Errors Reading 1050 Disk Drive

[russg]

 

I am wanting to get that, and I am watching one on eBay, but those things are so overpriced, it's a killer...

Am I able to daisy chain the SIO2PC out of the SIO port on the 1050? If my OS is on the thing I would think that would be important, but I heard something about the SIO2PC using all IO channels for virtual drives and I don't know if that would conflict with the physical drive, being on D1 currently.

I actually did get a strobe light today, I can fine tune it too, it's excellent. I just need to know exactly what to tune it to. It gives me a digital readout on the top of the strobe, I -think- that it displays rotations per minute. I could be wrong. What do y'all think? It's this expensive thing: http://www.grainger.com/product/5AY35?gclid=CK7rkPiBlbwCFat7QgodSkwAQg&cm_mmc=PPC:GooglePLA-_-Test%2520Instruments-_-Nonelectrical%2520Properties%2520Testing-_-5AY35&ci_src=17588969&ci_sku=5AY35&ef_id=UYrAoQAAALhAZ3tT:20140123191932:s

'Am I able to daisy chain the SIO2PC out of the SIO port on the 1050? '

Yes. You would boot from SIO2PC/APE .atr with your real 1050 set at D2:. You would plug your SIO2PC into the 1050 and the 1050 into the computer.

There are instructions for making your own SIO2PC. It uses a max232 or 1489.

 

ATARI=>() IBM=>[]

SIO CABLE 9 pin serial

------U-------

:* :

(5)------1:in-: :14------(10)

: \ / : :

*NC 2: o 1 :-in:13 ---

: : 4 \ / :

[2]------3:out 8 o :12 *NC

: 9 : :

(7)------4:in-: out:11 *NC

: \ / :

*NC 5: o :-in:10------[3]

: : \ / :

[9]------6:out o :9 *NC

: : :

(6)&[5]--7: out:8--:<--(3)

: : *DIODE

--------------

*NOTES: If you're positive the cable will never be used with other

peripherals like a 1050, you don't need the diode. If you do use a

diode <recommended> make sure it has a low forward voltage. In the

prototype I tried one with a 1.2V forward drop & it failed to operate,

but a diode with a .65V forward voltage drop worked like gang busters.

This means you can surely get away with any germanium diode & probably

most small signal diodes like the 1N914.

NC means not connected, you don't tie it to anything. Only pins 13 & 14

are tied together. If you feel bad because the thing is too easy, you

can add a few more parts like a .1 Micro Farad disk capacitor between

pins 14 & 7 for decoupling. The funny \ /

o on the 1489 are just an

attempt to show the inverting logic of the 1489. All parts can be had

from your local Radio Shack, but once again, I hate Radio Shack & would

recommend going anywhere else. The SIO pin out when looking at the back

of the Atari is

2 4 6 8 10 12

1 3 5 7 9 11 13

Colors I've found in addition to what Nick Kennedy found

4 & 6 Black & or bare wire. Both of these are grounds, one logic & the

other shield. If you really want to be sure you have a good ground you

can use both of them. Black wire is pretty safe if you're in a hurry &

don't want to bother.

10 +5V Blue in every cable I've seen.

5 DATA OUT It's Green or Dark Green on cables with two green wires.

7 COMMAND Violet on some, light green on others.

3 DATA IN & I don't think it comes in any color but orange.

That's it, 5 connections to the Atari serial cable, 4 to the IBM serial

cable, 1 chip, 1 diode. The IBM connector will have the numbers on the

plastic, but you may need a magnifiying glass to see them. If somebody

wants help building a cable, you're better off asking me about this one

vs the MAX or 7400 as I'm quite bored with them & this circuit is IMO:

The obvious way to go.

Rick

(Rick Cortese post on SIO2PC) ^^^^^

I'm not sure that formats correctly the chip pins. Here's a txt file.

1489.txt

Edited January 23, 2014 by russg

[bitfracture]

Are these even stock caps? Every pic I find of the Atari 1050 seems to have different ones.

