4 Nov
2010
4 Nov
'10
5:54 p.m.
OK, this is seriously weird.
I have two Amstrad 3-inch disc drives: an EME-156, and an EME-231.
Both drives will spin up, select and generally "work". For varying
values of "work". On both, the "activity" (selected, whatever) LED is
stuck on at half-brightness, but switches to full brightness when the
drive is selected.
For reference: both drives are showing the exact same symptoms.
I can select the drive, spin the motor, and seek around the disc.
Writing seems to work (more or less) -- when WR GATE goes low, the coils
of the head are driven to ~10V with smaller positive and negative pulses
(sort of like an exponential curve, synchronised with the falling edge
of WRDATA). When WR GATE goes high again, the head voltage falls back to
about 2V.
When I try and read anything, the head shows no response whatsoever. No
pulses on either side of the head at any amplitude (measured with a 1:10
probe on a Tek TDS2024B). Similarly, RD DATA shows no pulses either,
except for a brief glitch (2.8us or so) when the drive is selected.
The drive controller ASIC also gets rather hot.. like, too hot to touch.
I've confirmed that the power is correct -- 12V and 5V have been applied
to the correct pins.
I'm using "new old stock" Amsoft discs which I bought from an ebay
seller. No idea if they're good or bad, but they arrived in
shrink-wrapped side-opening plastic boxes and were apparently "Made in
Japan."
My current suspects are:
- Discs. The discs aren't actually magnetically coated in any way, or
the coating has failed in some way. Catch is, I don't have a known-good
disc to test with.
- Heads. Dirty, out of alignment or otherwise completely pooched.
Can anyone suggest some things I could check, or are these drives likely
to be toast?
Thanks,
--
Phil.
classiccmp at philpem.me.uk
http://www.philpem.me.uk/
5 Nov
5 Nov
3:53 p.m.
OK, this is seriously weird.
I have two Amstrad 3-inch disc drives: an EME-156, and an EME-231.
I have scheamtics that might be related. The main ASIC seems to be an
M51017Ap in both cases.
Both drives will spin up, select and generally "work". For varying
values of "work". On both, the "activity" (selected, whatever) LED is
stuck on at half-brightness, but switches to full brightness when the
drive is selected.
I an't see why it would do that. According to the schematic, the LED is
driven by pin 44 of the ASIC, have you stuck a 'scope o nther to see if
the half-brightness is actually a square wave?
For reference: both drives are showing the exact same symptoms.
If getting the power lines backwads does do damage, is it psosigle that
the otehr drive was damaged before you got it?
I can select the drive, spin the motor, and seek around the disc.
Writing seems to work (more or less) -- when WR GATE goes low, the coils
of the head are driven to ~10V with smaller positive and negative pulses
(sort of like an exponential curve, synchronised with the falling edge
of WRDATA). When WR GATE goes high again, the head voltage falls back to
about 2V.
I am not convinced that's right. What is the centre-tap of the head doing
(ASIC pin 18 for side 0 and pin 20 for side 1)?The head should be a
low-resistnce device and if you're getting 11V across it, something is wrong.
When I try and read anything, the head shows no response whatsoever. No
pulses on either side of the head at any amplitude (measured with a 1:10
probe on a Tek TDS2024B). Similarly, RD DATA shows no pulses either,
except for a brief glitch (2.8us or so) when the drive is selected.
Anything on TP1,2 on the drive PCB? Those are the differential outputs of
the read amplifier, used for head alignment.
FWIW, the peripheral components around that bit of the ASIC look very
like the ones used on a MC3470.
- Discs. The discs aren't actually magnetically
coated in any way, or
the coating has failed in some way. Catch is, I don't have a known-good
disc to test with.
I think that's unlikely
- Heads. Dirty, out of alignment or otherwise
completely pooched.
Well, clean the heads!. Check the DC resistance (should be very low
between the 4 wires going to a head). Out of alignment won't make a
difference, you are reading and writing with the same head, in the same
position.
Can anyone suggest some things I could check, or are
these drives likely
to be toast?
Do you have the schemciats?
-tony
6:03 p.m.
