Jump to content
IGNORED

Atari 1050 UDN5713M chips question


kbj

Recommended Posts

Hi all,

 

I have a 1050 disk drive that has no stepper motor drive at all.  I've done a swap out of chips, replaced the caps and checked the diodes between the 3 large caps and they're all fine.  I've checked the windings of the stepper motor for the correct resistance as per the service manual and all appears good.  I've checked the voltages on the J15 connector - all 12v. The drive motor spins and stops then repeats itself but the stepper never moves the head at all.

 

I'm suspecting that one or both of the UDN5713N chips is bad now.  Would that be a correct assumption?

 

I was also told that it may be possible to replace these chips with a SN75478 - does that sound like a flyer?

 

Alternatively, does anyone have a source of the UDN5713N chips?

 

Appreciate any insight!

 

Cheers,

 

KbJ

Link to comment
Share on other sites

Hi TGB1718,

 

Thanks for the info - I checked pin 10 on U11 - nothing at all - 0v.  It doesn't change at all even at switch on.  Replaced with a new 7402 but still getting the same issue.  One thing that did happen is the stepper motor briefly moved the head back and forward very quickly - like it vibrated the head - momentarily and then stopped.  I now can't get it do do it again..

 

This drive is really kicking my a** today ?

 

Link to comment
Share on other sites

2 hours ago, kbj said:

The drive motor spins and stops then repeats itself but the stepper never moves the head at all.

 

This is something like a "Windows bluescreen" from that era ? - When the drive starts spinning for 1-2 seconds, then 5-10 seconds off and repeat, the firmware signals a communication error this way. Mostly the FDC chip is defect, sometimes the RIOT (6532) or the 6810 SRAM chip.

 

But first - remove all five chips in sockets, clean the pins with a fiber-glass pencil, clean the sockets with an old toothbrush, use compressed air to clean all over. Bend the pins of the chips slights to the middle of the chips to have more  pressure to the ultra-low-cost sockets Atari used. In 7 of 10 cases you´re done and the drive will work normal after that.

 

  • Like 2
Link to comment
Share on other sites

Thanks mate! ?

 

I've changed the RIOT and the 6507 (with a UM6507 that I know works).  I've also checked the 6810 SRAM and that appears functional too.  It strange though.. when I check the address and data lines on the 6507, sometimes on power up they are obviously just noise and nothing appears to be running code - and when this happens the motor just spins constantly.  Other times, when the drive is doing 2 second spins then stop and repeating, the address and data lines all look normal and look as if code is running - have you seen that before?

 

I hope it's not the controller chip - can't find one of those at all - and I don't have a spare to try it with ?

 

I'll give the chips and sockets a clean as you say and report back.

 

Take care,

 

Mark.

Link to comment
Share on other sites

Ok, so cleaned all of the chips with a fiberglass pencil and cleaned the sockets with IPA and a toothbrush - no joy.  Burned a new Eprom with a copy of the Tandon REV L ROM which is what the original ROM marking said (but to be honest, read the original ROM and compared it with another copy and it was good).

 

All voltages are steady (+5v line is +5.01v and the +12v is +11.94v).

 

I've tried another RIOT in it - the original part had an Atari C0xxxxx number but my replacement part has the markings:

 

R6532GD

R6532-31

Mexico K3

8845 5941

 

Are there differences in the 6532 parts?  I use these to repair 2600 consoles..

 

I guess that it's looking like the FDC then?  Or the stepper?

 

One other thing I thought - I'm not using the original 9VAC 31VA psu (because I don't have one) but I do have a 9VAC 27VA adaptor - surely 4VA difference is not enough to stop the stepper getting enough current to work?

 

Cheers,

 

Mark. 

Edited by kbj
  • Like 1
Link to comment
Share on other sites

1 hour ago, kbj said:

surely 4VA difference is not enough to stop the stepper getting enough current to work?

 

If the DC voltages are steady all the time, then your Power Supply is adequate.

 

one more thing to try, is pin 3 of U4 "clocking" , it provides the input to U11 pins 8 and 9 ?

it's a nice cheap NE555 timer chip.

 

Don't know where your located, but a quick look on Fleabay :) and there's quite a few in the UK

Edited by TGB1718
  • Like 1
Link to comment
Share on other sites

Hey TGB1718,

 

Thanks for the info - I'll check U4 - that's for the reset circuit for the 6507?  If that's all it's used for, Pin 3 goes high (+4.8v) at switch on.

 

The PSU voltages are completely steady - even when the motor spins for the couple of seconds, then repeats on and off.

