Atari 1050 UDN5713M chips question

[_The Doctor__]

I'm just impressed by the fact you were able to break the pins for the mechanisms plugs...hope nothing is damaged inside the connector.

also when plugging them back in are they all facing the same direction? certain models have one plug rotated/reversed

 

Edited March 1, 2021 by _The Doctor__

[+DrVenkman]

38 minutes ago, kbj said:

The 6532 I have is from a working 2600 - which works fine - controllers and all.

Bear in mind, the 6532 can fail in subtle ways that aren't immediately obvious - I've encountered two bad ones from 4-switch 2600's. In each case, the failure was ONE difficulty switch was not working despite the switch itself being fine. The first time I ran into it I had just purchased a 4-Switch Woody that I was using as a daily driver I was mostly playing later games that didn't always use the difficulty switches so I failed to notice the issue. The first time I tried to play Space Invaders, I realized I was stuck with the "A" difficulty "fat base" and it wouldn't go back to the normal, narrower shape for "B." Player 2 was fine. I cleaned the switch which had no effect, then took the unit entirely apart and verified continuity for the switch contacts and PCB traces. That's when I suspected a bad RIOT. Replaced the chip and problem solved. About a year later I picked up a Vader super cheap and it had the same failure, but it was the Player 2 switch. In each case everything else was working fine.

 

So the point being, the 6532 has a bunch of different functions - "working" may mean functional in a 2600, or even in some 2600 games but not others, but still fail when used in a 1050. If you have a second chip handy, it would be good to test. But as Nezgar points out too, the original sockets aren't necessarily the best quality though they're usually okay. But inserting and removing chips a couple times might be enough to cause contact issues on one or more pins as well. 

[+Nezgar]

Some further testing with a 1050 here produced the following findings:

• powering up without a 6532 present results in continuous motor-on
• powering up without a 279x controller present results in the 2 second motor on/off pattern
• powering up without CPU or 6810 present = completely unresponsive

So @kbj based on your current symptoms of "always 2 second on/off pattern," I would prioritize your efforts on the WDC controller socket first.

[kbj]

13 hours ago, _The Doctor__ said:

I'm just impressed by the fact you were able to break the pins for the mechanisms plugs...hope nothing is damaged inside the connector.

also when plugging them back in are they all facing the same direction? certain models have one plug rotated/reversed

 

Haha! I was impressed myself!  I think that pulling J15 off and on so many times, I bent the connector at one point and think I just stressed one of the pins..

 

Ah well.. ?

[+kheller2]

Maybe I missed it but has someone verified that the 2 second motor on/off pattern occurs when you pull the stepper connection?  I didn't think the 1050 OS checked for a faulty stepper on POST.  

 

Or does it...

 

PORT A

PORT B

FDC

PULL Head back

ROM Checksum

RAM TEST

 

If any of those fails, it does the 2 second warning. It would have been nice if it varied the cycle by the condition.

 

 

[+Nezgar]

1 hour ago, kheller2 said:

Maybe I missed it but has someone verified that the 2 second motor on/off pattern occurs when you pull the stepper connection?  I didn't think the 1050 OS checked for a faulty stepper on POST.

@kheller2 Yes, you missed that I tested that in post #30. The POST succeeds in my drive with stepper detached, as long as head is positioned at track 0. It also succeeds if the track 0 sensor is disconnected at J10, making it think the head is always at track 0.

 

1 hour ago, kheller2 said:

PORT A

PORT B

FDC

PULL Head back

ROM Checksum

RAM TEST

Awesome, thanks for digging out that list. Interesting - so, since we confirmed his drive ALWAYS does the 2 second on/off regardless of head position (even with J10 disconnected) it must be failing at the FDC test. This also corroborates my process of elimination that pointed at the WDC in post #53. Failing at the FDC check would also explain why the head never pulls back when not at track 0.

 

Since @kbj has tested known good FDC chips in this drive, it would suggest a connection issue with the socket, or possibly bad solder point, trace or other nearby component.