On 05/11/10 22:53, Tony Duell wrote:
I haven't checked the centre tap. I've just found a test/adjustment
procedure in the Amstrad manual which suggests that the ASIC tries to
float the head in relation to ground, but that the absolute voltage over
the head is zero unless it's in use. That might explain the strange
measurements.
That's good to hear...
Done that (with a cotton bud and Servisol IPA-170 Isopropyl Alcohol). No
change in behaviour, no dirt on the cotton bud I used to clean them.
I have the PCW8256/8512 service manual, the 8256-only service manual and
the 9512 service manual as PDFs, which include the drive schematics.
Thanks,
--
Phil.
classiccmp at philpem.me.uk
http://www.philpem.me.uk/
I have scheamtics that might be related. The main ASIC
seems to be an
M51017Ap in both cases.
Yep. I think I might have the same schematics. Amstrad PCW 8256/8512
Service Manual?
I an't see why it would do that. According to the
schematic, the LED is
driven by pin 44 of the ASIC, have you stuck a 'scope o nther to see if
the half-brightness is actually a square wave?
Not yet, no.
If getting the power lines backwads does do damage, is
it psosigle that
the otehr drive was damaged before you got it?
I'd go with "probable". They were sold to me as "spares or repair, no
idea if they even work."
I can select
the drive, spin the motor, and seek around the disc.
Writing seems to work (more or less) -- when WR GATE goes low, the coils
of the head are driven to ~10V with smaller positive and negative pulses
(sort of like an exponential curve, synchronised with the falling edge
of WRDATA). When WR GATE goes high again, the head voltage falls back to
about 2V.
I am not convinced that's right. What is the centre-tap of the head doing
(ASIC pin 18 for side 0 and pin 20 for side 1)?The head should be a
low-resistnce device and if you're getting 11V across it, something is wrong. Anything on TP1,2 on the drive PCB? Those are the
differential outputs of
the read amplifier, used for head alignment.
Nope, nothing on either.
- Discs.
The discs aren't actually magnetically coated in any way, or
the coating has failed in some way. Catch is, I don't have a known-good
disc to test with.
I think that's unlikely - Heads.
Dirty, out of alignment or otherwise completely pooched.
Well, clean the heads!. Check the DC resistance (should be very low between
the 4 wires going
to a head).
On the to-do list (it's a bit late to be getting out the DMM at this
time of day!)
Out of alignment won't make a
difference, you are reading and writing with the same head, in the same
position.
Good point.
Can anyone
suggest some things I could check, or are these drives likely
to be toast?
Do you have the schemciats? 
6:12 p.m.
On 6 Nov 2010 at 1:03, Philip Pemberton wrote:
Yep. I think I might have the same schematics. Amstrad
PCW 8256/8512
Service Manual?
Got an 8256 here (those things should be as common as fleas in the
UK, I'd think. Getting a 120V 60Hz one here was a little less easy).
I've installed two 720K (Teac FD235F) 3.5" drives, both to get
something that will work with available media and also to get more
storage. The Joyce uses single-sided 3" disks. With a simple one-
byte change in the boot sector, I can use larger 720K DSDD floppies.
But anyway, back to your problem. About all I can verify right now
without hauling out the scope and screwdriver is that the activity
LED is normally about 1/3 brightness when not selected and that it
blinks briefly when power is first applied.
--Chuck
6 Nov
6 Nov
2:14 p.m.
On 6 Nov 2010 at 1:03, Philip Pemberton wrote:
The service manuals I have to hand make no reference to a 120V model. The
PSU is a switcher, producing 350V DC (or so) from our 240V mains, and it
would be easy to modify that to a voltage doubler for 120V mains input,
but the smoothing capacitor is just a single componentm not 2 in series,
so the changes would be more than in some other supplies.
Yep. I think I might have the same schematics.
Amstrad PCW 8256/8512
Service Manual?
Got an 8256 here (those things should be as common as fleas in the
UK, I'd think. Getting a 120V 60Hz one here was a little less easy).
