The 1050 Repair Saga: Drive 5 seems to have had an interesting life...

[SlagOMatic]

Thanks again to all the help and guidance I got from AtariAge. Four of my 1050s have been electronically restored/upgraded and are perfectly functional. (Cosmetic restoration to come.) Now my last 1050 is on the bench. I've noticed a few things and was wondering if there were any conjectures as to what's going on.

 

So, this drive's symptoms are it turns on (LED lights up for power) but there's no drive activity. When I opened it up the first thing I noticed was the three big capacitors in the back were slightly swollen, but more curious to me is the fact that none of them have scores on the tops of the capacitors (first picture). I've legitimately never seen that before. I was always under the impression that the score marks on the tops of caps were essentially a safety release; if the caps got too swollen they would burst along the score lines. These caps don't have that so...less safe? Cheaper design? Or is it a not-uncommon thing that I've just not encountered before?

 

Pulling the drive mechanism off revealed the RF shield was missing so someone was clearly in here before me. Further proof of that is the fact that the U8 (6810 static RAM) and U10 (OS ROM) chips are missing (second picture). This would certainly cause the lack of drive activity even if the caps were good.  🙂  Fortunately I have extra U8 chips, and since I intend on installing a Happy 1050 upgrade I don't actually need the U10 chip. But I'm guessing that this drive formerly housed a US Doubler modification. One of my 1050s had a US Doubler in it (installed by me ~35 years ago) and removing those chips was necessary to do the installation. Unless there's another potential reason for removing those two chips?

 

The last thing I noticed is an odd bit of resistor/capacitor work attached to C36 (third picture) which as I understand it has something to do with the drive alignment? Not sure. But I've not seen this in any of my other four drives. (This is a Singapore drive as are three of my other drives.) It looks too clean and perfect to be a hack job but what do I know. This is the bit that concerns me the most. What would it be doing? Should I remove it?

[xrbrevin]

the capacitor is detailed here:

 

the variable capacitor is 1 of 3 components used in calibrating the FDC chip to it's drive

the additional components on your drive could suggest either a repair to bring the reading into spec or (dont quote me!) maybe a mod for copying protected disks / skew mod?

[blacka013]

I would check the +5V and +12V supplies and if they're fine fit the Happy board and check it operates correctly.

 

If it doesn't, put the board back to stock, (will need an OS chip and initially leave the added resistor and capacitor in place) and fault find from there.

 

If you're not happy with the 3 capacitors, replace them.

[TZJB]

11 hours ago, SlagOMatic said:

Thanks again to all the help and guidance I got from AtariAge. Four of my 1050s have been electronically restored/upgraded and are perfectly functional. (Cosmetic restoration to come.) Now my last 1050 is on the bench. I've noticed a few things and was wondering if there were any conjectures as to what's going on.

 

So, this drive's symptoms are it turns on (LED lights up for power) but there's no drive activity. When I opened it up the first thing I noticed was the three big capacitors in the back were slightly swollen, but more curious to me is the fact that none of them have scores on the tops of the capacitors (first picture). I've legitimately never seen that before. I was always under the impression that the score marks on the tops of caps were essentially a safety release; if the caps got too swollen they would burst along the score lines. These caps don't have that so...less safe? Cheaper design? Or is it a not-uncommon thing that I've just not encountered before?

 

Pulling the drive mechanism off revealed the RF shield was missing so someone was clearly in here before me. Further proof of that is the fact that the U8 (6810 static RAM) and U10 (OS ROM) chips are missing (second picture). This would certainly cause the lack of drive activity even if the caps were good.  🙂  Fortunately I have extra U8 chips, and since I intend on installing a Happy 1050 upgrade I don't actually need the U10 chip. But I'm guessing that this drive formerly housed a US Doubler modification. One of my 1050s had a US Doubler in it (installed by me ~35 years ago) and removing those chips was necessary to do the installation. Unless there's another potential reason for removing those two chips?