[kbj]

15 minutes ago, Nezgar said:

@kheller2 Yes, you missed that I tested that in post #30. The POST succeeds in my drive with stepper detached, as long as head is positioned at track 0. It also succeeds if the track 0 sensor is disconnected at J10, making it think the head is always at track 0.

 

Awesome, thanks for digging out that list. Interesting - so, since we confirmed his drive ALWAYS does the 2 second on/off regardless of head position (even with J10 disconnected) it must be failing at the FDC test. This also corroborates my process of elimination that pointed at the WDC in post #53. Failing at the FDC check would also explain why the head never pulls back when not at track 0.

 

Since @kbj has tested known good FDC chips in this drive, it would suggest a connection issue with the socket, or possibly bad solder point, trace or other nearby component.

Hey guys,

 

I didn't have time to post earlier as it's been a busy day with other repairs..

 

I changed the FDC socket and still the same issue.  I tried 2 other FDC chips that I just had delivered.. same issue.. what's the chance of that.. 

 

I did however have a few cold solder joints (one on the crystal osc which actually was so bad on one pin, I was able to pull it out of the board..)  resoldered into the board and the clock still checks out at 1Mhz.

 

Still got good activity on the EPROM/CPU and SRAM.. data and address pins all look like they're strobing nicely.

 

The only thing that looks wrong is the outputs of the 6532 to the 2 UDN chips.. they just have noise on them.. I put the 6532 into a 2600 with a diag cart and harness and all tests past fine so I assume that the 6532P I have is good.  Same with the 6507 - put it in the same 2600 and all tests past.

 

I am waiting for a 'supposed to be' drop in replacement for the track 0 sensor - but the noise on the outputs of the 6532 are bugging me.. I don't see the same waveforms that I should see according to the service manual on either the UDN or 6532..

 

11.33pm here so think I'll sleep on it and see tomorrow..

 

I really appreciate all of the help guys,

 

Mark.

[+kheller2]

1 hour ago, Nezgar said:

 

Awesome, thanks for digging out that list. Interesting - so, since we confirmed his drive ALWAYS does the 2 second on/off regardless of head position (even with J10 disconnected) it must be failing at the FDC test. This also corroborates my process of elimination that pointed at the WDC in post #53. Failing at the FDC check would also explain why the head never pulls back when not at track 0.

 

Since @kbj has tested known good FDC chips in this drive, it would suggest a connection issue with the socket, or possibly bad solder point, trace or other nearby component.

it has been a while since i dug into the source code on the 1050, and the service guide for the step feels like it contradicts what the code is doing.. but I'm rusty.  There is a pull back test and if it doesn't find T0 after so many steps it bombs.

 

Also, Mark did say that the stepper did do the stroke thing once.

 

Mark where are you located? I almost feel like sending you parts to swap around to end this painful fix.

 

[kbj]

8 hours ago, kheller2 said:

it has been a while since i dug into the source code on the 1050, and the service guide for the step feels like it contradicts what the code is doing.. but I'm rusty.  There is a pull back test and if it doesn't find T0 after so many steps it bombs.

 

Also, Mark did say that the stepper did do the stroke thing once.

 

Mark where are you located? I almost feel like sending you parts to swap around to end this painful fix.

 

Hi Karl,

 

The head made a fast and quite loud 'vibration' one time (that's the only way I can describe it, a very fast and short back and forward at speed) for about a couple of secs until I powered off the drive. It didn't do it again - that's why I thought it may be the UDN chips or the stepper..

 

I'm in Scotland, UK I'm afraid - our respective governments' 'special relationship' doesn't extend to good shipping fees ? or I would love to pop parts in this thing and get it working again!

 

I'm still unsure about the stepper though - when J15 is unplugged and I measure as @Nezgar asked me to, the windings appear open - but when J15 is plugged in, the windings appear to read normally... I find that bizarre..

