28 Nov
2010
28 Nov
'10
12:52 a.m.
Hi guys,
As you might have gathered from my other thread, I'm toying with the
idea of building a replacement controller card for my two dead
double-sided Amstrad disc drives. Ideally I'd like to use as many
"standard" parts as possible to ease debugging and future repair. I'll
probably use a GAL to handle head-stepping and the write protect
lockout, and LSTTL for the bus interface.
I'm also planning on adding an overvoltage crowbar (or maybe some form
of a MOSFET OVP switch) on the 5V rail to protect against the common
killer of these drives: plugging them into a PC PSU (the power plug is a
Berg floppy power connector, but with a mirrored pinout).
I've got my hands on a couple of Texas Instruments MC3470A floppy disc
read amplifier ICs, and the Apple Disk II schematics suggest how this
can be tied to a drive mechanism. What I'd like to do is look at a few
other drives which use this chip to make sure I understand everything
correctly (a double-sided drive with a single 3470A would be handy). If
I'm being honest, the TI datasheet for the chip is utter crap...
So does anyone know of any other FDDs -- besides the Disk II -- which
used the MC3470A as the head amplifier, or maybe a few applicable
application notes?
Thanks,
--
Phil.
classiccmp at philpem.me.uk
http://www.philpem.me.uk/
2:08 a.m.
On 11/27/10 6:02 PM, Al Kossow wrote:
On 11/27/10 4:52 PM, Philip Pemberton wrote:
shugart/SA850_450_Read_Channel_Analysis_Dec79
siemens/FDD100-8_V1 starting at page 3-32
If I'm being honest, the TI datasheet for
the
chip is utter crap...
Dig around on bitsavers. There is a detailed analysis of the 3470 in one of the disk
directories (moto, shugart?)
2:40 a.m.
On 27 Nov 2010 at 18:02, Al Kossow wrote:
Dig around on bitsavers. There is a detailed analysis
of the 3470 in
one of the disk directories (moto, shugart?)
3470's were used in a LOT of drives.
Qume drives used them. The other half of the puzzle is the MC3471
write amplifier which seems to be a bit more uncommon. I've only
seen it on MPI/CDC drives.
Another popular circuit used on Tandon and Micropolis drives as well
as quite a few other brands was the NE592 (identified as a video
amplifier, but perfect for floppy or tape read amplifier)
--Chuck
9:40 a.m.
On 28/11/10 02:40, Chuck Guzis wrote:
Qume drives used them. The other half of the puzzle
is the MC3471
write amplifier which seems to be a bit more uncommon. I've only
seen it on MPI/CDC drives.
Though the Motorola appnote Al linked to explains that you can do the
same thing with a constant-current source, DFF and a pair of switching
transistors. I suspect "fast" is the order of the day for the
transistors, which probably means the BC547 and BC847 aren't suitable.
Another popular circuit used on Tandon and Micropolis
drives as well
as quite a few other brands was the NE592 (identified as a video
amplifier, but perfect for floppy or tape read amplifier)
That's... tempting. The 592 is still a current part, which at least
makes repair a little easier. Though with five MC3470A chips in stock,
I'm not overly concerned about availability issues...
Although whether I've got enough clearance to shove a PDIP chip in there
without fouling the drive belt or flywheel is another question entirely...
--
Phil.
classiccmp at philpem.me.uk
http://www.philpem.me.uk/
9:10 p.m.
On 28 Nov 2010 at 9:40, Philip Pemberton wrote:
I have never sene that version used in a disk drive, and I think for a
good reason. The 8 pin package doesn't bring out at least one pair of the
'gain select' pins. These are common used in disk drive read amplifiers
to connect am LCR network to set the frequency respsonce of the
amplifier.
-tony
Although whether I've got enough clearance to
shove a PDIP chip in
there without fouling the drive belt or flywheel is another question
entirely...
Well, the 592 does come in an 8-pin PDIP if that's any help.
8:16 p.m.
Though the Motorola appnote Al linked to explains that
you can do the
same thing with a constant-current source, DFF and a pair of switching
transistors. I suspect "fast" is the order of the day for the
transistors, which probably means the BC547 and BC847 aren't suitable.
I don;t think the transistors used were that exotic. Things like 2N3904s,
for example.
The semi-0discrete solution ('74 as a divide by 2 and a handful of
transistors) turns up all over the place. It's easy to design and reliable.
There were combined read/write chips (I can try to find some details),
but they are very few and fr between.
Although whether I've got enough clearance to
shove a PDIP chip in there
without fouling the drive belt or flywheel is another question entirely...
Do you have to pack it into the same volume as the original drive?
-tony
10:10 p.m.
On Sun, 2010-11-28 at 20:16 +0000, Tony Duell wrote:
I'd like to, for the sake of tidiness. It's not essential, though. Also
extending the mechanism cabling would be a bit of a pain.
--
Phil.
classiccmp at philpem.me.uk
http://www.philpem.me.uk/
I don;t think the transistors used were that exotic.
Things like 2N3904s,
for example.
Hmm, I wouldn't have thought they'd work at ~1MHz type speeds. They
certainly didn't work as voltage translators on a parallel-port JTAG
cable (I ended up using a TI level translator chip instead).
The semi-0discrete solution ('74 as a divide by 2
and a handful of
transistors) turns up all over the place. It's easy to design and reliable.
Always nice to know. I just need to figure out how much write current
the Panasonic chip provides (all I have is a short-form datasheet, which
doesn't include any figures or calculations, just a pinout and a few
electrical specs).
Although
whether I've got enough clearance to shove a PDIP chip in there
without fouling the drive belt or flywheel is another question entirely...
Do you have to pack it into the same volume as the original drive?
29 Nov
29 Nov
8:49 p.m.
The
semi-0discrete solution ('74 as a divide by 2 and a handful of
transistors) turns up all over the place. It's easy to design and reliable.
Always nice to know. I just need to figure out how much write current
the Panasonic chip provides (all I have is a short-form datasheet, which
doesn't include any figures or calculations, just a pinout and a few
electrical specs).
28 Nov
28 Nov
10:37 a.m.
Hi guys,
As you might have gathered from my other thread, I'm toying with the
idea of building a replacement controller card for my two dead
double-sided Amstrad disc drives. Ideally I'd like to use as many
"standard" parts as possible to ease debugging and future repair. I'll
probably use a GAL to handle head-stepping and the write protect
lockout, and LSTTL for the bus interface.
Hmmm.. I am not sure that easing future repair and GALs (or any other PLD) go
together. Well, not unless you etch the logic equations into the PCB :-)
I've got my hands on a couple of Texas Instruments
MC3470A floppy disc
read amplifier ICs, and the Apple Disk II schematics suggest how this
can be tied to a drive mechanism. What I'd like to do is look at a few
other drives which use this chip to make sure I understand everything
correctly (a double-sided drive with a single 3470A would be handy). If
I'm being honest, the TI datasheet for the chip is utter crap...
'My' schematics for the HP9826/36 (available from http://www.hpmuseum.org/)
include schematics of the MPI drive, which is a double-sided 5.25" unit using
the 3470. There are also schematics of a Tandon-based drive using a read
amplifier based on the uA733 and NE592 amplifiers (another common circuit -- it
works well, but do take care with the layout!).
-tony
5154
days inactive
5155
days old
11 comments
4 participants
participants (4)
-
aek＠bitsavers.org -
ard＠p850ug1.demon.co.uk -
cclist＠sydex.com -
classiccmp＠philpem.me.uk