Dell system 200 memory failure errors

[modem7]

Probably have some M27C256 eeproms arriving today too. Always handy to have some spare.

Presumably Winbond W27E257 EEPROM's.

 

[roadrash]

Since i didn't get the RDR roms to work I downloaded the Landmark supersoft roms from minus zero degrees site to test my programmer was still working properly. I notice the landmark roms from minus zero degrees site are different to the ones I've been using. I dont remember where I got them from now but they both give a Diffrent output.
This is a screenshot of the output from the minus zero degrees landmark diagnostic rom. Does it help diagnosing the problem.

 

[modem7]

Since i didn't get the RDR roms to work I downloaded the Landmark supersoft roms from minus zero degrees site to test my programmer was still working properly. I notice the landmark roms from minus zero degrees site are different to the ones I've been using. I dont remember where I got them from now but they both give a Diffrent output.
This is a screenshot of the output from the minus zero degrees landmark diagnostic rom.

The screenshot in your post shows the output of the draft AT version of Ruuds Diagnostic ROM (RDR) that I DM'ed to you at post #15.

This draft version was tested successfully on my IBM AT's. Testing included removing select RAM chips from banks 0 and 1, verifying that this draft version displayed the problem as expected. On a fully functional IBM AT motherboard, this draft version passed the 'Check first 2 KB of RAM' test, then went on to correctly determine the top-of-RAM (256 KB or 512 KB for my IBM AT motherboards), and then successfully test that RAM (data, refresh, address, ...)

Does it help diagnosing the problem.

As I expected, the 'Check first 2 KB of RAM' test failed.

The first thing to note is that RDR is not showing you a misleading address (the Supersoft Landmark Diagnostic ROM [SLDR] is showing you address 6E40). RDR found a problem at the first address it tested, address 00000 (the first address in bank 0), and then RDR aborted.

A reminder that unlike the AT version of SLDR, RDR tests all bits at the failing address before displaying the 'bad bits'.

Bit 14 in bank 0

Quite possibly the bit 14 RAM chip in bank 0. But it could be something else.

I see that earlier, we had identified bit 14 in bank 0 as a problem, but we were diverted when SLDR then started showing bits 2,3, and 4, in error.

If the diagram at [here] is correct, and the 'bit 14 in bank 0' problem is caused by the corresponding RAM chip, then U86 is the chip to replace.

Bits 0 through 7 in bank 0

In post #9, you replaced the RAM chips in bank 0 that correspond to bits 2, 3, and 4. Assuming that was done correctly, then in respect of bits 2, 3, and 4, the corresponding RAM chips are not the problem.

Bits 0 through 7 are all bits in the low byte of the word. That tends to point to something like an 8-bit transceiver chip, or the circuitry that instructs the transceiver how to behave (e.g. enables, direction).

Going forward

I suggest that you replace U86 and see if that fixes the bit 14 problem.

As for bits 0 through 7, I am going to review our 'Dell system 200' in-confidence correspondence that we had years ago. From my memory of the content, I may be able to identify one or more non-RAM chip for you to replace.

 

[roadrash]

The screenshot in your post shows the output of the draft AT version of Ruuds Diagnostic ROM (RDR) that I DM'ed to you at post #15.

This draft version was tested successfully on my IBM AT's. Testing included removing select RAM chips from banks 0 and 1, verifying that this draft version displayed the problem as expected. On a fully functional IBM AT motherboard, this draft version passed the 'Check first 2 KB of RAM' test, then went on to correctly determine the top-of-RAM (256 KB or 512 KB for my IBM AT motherboards), and then successfully test that RAM (data, refresh, address, ...)


As I expected, the 'Check first 2 KB of RAM' test failed.

The first thing to note is that RDR is not showing you a misleading address (the Supersoft Landmark Diagnostic ROM [SLDR] is showing you address 6E40). RDR found a problem at the first address it tested, address 00000 (the first address in bank 0), and then RDR aborted.