 

Thanks again ?

 

Mark.

[+kheller2]

If you unplug the stepper and track zero, on a good drive I think it was mentioned it would pass POST. 

[kbj]

18 hours ago, kheller2 said:

If you unplug the stepper and track zero, on a good drive I think it was mentioned it would pass POST. 

That's right.. and mine just keeps doing the same 2sec off/on.. So I may have a problem with logic somewhere else on the motherboard then?

[+kheller2]

 

9 hours ago, kbj said:

That's right.. and mine just keeps doing the same 2sec off/on.. So I may have a problem with logic somewhere else on the motherboard then?

yes.    This isn't a modified drive correct? It's just a stock Tandon mech unit?

The failure is probably with one of these chips:

 

6532

279x

ROM - (Checksum failure, I doubt that, but you should probably put the original ROM back in)

RAM

[kbj]

25 minutes ago, kheller2 said:

 

yes.    This isn't a modified drive correct? It's just a stock Tandon mech unit?

The failure is probably with one of these chips:

 

6532

279x

ROM - (Checksum failure, I doubt that, but you should probably put the original ROM back in)

RAM

Hi @kheller2

 

Nope, it's a stock Tandon unit.  

 

I've tested both the 6532P and the 6507 in a 2600 with the diag cart and harness and all tests pass.

 

I've done a binary compare of the EPROM with bin dumps from the same rev L EPROM from 2 other sources (working drives) and it's good.

 

I've checked the 6810 in my chiptester and the RAM is good (multiple times) but just in case, tried another 6810 and same issues.

 

I've replaced U11 as that's in circuit and no difference.

 

Even if I remove J10, the stepper still won't move and I get the same 2sec on/off...  I did get another track 0 sensor today and will dump that in, although I don't think that's the issue here either.

 

All the data and address lines on RAM/EPROM/CPU look normal to me.  The only issue is the pin outputs from the 6532 to the stepper circuit - they are just noise.. I would expect pulses of some sort?

 

I'm still drawn to the fact that removing J15 and measuring the windings in the stepper from the plug, I get open circuits (where Nezgar provided me with measurements from his working drive) but if I plug J15 back in and measure from it, I get correct readings... really boggling my brain that one..

 

Cheers,

Mark.

[+kheller2]

I would expect pulses too... if it got that far into the code, which it isn't.

I'm also assuming that the track zero sensor, is always showing continuity unless broken by the head (which would explain why unplugging it lies to the OS).

 

This makes me want to re-write some of the 1050 code and have a diagnostic ROM to flash the activity light as to what component is at fault.   Maybe some day.

 

I'll toy with one of my 1050s tonight during a hockey game to see if what has been stated here is 100% accurate.

 

You do have the mech plugged in correctly (remember, one of the connectors is backwards from the other for a Tandon)?

 

Perhaps a picture of the top and bottom of your board too.

 

[_The Doctor__]

also male sure the rom is a tandon rom...

I had a guy ship me his drive that was broken and he had a world storage rom with a tandon mech... that didn't work too well...

[kbj]

21 minutes ago, kheller2 said:

I would expect pulses too... if it got that far into the code, which it isn't.

I'm also assuming that the track zero sensor, is always showing continuity unless broken by the head (which would explain why unplugging it lies to the OS).

 

This makes me want to re-write some of the 1050 code and have a diagnostic ROM to flash the activity light as to what component is at fault.   Maybe some day.

 

I'll toy with one of my 1050s tonight during a hockey game to see if what has been stated here is 100% accurate.

 

You do have the mech plugged in correctly (remember, one of the connectors is backwards from the other for a Tandon)?

 

Perhaps a picture of the top and bottom of your board too.

 

Thanks for the reply ? 

 

The connectors are plugged in correctly (I hope!) ?  I've attached some pics - hope they're ok and of use.. let me know if you need further pics - I really appreciate the assist with this!

