• Updated 2023-07-12: Hello, Guest! Welcome back, and be sure to check out this follow-up post about our outage a week or so ago.

400k Drive Repair Guide - A Call to Arms!!!!!!

pbertolo

Well-known member
Hi guys,

follow up notes after replacing the capacitors on the motor board.

Honestly, I could not see any difference between before and after. Either way, the disk drive passed more than 2000 random test units without any fault.

I'm not saying capacitors may not need replacement (I have replaced hundreds of them in my Mac and NeXT systems), but I would personally spare the trouble if not strictly needed. Surface mounted stuff like the ones you can find on SE/30s surely need replacement, but my direct experience is telling me that radial electrolytic are normally OK even after so many years. I only remember one single case where radial electrolytic capacitors miserably failed and caused complete malfunction of the drive and that's about the (in)famous MO drive in the NeXT Computer. For what that matters, I have recapped some of those units too and they all failed in the overheated motor board only, analog boards being normally OK.

Anyhow, pushed by the enthusiasm of this successful story, I took out of the closet a "cold case". Still a M0130 floppy drive (external unit), which did not show any sign of life, except a constant buzz. 

Positive thing about the disk utilities integrated in the "Sony Test" unofficial pre-release is that you can really have a very deep insight of what is working and what not.

So, as soon as I tried to read a disk, the unit threw the message: "Cannot step head" (or something similar, don't remember the exact phrasing). Clearly, the step motor was not able to move (should move to track 00 at start up, which it didn't).

So, as the unit was working perfectly with a tested main PCB, I focussed on it and - proceeding from the stepper connector backward - I did a little bit of component swapping between the faulty and a good board  and eventually it came out that component Q107 is the culprit.

Now, my issue is that I could not find much information about this component, which I'd say must be a diode or transistor of some kind (the back of the board is reading "E" and "B" for the two outermost pins). Component is labelled "B822P M", which so far hasn't lead me anywhere...

Any clue / suggestion?

Thanks in advance!

Paolo

IMG_2437.jpg

 

JDW

Well-known member
I can only see Q107 from the top in your photo, but “Q” almost always refers to transistors, and your part number of “B882P“ seems to be the 2SB882P, which is a high power, 3A, NPN transistor. The modern part number equivalent fir such an NPN would be 2SD882. Download a data sheet for the D882 and make sure that E, C and B pins of the D882 line up with the markings on your board.

UPDATE: I just had a look at my circuit board and can now see that particular transistor. it’s only 4 mm or so in height, so a D882 would only work if you were to bend the pins 90° and lay it down on its back. Even then, I’m not sure if there’s enough space for it.

To test your existing transistor, though, you should desolder it, put it on a breadboard, use a 1k-ohm Base resistor tied to 5v, tie the Emitter to ground, and then tie the Collector to the ground side of an LED, and then tie the positive side of the LED to a 330-ohm (or so) resistor, and the other side to 5 V.  When you put 5V on the base resistor, the transistor should switch, if it’s good, and light the LED. And when you disconnect the base resistor, the LED should turn off.

 
Last edited by a moderator:

pbertolo

Well-known member
“B882P“ seems to be the 2SB882P,
Actually the label on the package of the unit I have is reading “B822”, not “B882”. Is yours different?

By the size of the tiny package and the absence of any heat sink, I would not believe it to be high power stuff rated at 3A...

 

JDW

Well-known member
Actually the label on the package of the unit I have is reading “B822”, not “B882”. Is yours different?

By the size of the tiny package and the absence of any heat sink, I would not believe it to be high power stuff rated at 3A...
I am very sorry.  I misread your previous post, and I don't have access to my PCB right now to verify my board's p/n.  I know my board won't be different that yours though, so what you have (2SB822) is a very short (<4mm height) PNP transistor rated at 2A.  Here's the datasheet:

https://datasheetspdf.com/pdf-file/70749/Rohm/2SB822/1

What you need is a 2A rated PNP that is very short like the original.

I still think you should desolder your PNP and test it.  From the board markings you know which legs are E, B, and C.  If you test the PNP outside the circuit and find it is good, then it means something else is bad.  But if you test and find the PNP is bad, then you can safely proceed to secure a replacement.  The key is finding a 2A-rated PNP that is 3.6mm tall (most are much taller).  I cannot find the original p/n on Mouser, unfortunately.

 

pbertolo

Well-known member
Hello,

thanks for the very useful feedback, thanks to it I guess I've eventually found the missing piece of information.