 

The last thing I noticed is an odd bit of resistor/capacitor work attached to C36 (third picture) which as I understand it has something to do with the drive alignment? Not sure. But I've not seen this in any of my other four drives. (This is a Singapore drive as are three of my other drives.) It looks too clean and perfect to be a hack job but what do I know. This is the bit that concerns me the most. What would it be doing? Should I remove it?

 

 

 

The capacitors look fine and are fully insulated so the metal can is not visible. Leaking would occur at the board level if there were a dielectric expansion problem.

It looks like the previous owner pulled out a US Doubler to sell seperately, without restoring the original components. The jumpers have been moved to JP2 and JP4 which indicates an EPROM was fitted and not a masked ROM. Obviously without the ROM and RAM components the drive is disabled.

The extra components connected to C36 are most certainly to bring the WD2793 chip into spec. as @xrbrevin stated. Definitely leave them in. A copying mod would add an additional potentiometer, not a capacitor. I have one an a 1050 Archiver.

[xrbrevin]

53 minutes ago, TZJB said:

A copying mod would add an additional potentiometer, not a capacitor. I have one an a 1050 Archiver.

Aha - thanks for the extra detail

every day is a learning day

[SlagOMatic]

As always, thanks for all the input.

 

Like I did with the other four drives, Drive #5 has been cleaned, new caps and regulators put in, replacement 6810 and Happy upgrade installed, and the mechanism cleaned and lubricated and a new belt installed.

 

Initially it seemed good. The drive spun up (it didn't before) and the computer recognizes it on the SIO chain (as D2). I booted the Happy diagnostic disk and give it a shakedown. The enhancement diagnostic and hi speed transfer diagnostics both pass. The RPM measure test fails with "ERROR can't measure" with the occasional flicker of an RPM speed that's north of 950rpm (which there's no way the disk is spinning that fast). The read/write test failed the first two times I ran it, then it passes on all subsequent runnings including after multiple cold boots of the drive and/or the computer.

 

Since that little capacitor/resistor package is attached to the variable capacitor that (partially?) controls drive speed, and since the drive speed diagnostic fails, that suggests to me that I should pull that capacitor/resistor package and try it again? Or should I try something else first?

[_The Doctor__]

the rpm resistor is where you adjust and then check- use a strobe light on the marked flywheel to dial in the speed if you don't trust what software reports

else a timing mark and sensor can count it for a timed interval and you do the math to get a real rpm number

Edited March 25, 2023 by _The Doctor__

[SlagOMatic]

13 minutes ago, _The Doctor__ said:

the rpm resistor is where you adjust and then check- use a strobe light on the marked flywheel to dial in the speed if you don't trust what software reports

else a timing mark and sensor can count it for a timed interval and you do the math to get a real rpm number

Okay, but...what about the resistor/capacitor?

[_The Doctor__]

You'd need to measure exactly what it changed the pulse timing to, perhaps they tried an experiment with a different FDC?

Was it done to fix some other form of damage?

I don't have it in my hands to test any of those theories, but anything is possible. Atari alchemists do strange things... cheapo shoestring attempt at a bit writer like mod?

Edited March 25, 2023 by _The Doctor__

[TZJB]

6 hours ago, SlagOMatic said:

Okay, but...what about the resistor/capacitor?

The variable capacitor C36 is connected to U13 WD2793 pin 26 VCO and adjusts the 125 KHz waveform. If you have an oscilloscope you would be able to calibrate the WD2793 using the article HowTo: Adjust analogue part of FDC 279x in a 1050 disk drive by @tf_hh posted above by @xrbrevin. Without one, just leave it alone as you could disable the drive completely. It does not control the RPM anyway which is adjusted by VR2 connected to U5 LM2917 pin 10.

[SlagOMatic]

5 hours ago, TZJB said:

The variable capacitor C36 is connected to U13 WD2793 pin 26 VCO and adjusts the 125 KHz waveform.

But I'm not and never have been talking about removing or disabling C36.

 

I think maybe my question/intent isn't clear.

