried to have the chock to the diode, and here something isn't correct as you explained. The black pole of the multimeter put on the +5V and the red put on the SIMM pins that you have me told before, my multimeter for the pin 15 and 14 result in a ZERO (0) no value like the other pins like 12, 8...

That is surprising. Do you mean it shows "OL.", which would indicate "no conductivity", like at https://youtu.be/s_ajW1dITFY?t=501 ? That's not necessarily wrong, as the 74F245 is a bipolar device, and conductivity from the input/output pins to +5V is not necessarily provided on them. But all the SIMM pins I indicated are driven by the same kind of 74F245 chip, so they should behave the same. Did you have any memory modules installed when you did that test?

mkarcher wrote on 2022-09-11, 10:25:

biessea wrote on 2022-09-11, 09:36:

Then I tried to have the chock to the diode, and here something isn't correct as you explained. The black pole of the multimeter put on the +5V and the red put on the SIMM pins that you have me told before, my multimeter for the pin 15 and 14 result in a ZERO (0) no value like the other pins like 12, 8...

That is surprising. Do you mean it shows "OL.", which would indicate "no conductivity", like at https://youtu.be/s_ajW1dITFY?t=501 ? That's not necessarily wrong, as the 74F245 is a bipolar device, and conductivity from the input/output pins to +5V is not necessarily provided on them. But all the SIMM pins I indicated are driven by the same kind of 74F245 chip, so they should behave the same. Did you have any memory modules installed when you did that test?

No, absolutely. Motherboard testing taken apart with no components on it.

I really don't know what to say, there is more test I can do? I hope to have understood good what to do, cause I don't want to give false results.

And the keyboard thing? What do you think?

biessea wrote on 2022-09-11, 10:35:

No, absolutely. Motherboard testing taken apart with no components on it.

OK. From what I see on the photos, the address pins on the SIMM sockets should not be connected to anything except the pins on a digital amplifier like U10. So all 8 low address pins should behave identically. Differences here might point to the issue you are having with the RAM at the moment. Please verify that the all of address pins at the SIMM sockets are connected between all four sockets of the first bank. Please also tell me whether the numbers you see in "diode check" mode are approximately the same or vary vastly between the pins, and please tell me the number(s) you are getting.

biessea wrote on 2022-09-11, 10:35:

And the keyboard thing? What do you think?

The keyboard thing sounds fine. No broken traces over there.

mkarcher wrote on 2022-09-11, 12:13:

OK. From what I see on the photos, the address pins on the SIMM sockets should not be connected to anything except the pins on a […]
Show full quote

biessea wrote on 2022-09-11, 10:35:

No, absolutely. Motherboard testing taken apart with no components on it.

OK. From what I see on the photos, the address pins on the SIMM sockets should not be connected to anything except the pins on a digital amplifier like U10. So all 8 low address pins should behave identically. Differences here might point to the issue you are having with the RAM at the moment. Please verify that the all of address pins at the SIMM sockets are connected between all four sockets of the first bank. Please also tell me whether the numbers you see in "diode check" mode are approximately the same or vary vastly between the pins, and please tell me the number(s) you are getting.

biessea wrote on 2022-09-11, 10:35:

And the keyboard thing? What do you think?

The keyboard thing sounds fine. No broken traces over there.

Please friend, explain me better what you want me to do.

I am not sure I have understood.

Please verify that the all of address pins at the SIMM sockets are connected between all four sockets of the first bank

What do you want me to do? Please I am unfortunately not a technician in electronics, so you have please to do a step to step every request you gave me to control over this board. This is the only way I can reach what you ask me to do and we can continue try to fix this unknown problem in this old great 386 board.

Please also tell me whether the numbers you see in "diode check" mode are approximately the same or vary vastly between the pins, and please tell me the number(s) you are getting

I have to write down a table with all the numbers I get by the multimeter an write here down, correct? In diode check mode I think. In red plug on GROUND and in the black plug on 5v? Correct?

If I have understood good I re-tried to measure the diode test and write down all the results:

SIMM slots, pin 15 (A7)
SIMM slots, pin 14 (A6)
SIMM slots, pin 12 (A5)
SIMM slots, pin 11 (A4)
SIMM slots, pin 8 (A3)
SIMM slots, pin 7 (A2)
SIMM slots, pin 5 (A1)
SIMM slots, pin 4 (A0)

These are the SIMM slot pins that I have to test, from the 15 in the middle to the pin 4 going to the "keyboard connector area".

