HP 35660A DSA Floppy Disc Drive Repair

[Smoky]

I bought this HP 35660A Dynamic Signal Analyzer two years ago. The only fail on the many pages of the self-test report is the Floppy Disk Drive.

While I'm inbetween repairs of classic oscilloscopes, I thought I'd give this repair a try.

I watched Dave's tear-down video on this DSA and I found the disassembly absolutely a breeze! The interior of the DSA is extremely clean and parts within the machine are dated 1986-1988. The Nichicon electrolytic capacitors are dated 1987.

I also watched a couple floppy disc drive repair videos on YouTube and most of the failures are a result of leaking electrolytic SMD capacitors. Hopefully, if that's the case with mine, the copper traces and pass-throughs have been spared (I'm having flashbacks of my TDS420 capacitor nightmare)

I did put a new and formatted disc in the drive during the self-tests.

Many pictures to follow!







The disc drive is made by Sony, model number MP-F52W-30. It spins silky-smooth too:



I can easily see two electrolytic caps near the rear opening:



...and three more through this opening. Time to drop the board:

[Smoky]

If you follow these steps, the electrolytic capacitors are totally accessible for testing and replacement:









Absolutely no parts fell out of place or anything! The mechanical assembly lifts off while leaving the PC board resting on your bench. This stuff is amazing

...I'll de-solder the capacitors tomorrow and test them for leakage.

Also, there are two of these tantalum capacitors on the board. 3.3uf at 10V. I'll test them too:



Ok, all of the electrolytic capacitors were tested. They all pass if they are of the 20% type but most of them tested right at the +20% level. I plan to replace them just to be safe. I found one bipolar electrolytic. It appears to be a .047uf capacitor that I thought at first was a 5% tolerance version. I come to find out that the "J" represents the voltage which equates to 6.3V, correct me if I'm wrong. I removed it and tested it in both directions, 49200pf with no leakage was what I read both ways. I reinstalled it.

What's nice is the way they marked the PCB for capacitors. Both the schematic symbol is used in addition to a dot representing (-) so that it can be verified after the part is installed:

[fzabkar]

FYI, I found this document:

http://bitsavers.trailing-edge.com/pdf/sony/floppy/9-975-174-01_MP-F52W_Product_Specifications_Sep85.pdf

[Smoky]

I'm in the process of putting the Floppy Drive back in.

While I'm in here, would anyone happen to know where the battery is located in this HP 35660A?

It will not keep the time and date when the unit is turned off and then turned back on again.

I did hear someone talk about "Super" capacitors when discussing some versions of HP DSA's.

What type of capacitor that is, I have no clue. If anything, would anyone know what one of those may look like?

Thank you!




For instance, on the memory board, it is mostly black socketed and soldered-in IC's but three of these capacitors are among them, both ends and in the middle. They are 5/8" long and 3/16" in diameter:

         

Scratch that. I found out that these are Kemet T140 Hermetically Sealed Tantalum capacitors.

[m3vuv]

should esr not be a better test as oposed to leakage??

[james_s]

The memory backup capacitors I remember seeing in that era were radial lead things about half an inch in diameter. I don't know where it is in that unit but I wouldn't think it would be all that hard to find.

[cuebus]

If it is anything like the 35665a, the battery is on the top logic board near the back (you have to pull it out). A 3.6V AA size lithium battery with leads. I ordered a Tadiran one from Mouser to replace mine.

[Smoky]

Thank you for the help!

M3vuv, I did also test using an ESR meter. I use a LC102 and an Atlas ESR+. One of the capacitors was below 1uf, so neither meter registered a reading for ESR. Both the old and the new capacitors tested very well.

James, The only part that resembles the one in your picture is this thing. I believe it's a buzzer/beeper from Murata:



With the boards removed from the machine, I tested various components on each one but could not find any residual DC voltage. Absolutely nothing.



I find no mention of the battery in the service manual.

I was in contact with one of the team members of HP that designed the 35660A, but unfortunately, his area of expertise didn't cover the power/memory portion of the 35660A. He thought the up and down and the right to left arrow buttons were ridiculous! HP went to a rotary control on the 35665A.



I spoke with George at GLK Instruments last year. He specializes in the 35665A. He had no clue where the battery was either.

As for the Floppy drive, this particular one was prone to failure and it was also changed to a different style in the 35665A.

Maybe I just won't shut this thing off

[SeanB]

(Attachment Link)

In the image, near the buzzer, is a black rectangle with red sides. I have seen those before, with a lithium CR2032 cell under the cap, from equipment used in the 1980's.

[Smoky]

Thanks for the help Sean.

I don't think it's a battery compartment. I checked it early on, and on the backside of the PCB, it reads "L900." I took for granted it was an inductor of some sort. It is also smaller than a CR2032 battery cell. Here is a picture of it from the top and bottom. I put red dots next to its two terminals. I couldn't find any information on it based on the numbers. It also looks like it could be a potting material that fills the red box to the top. My multimeter could not sense any voltage there whatsoever:

     

There is good news though, my Floppy drive works

I sort of wonder if the drive was so without lubrication, especially the worm-gear from the small motor and the many points and pivots, that it was stuck? After all of the cleaning and Teflon dry lube and such, the transport snapped into all positions with no effort. The mechanism didn't function like that before I cleaned it.

I'm sure those five little 105C, 5000hr capacitors I put in there are going to get quite the workout

What's funny too is that the quick-confidence test results are 3 pages long!

[james_s]

I wonder if it originally had a NVRAM with an internal potted battery and someone changed it with a standard SRAM? Dallas also used to make a "smart socket" that looks like an ordinary IC socket but underneath is a potted coin cell and a power control ic, you insert a standard SRAM chip and the socket makes it a NVRAM. I would look closely at any RAM chip you see.