I've installed two 720K (Teac FD235F) 3.5" drives, both to get
something that will work with available media and also to get more
storage. The Joyce uses single-sided 3" disks. With a simple one-
Actually, the standard configuration has drive 0 as a single-head 40
cylinder drive taking 'flippy' disks and the optional drive 1 as a
double-head 80 cylidner drive. The latter was much the same logically as
a noraml 720K drive.
-tony
3:02 p.m.
On 6 Nov 2010 at 21:14, Tony Duell wrote:
The service manuals I have to hand make no reference
to a 120V model.
Mine definitely says 120V 60Hz 1.1A on the back plate. Made in
Korea, FCC ID A7R9K2PC-4A. No mention of Amstrad UK, but rather
Amstrad International (USA) Inc., 3202 Doolittle Drive, Northbrook,
IL, 60062. The curious thing is that the label refers to it as a
"monitor". No reference to it being a computer of any sort.
Actually, the standard configuration has drive 0 as a
single-head 40
cylinder drive taking 'flippy' disks and the optional drive 1 as a
double-head 80 cylidner drive. The latter was much the same logically
as a noraml 720K drive.
Yes, but that's useless if what you want to do is boot from a 720K
and have 2x720K of floppy storage online.
--Chuck
7 Nov
7 Nov
11:47 a.m.
On 6 Nov 2010 at 21:14, Tony Duell wrote:
OK, I beleive you :-).
The service manuals I have to hand make no
reference to a 120V model.
Mine definitely says 120V 60Hz 1.1A on the back plate. Made in Korea, FCC ID A7R9K2PC-4A. No mention of Amstrad UK,
but rather
Amstrad International (USA) Inc., 3202 Doolittle Drive, Northbrook,
IL, 60062. The curious thing is that the label refers to it as a
"monitor". No reference to it being a computer of any sort.
Odd. I assume the main logic board is much the same as the UK version,
but the PSU/monitor board is different. Probably just differences in the
mains input circuitry.
The UK model uses and STK7308 switching regulator hybrid module. Do you
happen to know if yours does too?
Does it have 60 Hz vertcial sync? In these machines, the video and syncs
are generated bu the Amstrad 40028 gate array (IC 101 on the logic board)
which also contains the trivial bootstrap ROM (thre is no seprate ROM or
EPROM on the logic board). So I would assume that if the sync rate is
changed, this chip has ot be something different.
Actually, the standard configuration has drive 0
as a single-head 40
cylinder drive taking 'flippy' disks and the optional drive 1 as a
double-head 80 cylidner drive. The latter was much the same logically
as a noraml 720K drive.
Yes, but that's useless if what you want to do is boot from a 720K
and have 2x720K of floppy storage online.
3:32 p.m.
On 7 Nov 2010 at 19:47, Tony Duell wrote:
The UK model uses and STK7308 switching regulator
hybrid module. Do
you happen to know if yours does too?
Does it have 60 Hz vertcial sync? In these machines, the video and
syncs are generated bu the Amstrad 40028 gate array (IC 101 on the
logic board) which also contains the trivial bootstrap ROM (thre is no
seprate ROM or EPROM on the logic board). So I would assume that if
the sync rate is changed, this chip has ot be something different.
When I get a chance, I'll get a screwdriver out and check it, but my
memory is that I used the standard (European) schematic to locate and
replace one of those silly little circuit protectors in the power
supply--with a picofuse of the same rating.
That Amstrad gate array is the principal thing I really don't like
about the machine. Undocumented and probably irreplaceable.
Otherwise, a decent graphics display, lots of RAM, what's not to
like?
--Chuck
8 Nov
8 Nov
12:28 p.m.
When I get a chance, I'll get a screwdriver out
and check it, but my
memory is that I used the standard (European) schematic to locate and
replace one of those silly little circuit protectors in the power
supply--with a picofuse of the same rating.
What's silly about them (if picofuses are sensible)?
That Amstrad gate array is the principal thing I really don't like
about the machine. Undocumented and probably irreplaceable.
You and me both :-). To be fair, Amstrad did sell service manuals for the
machine at a sensible price. (THe one I have has 'Price \pounds 8.00'
printed on the cover), and spare parts [1] were available while the machine
was in production. So about 20 years ago you could have got a
replacement gate array chip with no problems. Which is more than you can
say for a lot of manufacturers.
