Tandon 810 won't boot or format

[TZJB]

7 hours ago, dukes909 said:

Do you mean adjust the low-frequency compensation adjustment screw on the probe? I ask because when I check the test waveform built into the scope I get a perfectly square signal:

Working on the rest of your suggestions...was away for a couple of days so it's going to take a bit to get caught up. Thanks

 

Yes. That looks excellent.

[dukes909]

2 hours ago, TZJB said:

 

Yes. That looks excellent.

But that is the test waveform without changing any settings...not from the 810.  Next question: should I be using 10X or 1X on the probes for checking signals on the 810 parts - and maybe that is the reason for wonky waveforms? (I told you I am new to oscilloscopes! 🤓)

[dukes909]

On 7/11/2023 at 10:47 AM, TZJB said:

That looks a lot squarer, but I think that you may need to adjust your probe to show a squarer signal.

Ok here is A201 pin 6same probe on 10x setting.  Not much difference. Since the 1KHz test signal is perfectly square I don't think I need to adjust the probes at all?

On 7/11/2023 at 10:47 AM, TZJB said:

Shame changing the 74LS74 didn't fix it yet, though. Obviously leave that one in now.

Okay!

On 7/11/2023 at 10:47 AM, TZJB said:

Looking back, your observation regarding multiple failures may be correct.

Shoot!

On 7/11/2023 at 10:47 AM, TZJB said:

There may be a problem with A206 as the output is a bit wonky (technical term) so definitely substitute that one from your other board also. Then if the fault persists record some more waveforms at A206 pin 7 and A205 pin 8. 

Reading still doesn't work.

A206 pin 7 produces this signal, but while the scope probe is attached to pin 7, the drive does not spin up when doing a disk directory command. Nothing happens at all. Then an ERROR 138 is returned. Here is what scope shows anyway for A206 pin 7:

And A205 pin 8:

On 7/11/2023 at 10:47 AM, TZJB said:

record some more waveforms at A206 pin 7 and A205 pin 8. If they are not similar, change A205 too.

Same thing as above. Won't read. Waveforms identical for A205 pin 8. With probe on A206 pin 7, drive won't spin. If I remove probe from A206 pin 7 before ERROR 138 is returned, the drive will spin up and then do its' convolutions and then ERROR 144.

On 7/11/2023 at 10:47 AM, TZJB said:

Lastly change A207 if it's still not playing ball.

Running out of things to change...LOL!

Changed A207 and...won't read...

But now the waveforms for the IC's above are a little different...?

A206 pin 7: (still get ERROR 138 though)

Lastly, A205 pin 8:

Uggghghhhhhh...............

[TZJB]

2 hours ago, dukes909 said:

Ok here is A201 pin 6same probe on 10x setting.  Not much difference. Since the 1KHz test signal is perfectly square I don't think I need to adjust the probes at all?

Okay!

Shoot!

Reading still doesn't work.

A206 pin 7 produces this signal, but while the scope probe is attached to pin 7, the drive does not spin up when doing a disk directory command. Nothing happens at all. Then an ERROR 138 is returned. Here is what scope shows anyway for A206 pin 7:

And A205 pin 8:

Same thing as above. Won't read. Waveforms identical for A205 pin 8. With probe on A206 pin 7, drive won't spin. If I remove probe from A206 pin 7 before ERROR 138 is returned, the drive will spin up and then do its' convolutions and then ERROR 144.

Running out of things to change...LOL!

Changed A207 and...won't read...

But now the waveforms for the IC's above are a little different...?

A206 pin 7: (still get ERROR 138 though)

Lastly, A205 pin 8:

Uggghghhhhhh...............

 

Your probe did look to be well adjusted so nothing to adjust on the probe as you said.

 

None of the waveforms after A201 pin 6 look very logical. Can you please verify that each logic chip is receiving 5V?

 

I would not expect the probe to affect the operation of the disk drive. It's a long shot but maybe add an extra ground wire to the board and attach your probe's ground to it?

 

Failing having to buy new logic chips. Do you have a way of testing the logic chips independently such as in a XGPRO TL866 II programmer or similar?

 

It may just be easier to buy new ones though, if the above tests prove okay but don't improve the situation.

 

7 hours ago, dukes909 said:

But that is the test waveform without changing any settings...not from the 810.  Next question: should I be using 10X or 1X on the probes for checking signals on the 810 parts - and maybe that is the reason for wonky waveforms? (I told you I am new to oscilloscopes! 🤓)

 

Sorry I missed this message. Normally X10 should be used I believe.