 

I've checked address and datalines - they all look good - strobing away nicely so it looks like the CPU/Memory/6532 are all addressing data and running code from the ROM.  The inputs into the 2 driver chips from the 6532 look just like noise with a bit of a DC component.. according to the service manual there should be a waveform on them but nothing at all.. makes me suspect that the UDN5713's are toasted?  If the stepper had gotten stuck, is there a chance that it would have killed one or both of them do you think?

 

Cheers,


Mark.

Edited by kbj
Link to comment
Share on other sites

Anything happening on Pins 2 and 7 on both U2 and U3, if they are in sockets, you could pop them out and check

the pins on powerup to see if anything's coming out of the 6532 pins 18,19,21 and 22 

 

btw. my bad, U4 is the power-on reset circuit (need to look more closely :) )

 

Edited by TGB1718
Link to comment
Share on other sites

15 hours ago, kbj said:

I've tried another RIOT in it - the original part had an Atari C0xxxxx number but my replacement part has the markings:

 

Are there differences in the 6532 parts?  I use these to repair 2600 consoles..

 

No, you can take them. 6532 are both the same in 1050 floppy and the 2600 console. The ones in the 2600 are graded for 2 MHz, the ones in the 1050 for 1 MHz only, but mostly it works also fine ?

 

Quote

I guess that it's looking like the FDC then?  Or the stepper?

 

Of course the stepper might be also a problem, but from my personal experience repairing these drives it´s the FDC 2793/2797. I´ve changed over 20 pieces the last years, but only 2 step motors.

 

Quote

One other thing I thought - I'm not using the original 9VAC 31VA psu (because I don't have one) but I do have a 9VAC 27VA adaptor - surely 4VA difference is not enough to stop the stepper getting enough current to work?

 

Absolutely not, the 31VA isn´t really needed. Atari make one power supply fit for several devices like the diskdrive 1050, 400/800 computers and some printers. The computers and the printers need - during start-up phase - much more current than the 1050. It´s fine to use the 27VA PSU.

 

Link to comment
Share on other sites

1 hour ago, kbj said:

The PSU voltages are completely steady - even when the motor spins for the couple of seconds, then repeats on and off.

 

I've checked address and datalines - they all look good - strobing away nicely so it looks like the CPU/Memory/6532 are all addressing data and running code from the ROM.

 

It´s simple: If the drives does these spin-on/spin-off used as an error-sign, then CPU and RAM at least must be ok. Also the NE555, which generates a delayed reset phase, is ok. You can exclude these parts.

 

Link to comment
Share on other sites

5 hours ago, kbj said:

I hope it's not the controller chip - can't find one of those at all

There are a few 2793 FDC available on eBay, whether they are functional is unknown. Also available from Best Electronics for $15.95 but there is a minimum order and Brad can be difficult to deal with. 

https://www.ebay.com/sch/i.html?_from=R40&_trksid=p2334524.m570.l1313&_nkw=%2B2793+dip40&_sacat=92078&LH_TitleDesc=0&_osacat=92078&_odkw=2793+dip40

 

Something to be aware of is that the drive should be recalibrated when the FDC is replaced.

 

Edited by BillC
  • Like 1
Link to comment
Share on other sites

Thanks for all of the info guys - it's really appreciated.

 

So, I guess it looks like its down to the 2 UDN 8 pin chips or the FDC then.  Can anyone tell me if I can replace the UDN5713 chips with (the easier to get) SN75478 chips?  I'd read in another forum that they were compatible replacements?

 

Thanks!


Mark.

Link to comment
Share on other sites

11 minutes ago, kbj said:

Can anyone tell me if I can replace the UDN5713 chips with (the easier to get) SN75478 chips?

Had a look at the datasheets and it does look like it's a direct replacement

https://pdf1.alldatasheet.com/datasheet-pdf/view/55137/ALLEGRO/UDN5713.html

https://www.ti.com/lit/ds/symlink/sn75478.pdf?ts=1613728346663&ref_url=https%3A%2F%2Fwww.ti.com%2Fproduct%2FSN75478

Link to comment
Share on other sites

I looked at those same datasheets myself just a moment ago - they do look like direct replacements don't they?

 

Thanks for getting back to me, I think I'll buy a couple and put them in sockets and see what happens.  I will remove them though and check activity on the RIOT.. maybe those little chips are internally shorted and are pulling the signals down.

 

Cheers!


Mark.