[1] I had no problem getting a replacement flyback transformer for an
Amstrad PC monitor, for example.
Otherwise, a decent graphics display, lots of RAM,
what's not to
like?
I still think the nicest CP/M machine I've seen is the Epson QX10.
-tony
2 p.m.
On 8 Nov 2010 at 20:28, Tony Duell wrote:
What's silly about them (if picofuses are
sensible)?
Maybe they're common on European gear, but the Joyce was the first
time I'd ever seen one--they're just not what I'd consider to be a
"universal" part.
I still think the nicest CP/M machine I've seen is
the Epson QX10.
That's a hard call--if we're talking about Z80-based CP/M machines
(not x86 or 68K), some of the other late Japanese machines were
pretty impressive. The QX10 was interesting to me mostly in its non-
CPM (TP/M?) software.
But I won't get behind any particular system--there were too many of
them.
--Chuck
9 Nov
9 Nov
12:14 p.m.
What's
silly about them (if picofuses are sensible)?
Maybe they're common on European gear, but the Joyce was the first
time I'd ever seen one--they're just not what I'd consider to be a
"universal" part. I still think the nicest CP/M machine I've
seen is the Epson QX10.
That's a hard call--if we're talking about Z80-based CP/M machines (not x86 or 68K), some of the other late Japanese
machines were
pretty impressive. The QX10 was interesting to me mostly in its non-
I did say 'nicest .. that I've seen' :-). The QX10 seems to have some
pretty nice hardware. Mostly stnadard chips too (alas mine has the later
video board with the gate array chip :-() and those lovelly voice coil
floppy drives.
I am not familiar with many other Japanese CP/M machines. What was nice
about them?
-tony
6 Nov
6 Nov
1:03 p.m.
On 05/11/10 22:53, Tony Duell wrote:
As it happens, yes. The scheamtics of at least the single-head drive
appear in other Amstrad service manuals too.
I have scheamtics that might be related. The main
ASIC seems to be an
M51017Ap in both cases.
Yep. I think I might have the same schematics. Amstrad PCW 8256/8512
Service Manual? I an't see why it would do that. According to
the schematic, the LED is
driven by pin 44 of the ASIC, have you stuck a 'scope o nther to see if
the half-brightness is actually a square wave?
Not yet, no. If getting the power lines backwads does do
damage, is it psosigle that
the otehr drive was damaged before you got it?
I'd go with "probable". They were sold to me as "spares or repair, no
idea if they even work." I haven't checked the centre tap. I've just
found a test/adjustment
procedure in the Amstrad manual which suggests that the ASIC tries to
float the head in relation to ground, but that the absolute voltage over
the head is zero unless it's in use. That might explain the strange
measurements.
The scehamtic of the double-head drive gives the pin votlages for the
ASIC. It owuld appear those are the votlages you'd get if the drive is
powered, but not conencted to a controller (all inputs pulled high). May
be worth checking the ones around the read section.
Anything on TP1,2 on the drive PCB? Those are the
differential outputs of
the read amplifier, used for head alignment.
Nope, nothing on either. - Heads. Dirty, out of alignment or otherwise
completely pooched.
Well, clean the heads!. Check the DC resistance (should be very low
between the 4 wires going
to a head).
On the to-do list (it's a bit late to be getting out the DMM at this
time of day!) Good point.
Do you have
the schemciats?
I have the PCW8256/8512 service manual, the 8256-only service manual and
the 9512 service manual as PDFs, which include the drive schematics.
5:46 p.m.
On 06/11/10 20:03, Tony Duell wrote:
It does sound like head or ASIC problems...
:(
Both of which would require a supply of Unobtainium and Magic Smoke (not
to mention the fabled Magic Smoke Inserter) to repair or replace. I dare
say these drives are going to become mechanical parts mules...
I have a Fluke 25 that lives in the cupboard, but I'm too lazy to wheel
my chair over to the cupboard to get it :)
The Agilent is downstairs. I'm trying to figure out why DiscFerret #2 is
having power-backfeed issues (the SENSE line from the USB port is
backfeeding onto +3V3 and Blowing Stuff Up.. and I've ruled out a
hard-short).