A reminder that unlike the AT version of SLDR, RDR tests all bits at the failing address before displaying the 'bad bits'.

Bit 14 in bank 0

Quite possibly the bit 14 RAM chip in bank 0. But it could be something else.

I see that earlier, we had identified bit 14 in bank 0 as a problem, but we were diverted when SLDR then started showing bits 2,3, and 4, in error.

If the diagram at [here] is correct, and the 'bit 14 in bank 0' problem is caused by the corresponding RAM chip, then U86 is the chip to replace.

Bits 0 through 7 in bank 0

In post #9, you replaced the RAM chips in bank 0 that correspond to bits 2, 3, and 4. Assuming that was done correctly, then in respect of bits 2, 3, and 4, the corresponding RAM chips are not the problem.

Bits 0 through 7 are all bits in the low byte of the word. That tends to point to something like an 8-bit transceiver chip, or the circuitry that instructs the transceiver how to behave (e.g. enables, direction).

Going forward

I suggest that you replace U86 and see if that fixes the bit 14 problem.

As for bits 0 through 7, I am going to review our 'Dell system 200' in-confidence correspondence that we had years ago. From my memory of the content, I may be able to identify one or more non-RAM chip for you to replace.

Going forward

I suggest that you replace U86 and see if that fixes the bit 14 problem.

As for bits 0 through 7, I am going to review our 'Dell system 200' in-confidence correspondence that we had years ago. From my memory of the content, I may be able to identify one or more non-RAM chip for you to replace.

Thanks Ray,
I replaced U86 and here is a screenshot of what your RDR rom reports. i hope this is getting more encouraging and we are getting nearer to the suspect chip. Ive still got another new 3 ram chips left.

 

[modem7]

I replaced U86 and here is a screenshot of what your RDR rom reports.

Good. 'Bit 14 in bank 0' problem fixed.

i hope this is getting more encouraging and we are getting nearer to the suspect chip. Ive still got another new 3 ram chips left.

I doubt that the 'bits 0 through 7 in bank 0' problem is due to 8 faulty RAM chips. Although we have had people report here with many faulty RAM chips on their motherboard, earlier (post #9), you replaced the RAM chips in bank 0 that correspond to bits 2, 3, and 4. Those bits are still showing in error, suggesting either:
- Wrong type of replacement RAM chips were used; or
- Poor de-soldering/soldering activity; or
- A PCB problem (e.g. open-circuit trace) that impacts on the operation of the low byte of RAM bank 0; or
- A faulty chip that supports RAM bank 0 operation.

I highly suspect the latter two due to all of the problem bits being the low byte of the word.

As for bits 0 through 7, I am going to review our 'Dell system 200' in-confidence correspondence that we had years ago. From my memory of the content, I may be able to identify one or more non-RAM chip for you to replace.

I cannot locate the correspondence. Do you still have that? It may have included a full or partial circuit diagram.

Otherwise, we have to take some educated guesses. For example, starting to get a picture of the circuitry, by working out (via multimeter) which pins of which chips that the data pin of RAM chip U70 (bit 0 in bank 0) connects to.

 

[roadrash]

Good. 'Bit 14 in bank 0' problem fixed.[/qoute]

glad something is fixed.

I doubt that the 'bits 0 through 7 in bank 0' problem is due to 8 faulty RAM chips. Although we have had people report here with many faulty RAM chips on their motherboard, earlier (post #9), you replaced the RAM chips in bank 0 that correspond to bits 2, 3, and 4. Those bits are still showing in error, suggesting either:
- Wrong type of replacement RAM chips were used; or
- Poor de-soldering/soldering activity; or
- A PCB problem (e.g. open-circuit trace) that impacts on the operation of the low byte of RAM bank 0; or
- A faulty chip that supports RAM bank 0 operation.

I highly suspect the latter two due to all of the problem bits being the low byte of the word.