 

I refer you to the third picture I posted at the opening of this thread. Coming off C36 is a capacitor/resistor piece that (a) isn't present on any of the other drives that I've managed to resurrect, and (b) unlike C36 itself this capacitor/resistor piece appears in none of the service documentation I've found. Given that, plus the fact that this drive was clearly worked on before I got to it, it leads me to believe that a previous owner made that modification for reasons we can only speculate about.

 

Now, you had previously said, "The extra components connected to C36 are most certainly to bring the WD2793 chip into spec. as @xrbrevin stated. Definitely leave them in." So I did. While I am not an electrical engineer and I'm completely feeling my way through this whole process, it seems more than a little coincidental to me that the only issue this drive seems to be having is in regards to measuring its RPM which is a function of drive control, and that capacitor/resistor piece just happens to be attached to part of the drive controller. None of the other drives have it, none of the other drives has an issue measuring RPM. This drive has it, this drive has an issue measuring RPM. It's literally the only unique thing about this drive, and this drive alone is having this unique issue.

 

So with that in mind I'm left to question your past advice of "definitely leave them in".

[+kheller2]

Random comments:

1) Is the RPM test you are using designed for Happy upgraded drives and does the track buffer need to be on or off?

2) If you want to adjust the RPM speed, then use the proper location (blue box VR2 I think).

[SlagOMatic]

I decided to just remove it and see what happens.

 

As before, I booted the Happy diagnostic disk. The enhancement diagnostic and hi speed transfer diagnostics both pass. The RPM measure test still fails with "ERROR can't measure" but now the occasional flicker of an RPM speed goes from around 288rpm (which seems accurate) up to about 600rpm (which seems not), and the read/write test now consistently fails at the beginning of the "medium density" part of the test.

 

However, in limited testing the drive appears to work. I'm able to boot, read, and write to/from single, enhanced, and double density media seemingly without issue. This doesn't sit well with me; I don't like when errors are generated but it still "works."  😕

 

I think at this point my only option is to buy and figure out how to use an oscilloscope to track down what's happening here.

[TZJB]

17 hours ago, SlagOMatic said:

But I'm not and never have been talking about removing or disabling C36.

 

I think maybe my question/intent isn't clear.

 

I refer you to the third picture I posted at the opening of this thread. Coming off C36 is a capacitor/resistor piece that (a) isn't present on any of the other drives that I've managed to resurrect, and (b) unlike C36 itself this capacitor/resistor piece appears in none of the service documentation I've found. Given that, plus the fact that this drive was clearly worked on before I got to it, it leads me to believe that a previous owner made that modification for reasons we can only speculate about.

 

Now, you had previously said, "The extra components connected to C36 are most certainly to bring the WD2793 chip into spec. as @xrbrevin stated. Definitely leave them in." So I did. While I am not an electrical engineer and I'm completely feeling my way through this whole process, it seems more than a little coincidental to me that the only issue this drive seems to be having is in regards to measuring its RPM which is a function of drive control, and that capacitor/resistor piece just happens to be attached to part of the drive controller. None of the other drives have it, none of the other drives has an issue measuring RPM. This drive has it, this drive has an issue measuring RPM. It's literally the only unique thing about this drive, and this drive alone is having this unique issue.

 

So with that in mind I'm left to question your past advice of "definitely leave them in".

I can appreciate that, but did you read and understand the WDC calibration process? I still stand by my advice, but you can do whatever you feel is best. Sometimes feelings can work. C36 and the added components in this instance are one in the same in my mind, sorry for not making this clear.

 

However I think you are confusing drive control with motor control. The WDC handles the data transfer to and from the floppy disk using frequency modulation (FM) and modified frequency modulation (MFM) and controls the head motor, but has no control over the drive motor which is the function of the LM2917.

 

But as you say, the drive has been worked on so you cannot even be sure that the WDC is the originally calibrated device. I would be tempted to put the drive back to stock with a Tandon OS and see how it behaves with an Atari test disk attached.

 

1050_Disk_Diagnostics_1983_Atari.atr

 

Regarding your tests, do the disks you make with this drive also work in another drive or is the calibration so far out it only works in this drive?