 

The EPROM is a copy of the original Tandon REV L ROM - but it binary compares with 2 other source EPROMs from working drives..

 

The diag rom idea sounds very cool!

 

Mark.

 

And I just noticed the dry solder joint on the bottom of the PCB just now.. resoldered it and no joy ?

 

Edited March 4, 2021 by kbj
Additional info

[kbj]

5 minutes ago, _The Doctor__ said:

also male sure the rom is a tandon rom...

I had a guy ship me his drive that was broken and he had a world storage rom with a tandon mech... that didn't work too well...

Hi @_The Doctor__

 

Yup - definitely the right Tandon Rev L EPROM..

 

Mark.

[_The Doctor__]

check Q4 for shorted legs, it looks pinched sideways and over. it's an accident waiting to happen in any case

Edited March 4, 2021 by _The Doctor__

[kbj]

14 minutes ago, _The Doctor__ said:

check Q4 for shorted legs, it looks pinched sideways and over.

I did look at both those transistors in that part of the circuit but they looked ok.  I've taken them out, tested them (ok) and put them back in - same issue.

 

Good spot though!

 

Mark.

[_The Doctor__]

setting the proper jumpers for the two different styles of (P)ROM or EPROM was a thing also, since I see an eprom make sure the jumpers are for tandon eprom... not mask rom perhaps? I am just taking a stab here as I don't remember the configs for the differing roms.. but I know it matters for jp1-jp7?

[+kheller2]

I don't see pin 25 out of the socket on the FDC.  (I'm going to get hit for that).

So, have you put all the original chips back in, and you don't have another 1050 right?

 

[kbj]

10 hours ago, kheller2 said:

I don't see pin 25 out of the socket on the FDC.  (I'm going to get hit for that).

So, have you put all the original chips back in, and you don't have another 1050 right?

 

Hey @kheller2 Pin 25 is out of the socket - it's the same chip that was in it originally.  Tried 2 others that I bought, bent pin 25 out the socket for each and no joy.

 

I have put all of the original chips back in and still the same issue.  At the moment the only non-original chips are the ERPOM (copy of the Tandon EROM that was in it) and the 6532.

 

Cheers,

 

Mark.

Edited March 5, 2021 by kbj

[kbj]

11 hours ago, _The Doctor__ said:

setting the proper jumpers for the two different styles of (P)ROM or EPROM was a thing also, since I see an eprom make sure the jumpers are for tandon eprom... not mask rom perhaps? I am just taking a stab here as I don't remember the configs for the differing roms.. but I know it matters for jp1-jp7?

It came with the stock EPROM from Tandon (Rev L) and I haven't changed the jumpers - as I'm pretty sure they should be okay as the drive worked at some time in the past until recently.  I guess someone could have had it to look at before me and stuck a tandon EPROM in it.. I'll check out the links just to be sure.

 

Cheers,

 

Mark.

[kbj]

Hey All,

 

Just an update on this 1050 drive issue.

 

I had another look at the logic board with a scope and noticed that when switching the drive on and off and scoping the address pins on the ROM that sometimes I would see good data for a few seconds and then it all would stall and it looked like the board had crashed

 

I checked the EPROM again in my programmer (read the contents and compared to a known good EPROM) and all looked good.  Just in case, I wrote another EPROM with data from the good file and plugged it back in to the board and it worked!  Drive is working fine..

 

Looks like I may have had a couple of bad 2732A EPROMS... (and a WDC/6532)

 

I took the EPROM that I had burned originally and wrote a batch file to continuously read its contents and compare it to the known working ROM dump - and about 2-3 times out of 5 it had errors.. damn!  

 

In the end, I replaced the WDC chip, 6532, EPROM, 2 sockets and did a recap but now the drive is working very sweetly!

 

I really can't thank you all enough for helping me troubleshoot this little pig (hehe!) and for the amount of knowledge you've all helped me gain throughout this ?

 

Cheers!

 

Mark.

[TGB1718]

Glad you finally cracked this nut