Edited July 13, 2023 by TZJB

[dukes909]

49 minutes ago, TZJB said:

Can you please verify that each logic chip is receiving 5V?

Double checked - all reading +5.03V.

 

50 minutes ago, TZJB said:

It's a long shot but maybe add an extra ground wire to the board and attach your probe's ground to it?

Ok...now to figure out how...

[dukes909]

  Sorry, had to leave computer for a bit while I was editing above comment. Added more info:

51 minutes ago, TZJB said:

Can you please verify that each logic chip is receiving 5V?

Double checked - all read +5.03V.

51 minutes ago, TZJB said:

It's a long shot but maybe add an extra ground wire to the board and attach your probe's ground to it?

Ok...now to figure out how...

51 minutes ago, TZJB said:

Do you have a way of testing the logic chips independently such as in a XGPRO TL866 II programmer or similar?

NoTL866 or the like here; looking at it, it looks like an EPROM programmer?.. not sure if I bought one how to use it to test chips.

I suppose I could rig up a breadboard test circuit for each?...may take me a while to build a circuit to test each one.  

I am thinking of just buying each of the chips that are still available and replacing all. Other than that, is there any chance that it is simply the drive mechanism / magnetic head and all of these other things are A-OK?

2 hours ago, TZJB said:

Sorry I missed this message. Normally X10 should be used I believe.

OK, good that is how I was checking today. Thanks again.

[TZJB]

11 hours ago, dukes909 said:

  Sorry, had to leave computer for a bit while I was editing above comment. Added more info:

Double checked - all read +5.03V.

Ok...now to figure out how...

NoTL866 or the like here; looking at it, it looks like an EPROM programmer?.. not sure if I bought one how to use it to test chips.

I suppose I could rig up a breadboard test circuit for each?...may take me a while to build a circuit to test each one.  

I am thinking of just buying each of the chips that are still available and replacing all. Other than that, is there any chance that it is simply the drive mechanism / magnetic head and all of these other things are A-OK?

OK, good that is how I was checking today. Thanks again.

 

Voltages seem good at each chip at over 5V.

 

One other thing to test for may be the 4 MHz clock going to A204 pin 5.

 

The metalwork of the Atari 810 is ground so maybe unscrew a bolt to add a ground wire and extend it the Data Separator board ground. It can't do any harm.

 

When you traced the signal from the head it looked good up to the Data Separator board which seems to be the problem here. For some reason I missed the obvious chart in the FSM ver. 3 below.

 

Your A201 pin 6 looks very close to the chart now that you changed the 74LS74, so keep changing chips until the rest look good.

 

A201 pin 5 is the trigger for the lower half of the readings which is the univerted output of A201 pin 6.

 

 

 

Regarding the TL866, it is a very versatile device and has more than one function. It can program devices and can also test certain devices including some SRAM and logic chips, even identifying unlabelled logic chips. The new version is T48.

 

 

[dukes909]

On 7/14/2023 at 4:52 AM, TZJB said:

For some reason I missed the obvious chart in the FSM ver. 3 below.

 

Your A201 pin 6 looks very close to the chart now that you changed the 74LS74, so keep changing chips until the rest look good.

A201 pin 5 is a real challenge to get to. Using pin 5 as a trigger and then any other pin at the same time with 2 scope probes is probably impossible for me to manage with the vertical orientation of the board and the tight proximity of the actual drive mechanism immediately to the right of it.  I may solder up a set of wires to header pins and sockets to be able to remove the board from the vertical position so that I can measure these accurately unless there are other ideas or suggestinons...? (Why are scope probe ground wires so short, BTW?!?) I tried a ribbon cable but the pins on it are not long enough to reach inside the sockets of the sideboard.  

 

On a different note, yesterday, for lack of anything else to try, and unable to check these signals above, I started building simple circuits on breadboards to test each chip. The 74LS04's worked as they should. The 74LS08's however all failed (I have 3 from Atari 810's). Thought I had an extra one somewhere but all I have in stock is a 74HC08. Pin for pin compatible so, with nothing to lose, I tried it. Drive still won't read.......  Not sure if that means another chip is dead somewhere or there is an incompatibility from using 74HC08 instead of the 74LS08's the board came with. Thoughts? I ordered new chips but it will be a little while before they get here (how I miss the ease of buying things like this locally at Radio Shack!)