Transistor specs: 2SB822

Packaging: "FTR" (see picture)

Testing done on the existing one (using a multimeter) is showing it is probably faulty, which is consistent with what assessed so far.

I will now try to source it on line, it's not very common but at www.donberg.ie they seem to have some units.

Will keep you posted.

Paolo

2SB822.png

 

brayne

Well-known member
I'm in the process of restoring a couple of 400K drives myself. Both have had their mechanisms cleaned and lubricated, and the disk load action is very smooth. However, one of them won't eject a disk, and the other won't spin up when you insert a disk (it knows the disk is in there, but the motor doesn't spin).

I happen to know that both motors work, they just don't fire up when they are supposed to.

Anyhow, I can see that the general consensus so far is to recap, so I'll give that a go and then post my results.


Here's my 400k floppy drive replacement capacitor list which I ordered recently:
Thanks for your reply JDW.

As it turns out, my issue with the eject mechanism wasn't related to capacitors. It was the NEC µPA2003C 16 pin NPN silicon epitaxial darlington transistor array, located just below the four diodes on the main controller board. I replaced this and now the eject mechanism is working beautifully.

Now I just need to move onto the other drive that's not spinning up, but there is some terrible corrosion on the PCB with the motor, so that's probably the cause.

 

JDW

Well-known member
As it turns out, my issue with the eject mechanism wasn't related to capacitors. It was the NEC µPA2003C 16 pin NPN silicon epitaxial darlington transistor array,
But did you also change out the capacitors? Those could use changing even if they were not the core problem.

 

pbertolo

Well-known member
Hi guys,

just to confirm that the transistor Q107 was indeed the culprit. Today I replaced it with a spare 2SB822 (thanks to JDW) and the drive works again like a charm.

 

pbertolo

Well-known member
Now I just need to move onto the other drive that's not spinning up, but there is some terrible corrosion on the PCB with the motor, so that's probably the cause.
According to a similar thread on some other forum, Hall effect sensors may also be the cause of a motor failing to spin when a disk is inserted (please see one of my previous posts on this thread).

 

JDW

Well-known member
The problem for me would be the shipping, as I am located in Japan.  I doubt they'd have all the caps I'd want for a particular recap like I can get at Mouser, so it really would be a single item purchase, and what a price that would be!  Luckly, that part hasn't burned out in any of my 400k drives.

I am curious why it burned out though.  Maybe the motor wasn't turning and it passed too much current and blew?  You don't want an expensive replacement to also burn out.

It's a crying shame they don't seem to manufacture that size anymore.  It's a useful low-profile size.

 

Mighty Jabba

Well-known member
I’m way over my head in this thread but I did recently get my first 400k drive in a bundle with an original Macintosh (upgraded to 512k). The seller claimed everything was working, but it was obvious that the grease in the drive had turned to glue and it wouldn’t even accept a disk. I did get a partial refund for that reason, but I decided I would try and fix the drive. To my amazement I was successful. 

I had to clear out the old grease with alcohol and then oil the moving parts. One metal piece also seemed to be very slightly bent, which was inhabiting the mechanism. The drive appears to work fine now and also ejects, but I noticed some mention of adjusting the tracking force (not sure what the actual term is) so that it won’t damage the disks. Is this something I should be concerned about, and how can I tell if it’s off?

 

pbertolo

Well-known member
I am curious why it burned out though. 
I suspect a short circuit, one trace was severed and I had to restore it.

Now the plot thickens, though, as the disk passed at first more than 1k random cycles in the test program, but then at the restart it suddenly start drifting and raising read / write errors. Then again everything OK, then again some random failures. Need to recheck everything again.

 

pbertolo

Well-known member
One metal piece also seemed to be very slightly bent
What I could see while restoring my drives is that the metal frame that is being engaged and pushed by the disks when inserted and that keeps the disk in its position shall not be deformed compared to factory status. Otherwise the alignment of the disk may not be totally correct and the drive may not be able to read good disks.

 

pbertolo

Well-known member
Quick update from my side. The repaired drive is able to format disks and to go through countless check runs. However, disks are not readable by other drives. Means: zero track alignment issue. Means: a sea of pain. I'm giving it a shot, I've been successful with some 800k drives, but this one seems to be particularly nasty and difficult, as you need to rotate the entire motor... Any suggestion is welcome.

 
Last edited by a moderator:
Top