[bitfracture]

Thank you russg, but I think I will end up buying one. I don't have access to many components nor the want to add any more personal error potential to my project

 

I removed the caps, they seem to charge and discharge exactly how I would expect them to. I cannot find my cap tester, darn it, but if I do I will tell the results.

I was testing these things using 12 volts 2Amp in parallel with a CPU fan. The voltage rises and falls like I would expect, but I also don't really know what to look for.

[BillC]

 

I am wanting to get that, and I am watching one on eBay, but those things are so overpriced, it's a killer...

Am I able to daisy chain the SIO2PC out of the SIO port on the 1050? If my OS is on the thing I would think that would be important, but I heard something about the SIO2PC using all IO channels for virtual drives and I don't know if that would conflict with the physical drive, being on D1 currently.

I actually did get a strobe light today, I can fine tune it too, it's excellent. I just need to know exactly what to tune it to. It gives me a digital readout on the top of the strobe, I -think- that it displays rotations per minute. I could be wrong. What do y'all think? It's this expensive thing: http://www.grainger.com/product/5AY35?gclid=CK7rkPiBlbwCFat7QgodSkwAQg&cm_mmc=PPC:GooglePLA-_-Test%2520Instruments-_-Nonelectrical%2520Properties%2520Testing-_-5AY35&ci_src=17588969&ci_sku=5AY35&ef_id=UYrAoQAAALhAZ3tT:20140123191932:s

I think the digital readout is just the speed setting of the strobe, I believe what you want is a photo tachometer.

 

http://www.ebay.com/bhp/digital-photo-laser-tachometer-non-contact

[bitfracture]

 

I think the digital readout is just the speed setting of the strobe, I believe what you want is a photo tachometer.

 

http://www.ebay.com/bhp/digital-photo-laser-tachometer-non-contact

 

 

If I make a single mark on the spinning part of the drive, I can set the strobe, then adjust the spin until the mark appears to not move.

[bitfracture]

It seems right now that the main trouble with this board is that it keeps going into that weird cycling mode, where the power to Q6 is cycled on and off instead of having a constant current, and the drive doesn't initialize. I adjusted the timing just perfect, but I can't test it. The drive is cold, all cables are correctly in place, and it was working moments ago. I don't get it...

Also I'm measuring this same on/off behavior, but at 7.3 or so volts on U5 during this.

Edited January 23, 2014 by bitfracture

[1050]

Cycling behavior is rumored to be one symptom of a bad 6532 RIOT (Register, In/Out, Timer) chip. You might see about getting a spare one, pop it in and see if the behavior goes away. Fleabay for used ones? I don't really know of a source, I don't think they are being made anymore, a few extra wouldn't be a bad idea anyway. If this is the problem please do get back to us so we can add a one count to the "bad 6532" case for the cycling fix, most just disappear without a word if the drive is dead or alive.

 

Which pin of U5 is cycling at 7.3 volts? The motor control line comes into the U5 circuit (base of Q4) from pin 11 of the 6532 RIOT. If the cycling signal is also at pin 11 of the RIOT then it might be the source and most likely the fault of the cycling? Just tossing a hat into the ring.

[bitfracture]

Cycling behavior is rumored to be one symptom of a bad 6532 RIOT (Register, In/Out, Timer) chip. You might see about getting a spare one, pop it in and see if the behavior goes away. Fleabay for used ones? I don't really know of a source, I don't think they are being made anymore, a few extra wouldn't be a bad idea anyway. If this is the problem please do get back to us so we can add a one count to the "bad 6532" case for the cycling fix, most just disappear without a word if the drive is dead or alive.

 

Which pin of U5 is cycling at 7.3 volts? The motor control line comes into the U5 circuit (base of Q4) from pin 11 of the 6532 RIOT. If the cycling signal is also at pin 11 of the RIOT then it might be the source and most likely the fault of the cycling? Just tossing a hat into the ring.