 

 

Edited July 16, 2023 by dukes909
Additional information posted

[TZJB]

7 minutes ago, dukes909 said:

A201 pin 5 is a real challenge to get to. Using pin 5 as a trigger and then any other pin at the same time with 2 scope probes are probably impossible for me to manage with the vertical orientation of the board and the tight proximity of the actual drive mechanism immediately to the right of it.  I may solder up a set of wires to header pins and sockets to be able to remove the board from the vertical position so that I can measure these accurately unless there are other ideas or suggestinons...? (Why are scope probe ground wires so short, BTW?!?) I tried a ribbon cable but the pins on it are not long enough to reach inside the sockets of the sideboard.  

 

On a different note, yesterday, for lack of anything else to try, and unable to check these signals above, I started building simple circuits on breadboards to test each chip. The 74LS04's worked as they should. The 74LS08's however all failed (I have 3 from Atari 810's). Thought I had an extra one somewhere but all I have in stock is a 74HC08. Pin for pin compatible so, with nothing to lose, I tried it. Drive still won't read.......  Not sure if that means another chip is dead somewhere or there is an incompatibility from using 74HC08 instead of the 74LS08's the board came with. Thoughts? I ordered new chips but it will be a little while before they get here (how I miss the ease of buying things like this locally at Radio Shack!)

 

 

 

I can imagine that a lot of those pins will be difficult to access, but as you have previously deduced that the 74LS74 and more recently that the 74LS08 are faulty, hopefully the replacements will work as expected and you won't need to access them.

 

Regarding oscilloscope ground probe leads, I agree they are too short. I extended mine with a 25cm double ended crocodile lead.

 

That is good that you are testing your logic IC's. It would be easier in a T48 programmer however, and it would also allow you to make an Archiver ROM later when you have fixed the Atari 810.

 

Unfortunately a 74HC part is only CMOS compatible and is not directly TTL compatible without additional pullup resistors which would mess things up even more.

 

Long ago we had Radio Shack on our high streets too in the form of Tandy, who were the UK version of Radio Shack selling most of their products, including components in neat packages.

 

We also had Maplin in Essex who sold electronics and components mainly mail order, and were a major Atari 8-bit proponent selling the full Atari range from an early age. They later expanded to the high street continuing to sell electronics and components.

 

Unfortunately both went bust at different times and closed all the shops...so everything is now mail order, just over the Internet. Tandy and Maplin are back online for orders here in the UK. Not so with Radio Shack it seems.

 

 

 

 

 

 

 

 

 

[dukes909]

2 minutes ago, TZJB said:

 

I can imagine that a lot of those pins will be difficult to access, but as you have previously deduced that the 74LS74 and more recently that the 74LS08 are faulty, hopefully the replacements will work as expected and you won't need to access them.

 

Regarding oscilloscope ground probe leads, I agree they are too short. I extended mine with a 25cm double ended crocodile lead.

 

That is good that you are testing your logic IC's. It would be easier in a T48 programmer however, and it would also allow you to make an Archiver ROM later when you have fixed the Atari 810.

 

Unfortunately a 74HC part is only CMOS compatible and is not directly TTL compatible without additional pullup resistors which would mess things up even more.

 

Long ago we had Radio Shack on our high streets too in the form of Tandy, who were the UK version of Radio Shack selling most of their products, including components in neat packages.

 

We also had Maplin in Essex who sold electronics and components mainly mail order, and were a major Atari 8-bit proponent selling the full Atari range from an early age. They later expanded to the high street continuing to sell electronics and components.

 

Unfortunately both went bust at different times and closed all the shops...so everything is now mail order, just over the Internet. Tandy and Maplin are back online for orders here in the UK. Not so with Radio Shack it seems.

I thought the HC version was not compatible voltage-wise 100% but thought I'd give it a go anyway. Thanks for that information (probably should have asked before trying it though 😛).

 

Looks like all I can do is sit back and wait on the new parts and maybe build a few more test circuits just for the practice.  I thought about extending the ground on the scope probe ground with just such a lead but then second guessed myself that they must have made it that short for a reason and didn't want to buy new probes and/or scope due to my ignorance! 

 

Thanks again for the help. If I come across a 74LS08 in anything laying around that I can verify works and then test it in the sideboard etc I will post updates. Once I get the new chips in I will post an update as well.

I asked before but not sure anyone replied: is it entirely possible (likely) the drive mechanism/head is faulty  and causing these problems?

 

Will buy you a pint or your choice the next time I'm over there!
Cheers

[TZJB]

23 hours ago, dukes909 said:

I thought the HC version was not compatible voltage-wise 100% but thought I'd give it a go anyway.

 

I thought about extending the ground on the scope probe ground with just such a lead

 

I asked before but not sure anyone replied: is it entirely possible (likely) the drive mechanism/head is faulty  and causing these problems?