I must admit i suspect possible poor de-soldering etc or track fault as in my de-soldering when i did the first ram chips years ago was not good. i de-soldered chips in one go all legs to remove chip. Where now i cut all legs to remove chip then de-solder each leg. its a lot tidier and its kinder on the pcb. If you want to tell me which ram chips to remove i will do this and then we can check each track to the next connected chip in the circuit.

I cannot locate the correspondence. Do you still have that? It may have included a full or partial circuit diagram.

Otherwise, we have to take some educated guesses. For example, starting to get a picture of the circuitry, by working out (via multimeter) which pins of which chips that the data pin of RAM chip U70 (bit 0 in bank 0) connects to.

I have looked through everything i have had from member bcc2, DM;s and emails and there is only a mention of a wiring diagram but i never received anything from him.

 

[modem7]

I must admit i suspect possible poor de-soldering etc or track fault as in my de-soldering when i did the first ram chips years ago was not good. i de-soldered chips in one go all legs to remove chip. Where now i cut all legs to remove chip then de-solder each leg. its a lot tidier and its kinder on the pcb. If you want to tell me which ram chips to remove i will do this and then we can check each track to the next connected chip in the circuit.

The reference diagram is the one at [here]. Chips U70 through U78.

 

[roadrash]

The reference diagram is the one at [here]. Chips U70 through U78.

Thanks Ray, The error in area 6E40 was a error I had at the very start of working on this Dell PC a long time ago and that time I replaced chips U70 & U80 which fixed the error. Do you think I should replace them again? I have been checking all connections between the ML1259 chips with my multi-meter and I have noticed that all pins on the 16 pin MT1259 chips are linked except for pins 2 and 11 is that correct?

 

[modem7]

Thanks Ray, The error in area 6E40 was a error I had at the very start of working on this Dell PC a long time ago and that time I replaced chips U70 & U80 which fixed the error. Do you think I should replace them again?

I do not see any point.

I have been checking all connections between the ML1259 chips with my multi-meter and I have noticed that all pins on the 16 pin MT1259 chips are linked except for pins 2 and 11 is that correct?

The MT1259 RAM chips being the ones in bank 0.

Good: Pin 2 is the data pin. So, pin 2 of the bank 0 chips are expected not to be connected together.

Bad: Pin 11 is address bit 4 (A4). Therefore, I expect that pin 11 of all bank 0 RAM chips be connected together. Surely at the least, some are connected together. Recheck this. If there was an A4 problem, and it was the only problem, I would expect the 'AT version of RDR' to pass the 'Check first 2KB of RAM' test, and then go to fail the address test.

But maybe there are multiple problems. You are seeing something unexpected in regards to A4 in bank 0, and so that needs to be investigated. For example, are you not seeing continuity between pin 11 of U83, and
pin 11 of U84 ?

 

[roadrash]

I do not see any point.


The MT1259 RAM chips being the ones in bank 0.

Good: Pin 2 is the data pin. So, pin 2 of the bank 0 chips are expected not to be connected together.

Bad: Pin 11 is address bit 4 (A4). Therefore, I expect that pin 11 of all bank 0 RAM chips be connected together. Surely at the least, some are connected together. Recheck this. If there was an A4 problem, and it was the only problem, I would expect the 'AT version of RDR' to pass the 'Check first 2KB of RAM' test, and then go to fail the address test.

But maybe there are multiple problems. You are seeing something unexpected in regards to A4 in bank 0, and so that needs to be investigated. For example, are you not seeing continuity between pin 11 of U83, and
pin 11 of U84 ?

View attachment 1322140

Sorry its pin 2 & pin 14 that have no connection between each chip.

 

[modem7]

Sorry its pin 2 & pin 14 that have no connection between each chip.

That's better.

And what you may see is that on each chip, pin 2 (data in) is connected to pin 14 (data out). Where that may not be the case, is on the parity chips. Confirm the pin-2-to-pin-14 connection on the data chips for me.