Aha...This remionds me ao a character I saw at a
couple of radio rallies
many yearsa ago. He had a pile of working hard drives. And another pile
of ''untensted' hard drives. Hmmm...
Read as: "I've tested all of these, but I'm selling the broken ones as
'untested'."
Then you get the fiercely honest ones -- I approached a seller at a
radio rally, and asked about some VRLA batteries he was selling.
"Ah, they're only a couple of months old -- brand new. Serial number's
here, and manufacture date here. Charged them up half-way the other day,
hang on, I've got a multimeter here somewhere.... Oh, and I'll wire a
car headlamp over one if you'd like to see it running under load..."
I wonder if theprevious owner plugged them into a PC
power supply and
then discovered that either they didn't work, or somebody told him he'd
probably done some daamge, so you ended up with them
Indeed. I've got a lead on a pair of double-sided drives from a PCW9512.
They're basically the same drive design, but without the write protect
sense pin (optical sensing!), a white front panel and a slightly
different mounting cage.
Fingers crossed...
Anything on TP1,2 on the drive PCB? Those are the
differential outputs of
the read amplifier, used for head alignment.
Nope, nothing on either. On the to-do
list (it's a bit late to be getting out the DMM at this
time of day!)
You put the DMM away? That suprises me ;-) I think I would start by checking the pin voltages on
the ASIC, and in
particular see what the head cnetre-taps are doing.
-- In write mode --
I'm seeing around 2V on all three pins of the head -- 'top', 'bottom'
and the centre tap. If I switch to AC coupling, 200mV/div, I see what
looks like a very noisy sine wave on the centre tap, and then the coils
have the same noisy sine, plus an alternating pulse. If I subtract the
CT from either coil, I can see the magnetic pulses. These look like the
0-180deg (positive going) part of a sinewave, then a gap, then the
180-360deg (negative going) part.
-- In read mode --
Not a sausage at 200mV. Dropping to 50mV/div, I can see a heck of a lot
of noise, but nothing that looks remotely like the 125kHz data I fed it.
There's nothing interesting on TP1, TP2 and TP3 either. With drive stopped:
TP1 -- 258.45kHz sine, ~200mV amplitude
TP2 -- same sine wave, 180 degrees out of phase
TP3 -- ground
Adding the two channels together gets me a flat line with a bit of
noise. Subtracting (per the instructions in the service manual) gets me
a 400mV pk-pk sine...
It's as if the ASIC were oscillating internally...
--
Phil.
classiccmp at philpem.me.uk
http://www.philpem.me.uk/
5:54 p.m.
On 07/11/10 00:46, Philip Pemberton wrote:
There's nothing interesting on TP1, TP2 and TP3
either. With drive stopped:
TP1 -- 258.45kHz sine, ~200mV amplitude
TP2 -- same sine wave, 180 degrees out of phase
TP3 -- ground
I should also mention that when writing, in MATH ADD mode I get a ~1.5V
(2V if you count the undershoot) square wave. Source signals (TP1 and
TP2) are oddly shaped spikes... almost a square wave but.. not quite.
Stop writing and the signal vanishes.
--
Phil.
classiccmp at philpem.me.uk
http://www.philpem.me.uk/
7 Nov
7 Nov
12:06 p.m.
On 07/11/10 00:46, Philip Pemberton wrote:
That could well be right. The read amplaifier is tryign to process the
signal being sent to the head in write mode.
I assume you've tried both sides of the doubler-head drive...
-tony
There's nothing interesting on TP1, TP2 and
TP3 either. With drive stopped:
TP1 -- 258.45kHz sine, ~200mV amplitude
TP2 -- same sine wave, 180 degrees out of phase
TP3 -- ground
I should also mention that when writing, in MATH ADD mode I get a ~1.5V
(2V if you count the undershoot) square wave. Source signals (TP1 and
TP2) are oddly shaped spikes... almost a square wave but.. not quite.
Stop writing and the signal vanishes.
12:51 p.m.
On 07/11/10 20:06, Tony Duell wrote:
That could well be right. The read amplaifier is
tryign to process the
signal being sent to the head in write mode.