 

Will buy you a pint or your choice the next time I'm over there!
Cheers

 

You can't harm anything with an HC substitute, it just probably won't work without the pull-up resistors. The datasheet supplies the details.

 

The oscilloscope probe ground lead is short to reduce inductance to prevent 'ringing' or overshoot, but if it doesn't reach it's not going to work anyway. Ideally it would be just a short piece of wire, but for our purposes at a few MHz, not that important.

 

The mechanism and drive head seem to be behaving correctly at this point, even erasing the disk as expected during a format.

 

You proved the Erase head when you formatted a disk and it erased it, and proved the Read/Write head when you traced the Read Data signal through to the Data Separator Board where it failed.

 

It would be nice to meet when you come to London so let me know!

Edited July 17, 2023 by TZJB

[dukes909]

On 7/17/2023 at 2:27 PM, TZJB said:

 

You can't harm anything with an HC substitute, it just probably won't work without the pull-up resistors. The datasheet supplies the details.

 

The oscilloscope probe ground lead is short to reduce inductance to prevent 'ringing' or overshoot, but if it doesn't reach it's not going to work anyway. Ideally it would be just a short piece of wire, but for our purposes at a few MHz, not that important.

 

The mechanism and drive head seem to be behaving correctly at this point, even erasing the disk as expected during a format.

 

You proved the Erase head when you formatted a disk and it erased it, and proved the Read/Write head when you traced the Read Data signal through to the Data Separator Board where it failed.

 

It would be nice to meet when you come to London so let me know!

I replaced all the logic chips on the data separator board with new ones:

1 x 74LS04

1 x 74LS08

2 x 74LS74

2 x 74LS193

No dice. Drive still won't read the disk. Same exact results as before. I haven't checked the signals with the scope but do I need to? So frustrating. 😞

[_The Doctor__]

did you have another 810 to simply swap mech and boards whole sale to limit the destruction/reconstruction to the problem board?

 

It'll be one of the SIO buffer/selection chips after all of this in combo with a cap or transistor at the outset as a cherry on top, just for kicks if the universe is having fun!

[dukes909]

59 minutes ago, _The Doctor__ said:

did you have another 810 to simply swap mech and boards whole sale to limit the destruction/reconstruction to the problem board?

Do you mean a known working 810 with identical parts? If so, no. I have another 810 that doesn't work either that I have swapped parts with but it has made no difference. I have been watching eBay for a verified, working 810 but nothing so far.

 

1 hour ago, _The Doctor__ said:

It'll be one of the SIO buffer/selection chips after all of this in combo with a cap or transistor at the outset as a cherry on top, just for kicks if the universe is having fun

Which  board are those chips on?

[TZJB]

On 7/20/2023 at 12:06 AM, dukes909 said:

I replaced all the logic chips on the data separator board with new ones:

1 x 74LS04

1 x 74LS08

2 x 74LS74

2 x 74LS193

No dice. Drive still won't read the disk. Same exact results as before. I haven't checked the signals with the scope but do I need to? So frustrating. 😞

 

Please check waveforms again on the Data Separator Board as they should be getting to A203 1771 FDC pins 26 & 27 via A205 pins 3 and 6 if the logic chips are working correctly.

 

On 7/20/2023 at 1:29 AM, dukes909 said:

Which  board are those chips on?

 

The read data path seems to go in and out of the Rear Board J102 & J106 connectors to finally end up at Rear Board J107 & J108 the actual SIO connectors.

 

The SIO buffer transistors are on the Side Board and are in A106 but there is also a Q105 on the Data Out.

 

[dukes909]

On 7/22/2023 at 5:00 AM, TZJB said:

A203 1771 FDC pins 26 & 27 via A205 pins 3 and 6

All these new chips are from Mouser, so hopefully not counterfeit ones from them. 🤔

 

First A205 pin 6

and matching A203 FDC pin 26

Then A205 pin 3

and matching pin A203 FDC pin 27

I also checked  A208 pin 9 which also feeds into A203 pin 27:

and A206 pin 5 that goes to A203 pin 26.

I tried 2 different A208 hex inverter chips (new, from Mouser), with same results as well as swapping the A203 with a couple of different ones. All looked the same on 26 & 27

On 7/22/2023 at 5:00 AM, TZJB said:

The read data path seems to go in and out of the Rear Board J102 & J106 connectors to finally end up at Rear Board J107 & J108 the actual SIO connectors.

 

The SIO buffer transistors are on the Side Board and are in A106 but there is also a Q105 on the Data Out.

