Two Atari 1050 hardware problems

[jmccorm]

1. MARGINAL DRIVE ID SWITCHES:

I have at least two Atari 1050s with marginal drive ID selection switches. They "prefer" to be Drive 1. When I switch the switches to Drive 2, even after a power cycle they want to stay as Drive 1. Sometimes, as I press against the switches or as I release them, they'll go back and forth between answering as Drive 1 and as Drive 2. I'm suspecting that the switches are oxidized with age. At least two different 1050s exhibit this problem. Is this a known issue? If I don't want to hard-wire a setting, are other people replacing these, and if so, from where? Or do they do something else?

 

2. INTERMITTENT. NO FIRMWARE FUNCTIONS AFTER POWER-CYCLE, REGULATORS ARE HOT TO TOUCH:

I have one Atari 1050 that I'd prefer to use because it does a slightly better job of reading marginal sectors, but it has an intermittent problem. When the drive works, it works great. On any subsequent power cycle, when you turn on the drive, it does nothing. When you close the drive's door, it does nothing. (Connected to the Atari or not, when it doesn't work, there is no busy light and no spin-up.) The large metal wall in the back of the drive seems hot to the touch on the back-left side (with the front drive door facing you) while a "good" 1050s might only seem very warm. Tried two power supplies, one rated 9.5VAC/40VA (11.2VAC actual) and the other rated 9.0VAC/31VA (10.5VAC actual), no change, and both power supplies work fine with other 1050s.

 

EDIT TO ADD: From memory, the regulator on the back-right side seemed to run fairly cool.

 

I'm hoping that these are known and/or common problems with well-worn answers. But I'll take anything you can offer.

 

Thanks, Atarians.

Edited February 12, 2018 by jmccorm

[+DrVenkman]

Can’t help you with problem #2, but as to #1, dirty switches and/or cold solder joints at the switch contacts to the PCB are common hardware problems. My first approach would be to squirt some electrical contact cleaner (Deoxit especially) into the switches - just a single small squirt is almost certainly enough - and cycle both the black and white switches a dozen times or more, then let them air dry for a few minutes. After that, they’ll probably be fine. If not, you might try touching up the contact points under the PCB with a soldering iron and/or a drop of liquid or gel flux to melt and reflow the solder.

[jmccorm]

Can’t help you with problem #2, but as to #1, dirty switches and/or cold solder joints at the switch contacts to the PCB are common hardware problems. My first approach would be to squirt some electrical contact cleaner (Deoxit especially) into the switches - just a single small squirt is almost certainly enough - and cycle both the black and white switches a dozen times or more, then let them air dry for a few minutes. After that, they’ll probably be fine. If not, you might try touching up the contact points under the PCB with a soldering iron and/or a drop of liquid or gel flux to melt and reflow the solder.

 

Next stop, Amazon. Thanks, DrVenkman.

[+Nezgar]

A quick comment on #1 - you don't need to power cycle the drive to change the drive number - this can be done while the drive is on, and will take effect immediately.

 

Re #2-Back left regulator is 5V and has consistent draw for the pcb electronics. Back right regulator is the 12V for the drive motor, and will generally heat up mostly when the drive is spinning. Might be plausible if you're keeping the drive busy in a long session of archiving.

 

Personally I've found most of the intermittent non post issues are either a bad 6810 RAM chip, or bad/corroded contacts on the socketed chips (ROM, CPU, RAM, RIOT). Worth popping out those chips if they're socketed, and reinserting maybe with a bit of deoxit or isopropyl alcohol action on the sockets and chip legs before.

 

Don't worry too much about the AC voltage of the power supply, it will drop under load. You can measure the DC voltage out of the regulators if you want assurance the drives internals are getting good 5v and 12v DC.

[_The Doctor__]

Some people think higher powering to the drive or using the higher voltage read on a ddm of power block on the drive is needed or good because of the motors involved but that's not the case the closer the input voltage to the voltage drop across the regulator in use plus the output voltage the more efficient it is and the less heat it will put out. We go for the sweet spot not too high not too low.

 

sometimes the regulators crack their solder joints from repeated hot cold cycles, re flow them.

 

sometimes the caps have become marginal replace them if all else checks out good and everything is clean and reseated. When the caps finally completely fail they sometimes take the regulators and other things with them...

Edited February 12, 2018 by _The Doctor__

[jmccorm]

I resocketed the chips, and I was hopeful, but the problem quickly returned. But actually, that's not so bad.

 

I've followed through on the easy answers, and now anything more will take a good deal more time and effort. The Doctor pointed out some of those possibilities. What I ended up doing is labeling that particular 1050 as a parts machine. I swapped its good mech into another 1050 that had been having seek issues. That's one more drive fixed. Now I'll also have a redundant set of drive selection switches that I can clean more aggressively.

 

Thanks, all, for driving me towards a better answer.

[xrbrevin]

Hi,

for fault #2, check the small electrolytic capacitor: C70.

In a doa drive i got, it metered up as a dead short, causing the regulated 12v DC to short to earth, heating up the heatsink (the metal wall) very quickly!

i replaced it and it solved the problem for peanuts

hope this helps!