Servicing NeXT (Canon) MO Drives

Started by korneluk, June 28, 2007, 06:27:00 pm

Previous topic - Next topic


I couldn't help noticing a whole pack of electrolytic capacitors (aluminum canister) near the motor driver. In my experience faulty electrolytics can be the cause of rapid failure, gradual failure, and failure of old-stock off-the-shelf electronics. I certainly have had a lot of experience with electrolytics being the cause of monitor failures and have had success reviving the monitors by replacing electrolytics (it's posted somewhere on Sometimes electrolytic failure is obvious - leaky, discolored, or even exploded canisters - but most of the time there is no evidence. As electrolytics fail sometimes they just drift out of spec for a while and you get all sorts of weird effects. Could be those little canisters on the boards.



This patent document has several flowcharts for the operation of the drive:

1. Load cartridge
2. Turn on motor spindle
3. Turn on autofocusing servo and autotracking servo
4. ...

Canon MO  prior art filing - PDF

Also, I was thinking about the integrity of the capacitors, however, looking at the flickr pictures, and examining the caps, they look impervious to the breakdowns seen in the last few years (Japanese product vs. Chinese/Taiwanese?)

- itomato[/url]



Thanks for the patent link. It certainly looks like it has some information that might eventually help figure out the stupid "spin up, spin down" loop that at least one OD (mine) goes into every time a disk is loaded now. It's hard to tell since I have very little idea of what they are talking about in most of the patent.

I noticed the patent was filed in 1995. In general, one would want to file for a patent before the device was produced and sold in great number. It appears to me that this patent is for an improvement meant to solve a magneto-optical drive problem, mainly the "servo control failure" mentioned in the first paragraph of the "Summary of the Invention" (second to last page in the PDF). This patent may be an attempt by Cannon to improve on the obviously failure-prone OD.

I was wondering how you can tell the capacitors are impervious to breakdowns? I've removed bad electrolytics that show no discernible external evidence of failure. I have used the 'heat, then cool' method to pinpoint bad caps that otherwise look fine. Unfortunately, the device needs to be at least somewhat working when the diagnostic is done, and it appears that would be hard to do with the OD's. An alternative is to remove the caps and test them off the board, or just replace them. It's a lot of work and whenever you remove components from boards you run the risk of 'pulling tabs' (de-laminating contacts from the circuit board). Incidentally, I'm not entirely convinced that failing electrolytics are the culprit in failed OD's, certainly other things could be the problem(s), as gtnicol success with reflowing the solder to get the things to work again would suggest.

Thanks again for doing some investigating. We are certainly learning a lot about these whacky drives.



I was curious if anyone knew if the Canon part number for these drives -- if there was one.

I'm musing whether or not Canon sold these drives to anyone other than NeXT?  The PC board would certainly indicate so as it has a 50 pin SCSI connector as well as what I believe is a NeXT proprietary MO connector.

Maybe I'm wrong -- it's been a while since I had to pull my smoked MO out of my cube.
ND \'040 NeXTcube [SN: ABA0004306], NeXTStation Turbo Color ADB, HP PA-RISC 712/80, HP PA-RISC 712/100, Sun SPARCSTATION 5, Sun SPARCSTATION 20


Well my brand new MO disk arrived today and I went to see if my drive even worked and...nothing happened :/ .
It didn't even pull the cartridge into the drive. It's dead as a doornail.
Well, it looks like I'm looking for another MO drive to try and tinker with.

-NeXT 68040 Cube with NS 3.3 and 64Mb ram


Good stuff korneluk & Pentium.

Can someone help out with the following burnt out component? Is it a capacitor - I cannot make it out, it's just charcoal... It's on the little board on the back of the drive, with the connecting ribbons.


Yuo got worse problems. The ribbon cable appears to have been damaged by the extreme heat.

-NeXT 68040 Cube with NS 3.3 and 64Mb ram


It looks like that, but it's actually the plastic "shield" that got melted... In either case, I still need to know what the component is :-)


Go up and to the left and you will see a black block with a white end on one side and marked "C105".
That's exactly what got burned up.

Edit: I just noticed that those little green capacitors on the spindle motor board (and I would like to remind you all that a constantly spinning or non responsive spindle motor is a known problem) that in the past we have suspected of leaking are all over the inside of these MO drives. If by chance they are indeed leaking, they should all be leaking by now. I have a hunch that just like my late 80's CCD-V5000 camcorder, once you replace the caps, it will work again. Also, I beg to differ on the risks of pulling the caps. So long as you got a decent iron and solder sucker you should be fine and replacing the caps would a a hell of a lot easier than reflowing the hundreds of possibly bad joints in the MO drive.

-NeXT 68040 Cube with NS 3.3 and 64Mb ram


Man, I'm impressed by the number of trim pots.  Must be fun to calibrate these drives....


Do you think replacing the motor control IC would help? I managed to locate a few and I'll probably try this.


to my knowledge, that might not be faulty. I'm still pretty ceatain it's the capacitors which have gone tits up.

-NeXT 68040 Cube with NS 3.3 and 64Mb ram


Quote from: "pentium"to my knowledge, that might not be faulty. I'm still pretty ceatain it's the capacitors which have gone tits up.

If it's just the caps, they should be easy to replace. I do think that chip is messed up, though -- at least mine looks VERY ugly.


Leakage from the caps can make some really ugly messes. Unless you see something really obvious like the PCB is burned, the chip is cratered or it's brown, it should be okay. The best you can do otherwise is reflow the joints and check for continuity in cases the leakage has been eating at the traces.

-NeXT 68040 Cube with NS 3.3 and 64Mb ram


Thank you Pentium for your disassembly photos. They helped cut the time.

I am determined to get ALL our MO drives functional. GTNICOL has graciously shipped me 3 non-functional units for a fair price.

I have finally gotten to this project and will get it finished.

I identify three PCB's in each MO drive. I refer to them as the Digital Board, Analog Board, and Motor Board.

On drive #1-
I saw no visual evidence of electrolytic capacitor leakage. However, I pulled all e-caps. Only 2/10 measured remotely close to the correct value when attached to a LCR meter, and one of those was out of 20% spec -

listing legend-

part # / part value on cap / temp rating/ radiusxlength mm / (measured value)


C3   2.2uf@50v   105C   4x6   (0.33uf)
C6   10uf@35v   105C   5x6   (0.02UF)
C7   10uf@35v   105C   5x6   (8.8uf)
C8   10uf@35v   105C   5x6   (0.23uf)
C9   10uf@16V   105C   4x6   (0.18uf)
C2   1UF@50V      105C   4x6   (0.5uf)
C4   10uf@35V   105C   5x6   (7.6uf)
C5   10uf@35V   105C   5x6   (none)

Soon I will measure the following other Electrolytics -





Obviously we are dealing, unsurprisingly, with degraded electrolytic capacitors which have been subjected to large amounts of heat and time.

I am not suggesting this is the only fault, but as you can see the capacitors so far measured on the motor control board are completely failed out of spec.