 

Not sure about all this. Should I be checking these or getting the data separator board looking right first?

Edited July 23, 2023 by dukes909

[TZJB]

On 7/23/2023 at 11:25 PM, dukes909 said:

All these new chips are from Mouser, so hopefully not counterfeit ones from them. 🤔

 

First A205 pin 6

and matching A203 FDC pin 26

Then A205 pin 3

and matching pin A203 FDC pin 27

I also checked  A208 pin 9 which also feeds into A203 pin 27:

and A206 pin 5 that goes to A203 pin 26.

I tried 2 different A208 hex inverter chips (new, from Mouser), with same results as well as swapping the A203 with a couple of different ones. All looked the same on 26 & 27

Not sure about all this. Should I be checking these or getting the data separator board looking right first?

 

The Data Separator Board needs to work correctly before proceeding to look elsewhere but these chips are on that board.

 

A205 pin 6 should have the same waveform as A203 pin 26 so there is a disconnect between them. The same is true for A205 pin 3. It might be easier to remove the board and proceed with an Ohmmeter in order to trace the tracks. Maybe the sockets need cleaning or replacing?

[dukes909]

21 hours ago, TZJB said:

The Data Separator Board needs to work correctly before proceeding to look elsewhere but these chips are on that board.

Ok

21 hours ago, TZJB said:

A205 pin 6 should have the same waveform as A203 pin 26 so there is a disconnect between them. The same is true for A205 pin 3. It might be easier to remove the board and proceed with an Ohmmeter in order to trace the tracks. Maybe the sockets need cleaning or replacing?

Pin to pin for both pairs, I get perfect continuity (well, 0.001 ohms) according to DMM. I can't make sense of it. 🤨

[dukes909]

I have rigged up a set of header cables from sideboard to power board that allows me to lay the sideboard down flat. I will go through to make sure I am checking the right pins one at a time. Stay tuned.

[dukes909]

2 hours ago, dukes909 said:

I have rigged up a set of header cables from sideboard to power board that allows me to lay the sideboard down flat. I will go through to make sure I am checking the right pins one at a time. Stay tuned.

EDIT: I spoke too soon. I suppose I can't use a set of 6" header cables to connect from the power board to side board due to some length limitation? Anyone?  With the header cable connected the drive will spin up and the LED goes on and off like it is supposed to but just sits there when I try to do a directory from DOS 2.0 then ERROR 138.

 

It is such a pain to try and get readings off the chips when the sideboard is mounted vertically in the drive. How did technicians back in the day do it?

[TZJB]

On 7/27/2023 at 5:45 PM, dukes909 said:

Ok

Pin to pin for both pairs, I get perfect continuity (well, 0.001 ohms) according to DMM. I can't make sense of it. 🤨

 

Is the continuity to both sides of the board, just in case the two sides have separated somehow?

[TZJB]

On 7/27/2023 at 11:47 PM, dukes909 said:

EDIT: I spoke too soon. I suppose I can't use a set of 6" header cables to connect from the power board to side board due to some length limitation? Anyone?  With the header cable connected the drive will spin up and the LED goes on and off like it is supposed to but just sits there when I try to do a directory from DOS 2.0 then ERROR 138.

 

It is such a pain to try and get readings off the chips when the sideboard is mounted vertically in the drive. How did technicians back in the day do it?

 

That seems a reasonable method of faulting the board. Do you have a nice thick ground wire connecting it to the main chassis though?

[dukes909]

8 hours ago, TZJB said:

 

Is the continuity to both sides of the board, just in case the two sides have separated somehow?

As near as I can tell. It's a lot of pins to check. There were tight going on. I have 5V on all of the chips on the sideboard.

EDIT: Checked all 30 connections + ground connections with 14 ga. ground wire. All good. Device powers on but will not take commands from 800.

Edited July 29, 2023 by dukes909
Additional information posted

[dukes909]

7 hours ago, TZJB said:

 

That seems a reasonable method of faulting the board. Do you have a nice thick ground wire connecting it to the main chassis though?

A 14 gauge between chassis and sideboard.

[TZJB]

19 hours ago, dukes909 said:

As near as I can tell. It's a lot of pins to check. There were tight going on. I have 5V on all of the chips on the sideboard.

EDIT: Checked all 30 connections + ground connections with 14 ga. ground wire. All good. Device powers on but will not take commands from 800.

 

There must be too much interference on the extension wires I suspect. 14 gauge is about 2mm squared so seems suitable for a ground wire. Would it be possible to screen the extension wires? And does removing the extension restore expected behaviour?