 

I think we ought to correct all power problems before I do any chip replacement. I register half a volt of AC coming through the voltage regs, which means those caps aren't doing their job. I think if we kick that problem out of the way then we can be sure nothing else is going to get fried.

I can't find anything labeled the 6532, what is the label name? I apologize for not having the most advanced electronics skills, please bear with me.

On U5: Pin 8 is over 8 volts with motor running, pin 9 at relatively 7 volts, 10 at 3 volts, 11 at 1 volt.

 

I don't know if that's supposed to be that way or not. All of the electronics I have dealt with were all 5 or 3.3, unless dedicated to driving higher or lower voltage equipment. This range seems weird to me.

[BillC]

 

I think we ought to correct all power problems before I do any chip replacement. I register half a volt of AC coming through the voltage regs, which means those caps aren't doing their job. I think if we kick that problem out of the way then we can be sure nothing else is going to get fried.

I can't find anything labeled the 6532, what is the label name? I apologize for not having the most advanced electronics skills, please bear with me.

On U5: Pin 8 is over 8 volts with motor running, pin 9 at relatively 7 volts, 10 at 3 volts, 11 at 1 volt.

 

I don't know if that's supposed to be that way or not. All of the electronics I have dealt with were all 5 or 3.3, unless dedicated to driving higher or lower voltage equipment. This range seems weird to me.

The 6532 is U7 and could also be labeled CO10750. The 1050 Field Service Manual is available at Atarimania:

http://www.atarimania.com/documents/atari-1050-field-service-manual.pdf

[bitfracture]

The 6532 is U7 and could also be labeled CO10750.

 

I reseated U7, and the drive seemed to init correctly more often than before, but given a few starts, right back to where we were before. So you recommend that I order a new one to fix the cycling problem? I am also going to have to order new caps and an SIO2PC then, eh? I just want to be as sure as everyone else on this forum before ordering things, because it always takes weeks for things to ship. Thanks for the help.

[1050]

I think we ought to correct all power problems before I do any chip replacement. I register half a volt of AC coming through the voltage regs, which means those caps aren't doing their job. I think if we kick that problem out of the way then we can be sure nothing else is going to get fried.

I can't find anything labeled the 6532, what is the label name? I apologize for not having the most advanced electronics skills, please bear with me.

On U5: Pin 8 is over 8 volts with motor running, pin 9 at relatively 7 volts, 10 at 3 volts, 11 at 1 volt.

Sorry I was away for a while and did not get back until now, you were NOT being ignored.

As far as I'm concerned, you've got mad skills posting the regulator voltages on your own accord too, simply brilliant work so far. Those caps appear to be correct as they fit the exact silkscreen outlines perfectly, the schematic calls for (3) 4700 uf, 35 volt electrolytic caps, C71 is the voltage doubler, the other two are filtering caps for the 5 volt and 12 volt circuits (one each) to smooth the resulting 30 Hz waveform and if you'll note the oscilloscope waveforms in the field service manual, they are RICH in AC so I don't think they are performing too poorly, but it's your drive, change them out if you want too - it IS the only way to have peace of mind that your caps are good if you lack a scope. What doesn't click for me is that the closest one in your photo appears to be bigger than the other two which do appear to be the same size as each other. Maybe it's just camera angle at work there since they are all supposed to be the same size electrically at least.

 

I don't know if that's supposed to be that way or not. All of the electronics I have dealt with were all 5 or 3.3, unless dedicated to driving higher or lower voltage equipment. This range seems weird to me.

U5 would have to deliver an analog voltage somewhat less than 12 volts in order to read it's tach inputs and provide what ever voltage is required to run the spindle at 288 RPM on the 12 volt motor - that voltage should rise a bit when you close the door on a disk versus closing the door without a disk in it at all.

 

The voltages you provide for U5 then seem to me to be quite within expected ranges, but I haven't gone there myself to verify that they are repeatable here. U5 has a digital front end for the tach sensor, but the output doesn't necessarily have to be a variable duty cycle digital voltage of 12 volts, it could be pure analog. I wouldn't worry about it too much at this point, lots of chips cross the line between digital and analog realms especially in the case of automotive chips for example. Often referred to as linear chips, the 3086 transistor array (U1) is a perfect example, 555 timer (U4) and op amps galore all listed as linear chips.