Link to comment
Share on other sites

Have you checked the service manual?  The busy light and motor cycling on and off is a “POST” failure as has been mentioned.  (And why is it called post? Pre would be better). The OS is detecting a hardware failure.  Yes I’m stating the obvious so don’t throw things at me. ? if you’ve replaced most of the big chips it’s probably the FDC. 

Link to comment
Share on other sites

51 minutes ago, kheller2 said:

The busy light and motor cycling on and off is a “POST” failure as has been mentioned.  (And why is it called post? Pre would be better).

I presume that’s a joke. :) But for anyone who legitimately doesn’t know, POST = Power-On System Test. 

  • Like 2
Link to comment
Share on other sites

So, I desoldered the 2 UDN5713's, socketed them and replaced them with 2 x SN75478 chips.  Same issue.  Only this time the 4 outputs to those chips from the 6532 are sitting high continuously (about +4.5v) with no stepper signal on them at all.  Replaced the 6532 (thinking I may really have a bad one now) and same issue.. Really is beginning to look like the FDC chip ?  Anyone know where I can get one for a decent price?  The only one I can find is £24 ?

 

I probably should also say that before putting the SN75478 chips in, I measured the 4 outputs from the 6532 and they were all the same (high, +4.5v)..

 

Everything else I checked on the CPU/RAM side appear to have nice pulsing lines (all data and address etc).  Reset line is working etc.

 

Thanks guys,

 

Mark. 

Link to comment
Share on other sites

Ahh! I didn't see those when I looked yesterday! Thanks!  I'll go buy one just now.

 

I guess that since my original one has pin 25 snipped, I'll do the same with this one (just bend the pin up out of the socket)?

 

Cheers,

 

Mark.

Link to comment
Share on other sites

On 2/18/2021 at 8:23 AM, kbj said:

I've also checked the 6810 SRAM and that appears functional too.

How did you check the 6810? The best test would be to try it in another 1050... checking in an EPROM programmer may not "stress" it the same way. I believe bad 6810 caused drive motor cycling on/off (post failure)

 

Identifying a problem with the stepper would also be quickly isolated by swapping mechs with another 1050.

 

If it really is the 279x chip definitely consider the calibration that has to be done as tf_hh mentioned - a fault with that would be a first for me, but I have had multiple 6810 failures... (and 6116 / 6264 type SRAMs in happy ehnacements too) (in 20+ drives...)

Link to comment
Share on other sites

57 minutes ago, kbj said:

I guess that since my original one has pin 25 snipped, I'll do the same with this one (just bend the pin up out of the socket)?

18 minutes ago, TGB1718 said:

Not sure why it's snipped, but it's not used anyway, it's the Side Select pin for drives with 2 heads,

the circuit diagram shows it going to a 1K pullup (R46)

All of the 1050's I've seen that had 2797 controllers all had pin 25 snipped. If it was me i would just bend out the pin on a new chip vs snipping it off too... Some interesting background on Pin 25 on the 2797 in this thread:

 

Link to comment
Share on other sites

@TGB1718 Thanks for the info - I saw a different post (think it was on this forum) about 25 being cut from the factory but wasn't sure.  Thanks for confirming ?

 

@Nezgar I've a chip tester Pro from the 8-Bit Museum - checked it with that multiple times and it showed good every time.  I'd love to try the guts of another 1050 but I don't have one to do that with.  That was my first thought but the prices are silly just now for the few that are on fleabay.  I'm not sure that it isn't the stepper - but getting one of those I think is near impossible - even NOS...

I'll do the calibration according to tf_hh when I replace the FDC - see where that lands me.

 

Cheers,

 

Mark.

Link to comment
Share on other sites

Not cut at the factory. One NEVER does that with a working brain stem, it needs to be told what to do and may be your entire problem. I would attempt to use a bit of thin wire stuck into the the socket and soldered onto what ever is left of the cut off pin right away. You don't leave pins to float since their logic state can change from internal chip leakage and that gets you random results any way you can slice that cake.

 

I'm pretty sure it's not the stepper because it's not passing the RAM test or other section of Power On Self Test code written into the ROM. Stepper code isn't able to even be run because it keeps looping thru the failed POST test to start over and that is what is flashing the light as a clue and spinning up the main motor. 6810 isn't prohibitive in cost and you could use some extras to make a US Doubler out of it as well - why not get some and swap it just to be sure? The good news is it appears that the CPU and ROM are working, but something else isn't quite right. And the RAM test is quite extensive here reading mirrored data images in banks showing up at different addresses that another test might not be testing for.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...