Some facts:
• Bank 0 : Low byte failing
• Bank 0 : High byte passing
• Bank 0 : The non-data pins of each bank 0 chip (i.e. 16 bits) have a verified connection to the same pin on the other chips in the bank. Includes ground and +5V.

We do not have a circuit diagram of your motherboard, but I bet that the bank 0 circuitry is very close to, or even the same as, the bank 0 circuitry in the IBM 5170. The 5170's bank 0 circuitry is shown at [here]. Let us pretend that it is your circuitry. In the diagram, note the chips and paths coloured in green. That green shows the known-working chips and paths associated with the working high byte of bank 0.

Pin 3, WRITE ENABLE, is verified connected to all 16 RAM chips, and therefore, a functioning WRITE ENABLE is reaching the low 8 RAM chips (bits 0 to 7).
Pin 4, RAS-0, is verified connected to all 16 RAM chips, and therefore, a functioning RAS-0 is reaching the low 8 RAM chips (bits 0 to 7).

In the IBM 5170, there are two CAS signals for bank 0:
• CAS-0-L for the low byte of bank 0, targeting pin 15 of each RAM chip in the low byte of bank 0
• CAS-0-H for the high byte of bank 0, targeting pin 15 of each RAM chip in the high byte of bank 0
I.e. The 'pin 15's of the low byte of bank 0 are not connected to the 'pin 15's of the high byte of bank 0.

But on your motherboard, you implied that the 'pin 15's of all 16 RAM chips in bank 0 are connected together. That would differ to the IBM 5170. Please verify what the story is in regard to your pin 15.

Back to the IBM 5170 diagram. For the symptoms, what I would suspect is:
• Faulty U5
• Faulty generation of CAS-0-L
• Open-circuit between the DIR pins of U5 and U11 (i.e. direction signal reaching U11, but not U5)
• XA0 not reaching U5

But not having the circuit diagram for your Dell motherboard makes things tricky. What you could do is see if there is an equivalent to U5 on your motherboard. That is, does pin 2 of U70 connect to one of the A/B pins of a 74245? If so, try replacing the 74245 (observing correct replacement, e.g. 74ALS245 for 74ALS245, etc.)

 

[modem7]

Back to the IBM 5170 diagram. For the symptoms, what I would suspect is:
• Faulty U5
• Faulty generation of CAS-0-L
• Open-circuit between the DIR pins of U5 and U11 (i.e. direction signal reaching U11, but not U5)
• XA0 not reaching U5

I enhanced the 5170 diagram. It turns out that the BIOS ROM's are on the memory data bus. So for the symptoms on a 5170, the possibilities would be:
• A problem in the generation of CAS for the low byte of bank 0.
• The ROM supplying the low byte is always outputting its data, resulting in bus contention, and always wining the contention 'war'. (Faulty ROM, or good ROM always commanded to output).

But not having the circuit diagram for your Dell motherboard, we don't know where in the circuitry the ROM's are. To see if they are on the same data bus as the RAM, see if there is continuity between the data pin of U70 (U70 = bit 0 of low byte of RAM) and the data 0 pin (pin 11) of whichever BIOS ROM (27C256's in your Dell) is supplying the low byte.

 

[roadrash]

I enhanced the 5170 diagram. It turns out that the BIOS ROM's are on the memory data bus. So for the symptoms on a 5170, the possibilities would be:
• A problem in the generation of CAS for the low byte of bank 0.
• The ROM supplying the low byte is always outputting its data, resulting in bus contention, and always wining the contention 'war'. (Faulty ROM, or good ROM always commanded to output).

But not having the circuit diagram for your Dell motherboard, we don't know where in the circuitry the ROM's are. To see if they are on the same data bus as the RAM, see if there is continuity between the data pin of U70 (U70 = bit 0 of low byte of RAM) and the data 0 pin (pin 11) of whichever BIOS ROM (27C256's in your Dell) is supplying the low byte.