I assume you've tried both sides of the doubler-head drive...
Indeed I have.
--
Phil.
classiccmp at philpem.me.uk
http://www.philpem.me.uk/
6 Nov
6 Nov
8 p.m.
On Sun, 7 Nov 2010, Philip Pemberton wrote:
Read as: "I've tested all of these, but
I'm selling the broken ones as
'untested'."
close
The "untested" ones are the ones that his "technicians" were
unable to repair, and may be missing parts that his "technicians"
"board-swapped" into other machines.
The "tested" ones are those where the motor turned and the LED lit.
Or did so after his "technicians" swapped parts.
You expected a surface scan?
--
Grumpy Ol' Fred cisin at xenosoft.com
7 Nov
7 Nov
12:45 a.m.
On 11/6/2010 5:46 PM, Philip Pemberton wrote:
On 06/11/10 20:03, Tony Duell wrote:
There's a white haired "gentleman" who still shows up at the
TRW/Northrup Grumman and ACP swap meets who has been around for 20
years. I recall when he was starting out that he would be rude and
abusive to his mob of kids, and go around and buy absolute crap drives.
I discovered him selling them a few months later as "untested" at swap
meets with his kids running the booth. I ran across him as he moved up
in the ripoff game buying in bulk from guys who would bring clearly
labeld "bad" and the same packaging would show up as working or
otherwise with the drives scrubbed of the markings.
When I found him at that time, I asked whether the prices were at scrap
aluminum or higher.
I do know he was banned from ACP, but still see him ripping of people at
TRW. You gotta be careful who you buy from.
Aha...This remionds me ao a character I saw at a
couple of radio rallies
many yearsa ago. He had a pile of working hard drives. And another pile
of ''untensted' hard drives. Hmmm...
Read as: "I've tested all of these, but I'm selling the broken ones as
'untested'."
Then you get the fiercely honest ones -- I approached a seller at a
radio rally, and asked about some VRLA batteries he was selling.
"Ah, they're only a couple of months old -- brand new. Serial number's
here, and manufacture date here. Charged them up half-way the other
day, hang on, I've got a multimeter here somewhere.... Oh, and I'll
wire a car headlamp over one if you'd like to see it running under
load..."
12:04 p.m.
On 06/11/10 20:03, Tony Duell wrote:
I suspect, and suspected at the time, that you areperfectly correct :-)
Before we get another flamewar, I will point out that just about every
seller I've dealt with at radio rallies has been extremely honest. If
something doesn't work, they say so. If it's missing important parts,
they say that too.
I feel that the enthusiast-type radio amateurs are more likely to be
totally honest than the traders (the former don't want to cheat a fellow
ham), but to be fair I've never had what I consider a bad deal from any
of them.
Aha...This remionds me ao a character I saw at a
couple of radio rallies
many yearsa ago. He had a pile of working hard drives. And another pile
of ''untensted' hard drives. Hmmm...
Read as: "I've tested all of these, but I'm selling the broken ones as
'untested'." Then you get the fiercely honest ones -- I approached
a seller at a
radio rally, and asked about some VRLA batteries he was selling.
"Ah, they're only a couple of months old -- brand new. Serial number's
here, and manufacture date here. Charged them up half-way the other day,
hang on, I've got a multimeter here somewhere.... Oh, and I'll wire a
car headlamp over one if you'd like to see it running under load..."
I've met plenty of sellers who are happy to let you remove covers and
have a look inside. And I rememebr a chap with a box of assorted valves
who provided a multimeter so you could at least check if the heater was
OK.
I wonder if theprevious owner plugged them into a
PC power supply and
then discovered that either they didn't work, or somebody told him he'd
probably done some daamge, so you ended up with them
Indeed. I've got a lead on a pair of double-sided drives from a PCW9512.
They're basically the same drive design, but without the write protect
sense pin (optical sensing!), a white front panel and a slightly
different mounting cage.
Fingers crossed...
Indeed.
On the to-do list (it's a bit late to be getting
out the DMM at this
time of day!)