 

You might try to shine up the pins of U7 with some form of emery cloth, grill pad, small wire brush, or anything else you have around handy and give it another try? I am a firm believer in using contact cleaner on clean pins to insure good contact, I've pulled some 40 pin chips on several 800XLs that were BLACK all the way down on both sides too, no clue what it could have been, but I scuffed it all off anyway. I like to shot the table top with the contact cleaner and then transfer the liquid to the pins with a Q-Tip dabbed into the pool of cleaner. WD-40 might work for this if you don't yet have a spray can of genuine contact cleaner just yet too, I find the drive select switches need a shot of cleaner every 10 years or so as the drive will not respond properly if those switches gunk up.

 

I really would get at least one 6532 for a spare which is what a parts drive is for as well. And after reading the field manual, it shows that I ascribed the wrong words to the RIOT acronym, it's Ram, I/O, Timer, but it does have some PIA port type registers as well.

[Paul Westphal]

Paul Alhart does a great job repairing drives. He has repaired a 1050, Percom and IndusGT for me. E-Mail : p.alhart@verizon.net

[bitfracture]

Those caps appear to be correct as they fit the exact silkscreen outlines perfectly, the schematic calls for (3) 4700 uf, 35 volt electrolytic caps, C71 is the voltage doubler, the other two are filtering caps for the 5 volt and 12 volt circuits (one each) to smooth the resulting 30 Hz waveform and if you'll note the oscilloscope waveforms in the field service manual, they are RICH in AC so I don't think they are performing too poorly, but it's your drive, change them out if you want too - it IS the only way to have peace of mind that your caps are good if you lack a scope. What doesn't click for me is that the closest one in your photo appears to be bigger than the other two which do appear to be the same size as each other. Maybe it's just camera angle at work there since they are all supposed to be the same size electrically at least.

 

Actually, that is what concerned me. Two of the caps (The smaller ones) are not 35 volts like the larger one, they are 25 volt caps. They are all rated at 4700 microfarad.

I will respond to the rest of your comments later, I'm in a rush. Thanks so much!

[bitfracture]

Paul Alhart does a great job repairing drives. He has repaired a 1050, Percom and IndusGT for me. E-Mail : p.alhart@verizon.net

 

Going to give this a good going myself, but thank you for the contact. I will use it if all else fails

 

You might try to shine up the pins of U7 with some form of emery cloth, grill pad, small wire brush, or anything else you have around handy and give it another try? I am a firm believer in using contact cleaner on clean pins to insure good contact, I've pulled some 40 pin chips on several 800XLs that were BLACK all the way down on both sides too, no clue what it could have been, but I scuffed it all off anyway. I like to shot the table top with the contact cleaner and then transfer the liquid to the pins with a Q-Tip dabbed into the pool of cleaner. WD-40 might work for this if you don't yet have a spray can of genuine contact cleaner just yet too, I find the drive select switches need a shot of cleaner every 10 years or so as the drive will not respond properly if those switches gunk up.

 

The pins looked perfect. I had one heck of a time getting the darn thing out, and worse getting it back in... I swear the whole board was about to shatter :/ I feel pretty confident in the contacts though. I am concerned I may have accidentally caused the problem with U7 though. I am really shaky and have trouble controlling what my meter probes touch, unfortunately.

 

I really would get at least one 6532 for a spare which is what a parts drive is for as well. And after reading the field manual, it shows that I ascribed the wrong words to the RIOT acronym, it's Ram, I/O, Timer, but it does have some PIA port type registers as well.

 

I am going to try to get a few if I can find them, maybe it will make an IC worth the shipping for once! And yeah, honestly, I wouldn't have noticed the difference with the acronym name, haha, I read it the other way elsewhere and didn't even notice