I enhanced the 5170 diagram. It turns out that the BIOS ROM's are on the memory data bus. So for the symptoms on a 5170, the possibilities would be:
• A problem in the generation of CAS for the low byte of bank 0.
• The ROM supplying the low byte is always outputting its data, resulting in bus contention, and always wining the contention 'war'. (Faulty ROM, or good ROM always commanded to output).

But not having the circuit diagram for your Dell motherboard, we don't know where in the circuitry the ROM's are. To see if they are on the same data bus as the RAM, see if there is continuity between the data pin of U70 (U70 = bit 0 of low byte of RAM) and the data 0 pin (pin 11) of whichever BIOS ROM (27C256's in your Dell) is supplying the low byte.

I have just spent ages going along the 2 rows of RAM chips (U70-U78 and U80-U88) checking that all pins from one chip to the next are connected for pins 1 to 16 except pins 2 and 14, I found a lot of missing connections between some chips and where there was no connection I linked between the 2 chips with a small bit of wire creating a connection. The Chips that had no connection were Chips U73, 74, 75, 76 and U77.
The Ram chips at the end of each row are U94 last chip in row 1 and after U78. Then U95 is last chip in row 2 and after U88. The last in row chips dont follow the same connection pattern as the other chips. Maybe these are the Parity chips you mentioned.
anyway I turned the dell on and here is a screenshot of what your diag Rom reported. i would think what Ive done has cured a lot of ram errors from the last diag check.

 

[roadrash]

And what you may see is that on each chip, pin 2 (data in) is connected to pin 14 (data out). Where that may not be the case, is on the parity chips. Confirm the pin-2-to-pin-14 connection on the data chips for me.

Sorry forgot to do this for you. I get continuity between pins 2 & 14 on every Ram chip except on U25 & U28.

Pin 3, WRITE ENABLE, is verified connected to all 16 RAM chips, and therefore, a functioning WRITE ENABLE is reaching the low 8 RAM chips (bits 0 to 7).
Pin 4, RAS-0, is verified connected to all 16 RAM chips, and therefore, a functioning RAS-0 is reaching the low 8 RAM chips (bits 0 to 7).

In the IBM 5170, there are two CAS signals for bank 0:
• CAS-0-L for the low byte of bank 0, targeting pin 15 of each RAM chip in the low byte of bank 0
• CAS-0-H for the high byte of bank 0, targeting pin 15 of each RAM chip in the high byte of bank 0
I.e. The 'pin 15's of the low byte of bank 0 are not connected to the 'pin 15's of the high byte of bank 0.

I can confirm there is no connection between pins 15 in bank 0 and pin 15 in bank 1,

But not having the circuit diagram for your Dell motherboard makes things tricky. What you could do is see if there is an equivalent to U5 on your motherboard. That is, does pin 2 of U70 connect to one of the A/B pins of a 74245? If so, try replacing the 74245 (observing correct replacement, e.g. 74ALS245 for 74ALS245, etc.)

The 2 74LS245 chips on my motherboard are U85 & U43. There is no continuity between U70 pin2 and any pins on the 74LS245 chips.
also pin 2 of U70 does not show any continuity to any pins on the 2

There is no continuity between pin 2 of U70 and any pins on the 2 Bios roms (27C256's)

Sorry for not getting everything done you asked me but because we have had good dry weather ive been going back and forth between fixing my motorbikes and then taking another look at this. Trouble is i keep forgetting where I left off the time before. We are having some poor weather for a little while so will dedicate more time to getting this Dell 200 sorted.

 

[modem7]

Sorry for not getting everything done you asked me but because we have had good dry weather ive been going back and forth between fixing my motorbikes and then taking another look at this.