These are the results:

+5V test (black stick of the multimeter in a 5V pin, i choose the keyboard connector pin 40 that it's eady to use)
BANK0 (from the lower SIMM bank to the fourth SIMM bank)
1) OL - OL - OL - OL - OL - OL - OL - OL
2)OL - OL - OL - OL - OL - OL - OL - OL
3)OL - OL - OL - OL - OL - OL - OL - OL
4)OL - OL - OL - OL - OL - OL - OL - OL
BANK1 (from the middle SIMM bank to the eight upper SIMM bank)
1) OL - OL - OL - OL - OL - OL - OL - OL
2)OL - OL - OL - OL - OL - OL - OL - OL
3)OL - OL - OL - OL - OL - OL - OL - OL
4)OL - OL - OL - OL - OL - OL - OL - OL

GRN test (red stick of the multimeter in a GRN pin, I choose the pin 10 of U10 chip)
BANK 0 (from the lower SIMM bank to the fourth SIMM bank)
1) 0,440 - 0,440 - 0,440 - 0,440 - 0,440 - 0,440 - 0,440 - 0,440
2) 0,440 - 0,440 - 0,440 - 0,440 - 0,440 - 0,440 - 0,440 - 0,440
3) 0,440 - 0,440 - 0,440 - 0,440 - 0,440 - 0,440 - 0,440 - 0,440
4) 0,440 - 0,440 - 0,440 - 0,440 - 0,440 - 0,440 - 0,440 - 0,440
BANK1 (from the middle SIMM bank to the eight upper SIMM bank)
1) 0,438 - 0,438 -0,438 - 0,438 - 0,438 - 0,438 -0,438 - 0,438
2) 0,438 - 0,438 -0,438 - 0,438 - 0,438 - 0,438 -0,438 - 0,438
3) 0,438 - 0,438 -0,438 - 0,438 - 0,438 - 0,438 -0,438 - 0,438
4) 0,438 - 0,438 -0,438 - 0,438 - 0,438 - 0,438 -0,438 - 0,438

I hope it gives you more interesting indication.

So tell me more about the other test that I have not understand what to do.

Thanks.

biessea wrote on 2022-09-12, 12:38:

If I have understood good I re-tried to measure the diode test and write down all the results:

[...]

I hope it gives you more interesting indication.

Good news: These values are the ones that I need to get to a conclusion. Bad news: These values all look fine, so it doesn't look like your issue is caused with the low address bits of the RAM.

I'm sorry, but at the moment I am out of ideas what to try with the equipment and skill you currently have. The board is very close to working fine, but somewhere is a minor fault, and it is difficult to locate that fault. The next thing I would do if it were my own board: Use an ISA POST card to trace the progress of the self test, and disassemble the BIOS (after reading it in an EPROM reader) to find out what specifically the BIOS is doing before it goes beeping. That means: In this scenario of remote debugging, I have to give up.

Your 386DX-40, the coprocessor and the RAM are still good components, and assuming the fault is on the board (which is quite likely), I would keep looking at ebay, craigslist or similar for another 40MHz-capable 386 board for an acceptable price.

mkarcher wrote on 2022-09-12, 14:47:

Good news: These values are the ones that I need to get to a conclusion. Bad news: These values all look fine, so it doesn't loo […]
Show full quote

biessea wrote on 2022-09-12, 12:38:

If I have understood good I re-tried to measure the diode test and write down all the results:

[...]

I hope it gives you more interesting indication.

Good news: These values are the ones that I need to get to a conclusion. Bad news: These values all look fine, so it doesn't look like your issue is caused with the low address bits of the RAM.

I'm sorry, but at the moment I am out of ideas what to try with the equipment and skill you currently have. The board is very close to working fine, but somewhere is a minor fault, and it is difficult to locate that fault. The next thing I would do if it were my own board: Use an ISA POST card to trace the progress of the self test, and disassemble the BIOS (after reading it in an EPROM reader) to find out what specifically the BIOS is doing before it goes beeping. That means: In this scenario of remote debugging, I have to give up.

Your 386DX-40, the coprocessor and the RAM are still good components, and assuming the fault is on the board (which is quite likely), I would keep looking at ebay, craigslist or similar for another 40MHz-capable 386 board for an acceptable price.

Oh my God, I never wanted to read this.

I really hoped to make this board work, I really hoped to solve the problem. Locate and solve it.

Now instead I have to give up, I'm really sad about that.

But in my past I had other big lucky moment, I assume that I have to accept that this is simply a sad moment where I cannot do anything.

Anyway I will take this board for a while, probably for my "museum" of old hardware show.

Nothing more, if you have other idea my dear, I am here. I really hope something other will come to your great mind.

Thanks a lot for all this useful help. It was fun. Not the end, but the path was fun.

Big hug,

Loris