15 Jun
2007
15 Jun
'07
3:41 p.m.
All other self-tests pass ok, and if I put a floppy
formatted with format
self-test in another drive (I have a 82902 also), I can read it with the
read self-test as described in the service manual.
So I'm almost sure that there is a problem in the read circuit. It can be in
the analog or digital part of it anyway, but there are more probability that
it's in the analog part.
Maybe, maybe not :-). I've not traced out any scheamtics yet, but the
read circuit is not simple. When tracing out connections to that MB8866
chip, I found the read data line from the drives [1] goes through what
appears to be a delay line made of L's C's and a 74LS14 chip. The problem
could be in that area. Or it could be in the VCO circuit, or even (let's
hope not) in the MB8866 chip itself.
[1] The 34 pin drive cable is the standard Shugart pinout, of course. The
drives (HP9130 IIRC) are Tandon drives, built to HP spec, with
gold-plated PCBs, and the like. I don;t know which company actually made
and asembled the PCBs.
FWIW, the older 9885 8" drive has a Shugart 800 mechanism with HP
electronics (this PCB is not a copy of the standard one. The service
manual for that unit actually mentioned this fact, and gives Shugart part
numbers for some of the mechancial bits.
Also, I made some reads with the oscilloscope, and it
seems there is a
shorted transistor. I ordered a few in a local store and I'm waiting for
I don't know what tests you made, but it's not uncommon for theere to be
transsitors with bace and collector linked together in this sort of
circuit. So in-cricuit tests may bot be relevant.
Most of the transistors seem to be 2N3904 (NPN) and 2N3906 (PNP). but
there's at least one with an 1855-xxxx number, that may well be a FET.
-tony