No apology required. I had a colonoscopy, and there were some other local matters to deal with. This is a hobby.

anyway I turned the dell on and here is a screenshot of what your diag Rom reported. i would think what Ive done has cured a lot of ram errors from the last diag check.

Comparing that screenshot to the one in post #24, bits 7 and 6 are now working. It throws out the hypothesis that a certain chip failure was impacting on the entire low byte of bank 0.

If in bank 0, we compare one of the working bits, bit 8, with one of the faulty bits, bit 0, we end up with the diagram at [here].
Let's look at U70.
U70: You have verified (from working bits) that pins 2 and 14 are connected together.
U70: Apart from the data pins (2+14), all other pins (including VCC {+5V} and VSS {ground}) on U70, connect to the corresponding pins on working chips.
Therefore, I think that the data pins (2+14) need to be investigated.

The 2 74LS245 chips on my motherboard are U85 & U43. There is no continuity between U70 pin2 and any pins on the 74LS245 chips.

I see three 74LS245, U54 being the third. It's next to ISA slot 6.

Note that your motherboard uses part of a chip set, the CHIPS and Technologies CS8220. Taking a look at that chipset:
- A data sheet for the chip set is at [here].
- In the data sheet, the block diagram reveals that the 82A205 chip, acts as the buffer between the SD bus and the MD bus. (IBM 5170 diagram at [here].)

Can you confirm that pin 2 on U80 (working bit) is connected to one of the pins on the 82A205 chip. There is a pin-out of the 82A205 on page 3 of the aforementioned data sheet. Expected is that the 82A205 pin used is pin 2 (MD8), but for all we know, Dell may not have used direct pin mapping.

If that connection exists, see if pin 2 on U70 (not working bit) is connected to the 82A205.

Have you re-seated the 82A205 ?

 

[roadrash]

No apology required. I had a colonoscopy, and there were some other local matters to deal with. This is a hobby.

Ouch! was that the camera up the bum. I had that done twice myself.

Confirm the pin-2-to-pin-14 connection on the data chips for me.

I get continuity between pins 2 & 14 on every Ram chip except on U25 & U28 is this correct?

But on your motherboard, you implied that the 'pin 15's of all 16 RAM chips in bank 0 are connected together. That would differ to the IBM 5170. Please verify what the story is in regard to your pin 15.

checked all pin 15 connections and this is what I found:
there is continuity between pin 15 on U70 to U75
and U94. But there is no continuity between U75 & U76 (Do I need to put a link between U75 & U76?) but there is continuity then between the remaining chips U76 to U78 and U95
On the second row there is continuity between U80 to U88 & U95.

Have you re-seated the 82A205 ?

I have now and here is screenshot by RDR

 

[modem7]

Have you re-seated the 82A205 ?

I have now and here is screenshot by RDR

That is odd. 64 characters wide instead of 80, and other. Maybe try re-seating the 82A205 again.

I get continuity between pins 2 & 14 on every Ram chip except on U25 & U28 is this correct?

Pins 2 and 14 are DATA IN and DATA OUT. When I look at the RAM diagram at [here], I do not see a U25 nor U28.

If you meant the parity chips, U78 and U88, then that is not unexpected. It comes down to the design. For example, in the IBM 5150, per [here], the data in/out pins are connected on each parity chip. But in the IBM 5160, per [here], the data in/out pins are not connected on each parity chip. If U78 and U88 are the chips you meant, then I think we can safely deduce that in the System 200 design, Dell has the data in/out pins pins disconnected on each parity chip.

checked all pin 15 connections and this is what I found:
there is continuity between pin 15 on U70 to U75
and U94. But there is no continuity between U75 & U76 (Do I need to put a link between U75 & U76?) but there is continuity then between the remaining chips U76 to U78 and U95
On the second row there is continuity between U80 to U88 & U95.

You have quoted U94 and U95. Those are in bank 1. We are only interested in bank 0 at this time.

Pin 15 is the CAS pin on an MT1259. In bank 0 (chips U70 to U88), all chips should have pin 15 (CAS) connected to each other. And so it sounds like you have found the cause of the current symptom of 'bits 0 to 5 in bank 0 are faulty'. Yes, connect pin 15 on U75 to pin 15 on U76.

 

[roadrash]

Pin 15 is the CAS pin on an MT1259. In bank 0 (chips U70 to U88), all chips should have pin 15 (CAS) connected to each other. And so it sounds like you have found the cause of the current symptom of 'bits 0 to 5 in bank 0 are faulty'. Yes, connect pin 15 on U75 to pin 15 on U76.

Thats it sorted I believe, check the screenshot of RDR. I dont understand how i didnt notice the break in the circuit between U75 & U76 when I checked all the other Ram chip connections. It might have been due to auto multimeter I had that kept giving me unreliable results so I threw it out and used a old cheap meter and got consistent results straight away. So much foe some modern expensive meters.[/quote]

I am so grateful to you Ray, thats twice you have helped me rescue this Dell prcission 200 {286). I have a 386 version that has given no problems. This 286 version system 200 has been a challenge from the start and would never have been sorted had we not originally found Dell had put a bit of code in the standartd IBM type keyboard controller and lucky a ex Dell employee had a New old stock chip we managed to get the code out off. At least its now on this site for others to get if they have the same issue, I fail to understand that after 5 years sitting in my warm dry loft all these memory failures happend. Some may have been caused by me doing a poor job de-soldering the chips (now using a better method now) but some the result of failed chips and degraded connections. Sometimes makes me wonder if trying to keep these old systems is worth it when they just fail over time and not use. I every few years warmed all my systems up but still they are failing over time

I wonder Ray if you are interested in helping me recover ) a old but now immaculate Apricot PC (in a new thread of course), I spent months restoring including treating all the cases to remove the yellowing and now look good as new. I did have it all working perfect, and it does still boot up but cant load a system up due to disk errors. I have a compaq demo disk that when booted is impressive and shows how advanced this suyetem was for its time. If I could get this working again it would be great. I bet there are almost none working now.

 

[modem7]

Thats it sorted I believe, check the screenshot of RDR.

Good news.

I fail to understand that after 5 years sitting in my warm dry loft all these memory failures happened. Some may have been caused by me doing a poor job de-soldering the chips (now using a better method now) but some the result of failed chips and degraded connections.

I have certainly had cases where I pulled out a vintage computer that was put away in a fully functional state, only to discover a problem that I then tracked down to a faulty chip. Rhetorical: Did the chip go faulty during the storage period, or go faulty at turn-on time? I do not recall being in that situation with it turning out that multiple chips had failed. I would then be saying to myself, surely those multiple chips didn't all go faulty at power-on time.

I wonder Ray if you are interested in helping me recover ) a old but now immaculate Apricot PC (in a new thread of course), I spent months restoring including treating all the cases to remove the yellowing and now look good as new. I did have it all working perfect, and it does still boot up but cant load a system up due to disk errors.

Create the thread. I'm sure that many here will be able to assist.

 

[roadrash]

Good news.

It was good news and i put it all back together and set up the bios. But had a problem with hard drivebutton on this PC giving errors so thought I would boot from a floppy just for now. But noticed floppy wasnt reading. anyway checked all cables were on correctly then tried turned it on again but got a 4, 2, 3 beep error. I have got the motherboard back on the bench and first tried the supersoft roms and it shows all had passed except a Gate A20 error. So put RDR roms in the tried again and it shows memory errors. Here is a screenshot of the output. Its a bit strange because its been starting up without erros many times. I occasionally had a problem when it failed to start and found if i briefly connected the two reset pins it restarted and went on to pass post. This has happened a few times. There is no reset button on this PC but could fit one if I had too.
Ive checked all the jumper wires I soldered on and all are still connected so could you please help trace this latest memory error.