You put the DMM away? That suprises me ;-) I'm seeing around 2V on all three pins of the head
-- 'top', 'bottom'
and the centre tap. If I switch to AC coupling, 200mV/div, I see what
looks like a very noisy sine wave on the centre tap, and then the coils
Trhis is for the selected head? That can't be right!~~~~~~~~~~~~~~~~~~~~
have the same noisy sine, plus an alternating pulse.
If I subtract the
CT from either coil, I can see the magnetic pulses. These look like the
0-180deg (positive going) part of a sinewave, then a gap, then the
180-360deg (negative going) part.
-- In read mode --
Not a sausage at 200mV. Dropping to 50mV/div, I can see a heck of a lot
of noise, but nothing that looks remotely like the 125kHz data I fed it.
There's nothing interesting on TP1, TP2 and TP3 either. With drive stopped:
TP1 -- 258.45kHz sine, ~200mV amplitude
TP2 -- same sine wave, 180 degrees out of phase
TP3 -- ground
Adding the two channels together gets me a flat line with a bit of
noise. Subtracting (per the instructions in the service manual) gets me
a 400mV pk-pk sine...
It's as if the ASIC were oscillating internally...
It may well do this. There's probablhy an AGC circuit that cranks up the
gain if there's no signal from the head.
I asusme the head windings are continuous, and that the heads are
actually touching the disk...
But that head CT signal sounds totally wrong.
-tony
1:28 p.m.
On 07/11/10 20:04, Tony Duell wrote:
Huh. This one has a mechanical switch. Maybe the SS drive with the mech
switch is out of a Spectrum +3 or CPC, and the DS drive is out of a
later PCW?
The DS drive shines a light at the disc if memory serves. Black for
"protected" (i.e. tab pushed in), reflected white for "unprotected".
"Hmm."
I'll try and get some scope traces later -- but first I need to find a
USB thumb-drive that's small enough for the DSO to accept...
--
Phil.
classiccmp at philpem.me.uk
http://www.philpem.me.uk/
I suspect, and suspected at the time, that you
areperfectly correct :-)
Well as I said, I'm bidding on another set of drives from a 9512.
Hopefully this seller is a bit more honest.
I'll be fishing the drivebelts out of the two drives once I can prove
they're shot, but I'll probably keep them for mechanical spares (read:
spindle motor).
Before we get another flamewar, I will point out that
just about every
seller I've dealt with at radio rallies has been extremely honest. If
something doesn't work, they say so. If it's missing important parts,
they say that too.
I can say the same -- I've yet to meet a dishonest seller at a radio
rally. A few traders have an over-inflated idea of the value of their
stock, but you see that *everywhere*...
I've met plenty of sellers who are happy to let
you remove covers and
have a look inside.
Seen that, too.
"Interested in that, sir? It won't power up, but it's full of mixers,
filters and other bits. Bound to be something good in there -- just let
me get my screwdriver and I'll show you!"
And I rememebr a chap with a box of assorted valves
who provided a multimeter so you could at least check if the heater was
OK.
One of these days I'm going to have to buy a couple of triodes to play
around with... "valve amplifier" is on the long-term projects list, as
is "fix that X-Y display you got half way through building". I was going
to build a few front-ends for that -- NBTV, CVBS, variable amplitude X-Y
and maybe an FPGA version of Asteroids and Battlezone.
Unfortunately I had issues with noise -- the trace on the DG7/32 shook
like a jelly in an earthquake.
Indeed.
I've got a lead on a pair of double-sided drives from a PCW9512.
They're basically the same drive design, but without the write protect
sense pin (optical sensing!), a white front panel and a slightly
different mounting cage.
Odd that the schematics in the PCW8256 manaul show an optical write
protect sensor (look at the circuitry connected to pin 42 of the ASIC). Ah, my multimeter lives on the workbench.
I haven't got a workbench, but I've appropriated the kitchen table for
DiscFerret manufacturing :)
Though it's currently taking me a day to make and debug each DF. I
really, *really* need to get that down a bit. "Solder paste stencils"
and "a tub of solder paste" are on the shopping list...
At the moment, it's a horrible
analogue thi9ng becuase I still can't find a digitial meter that I like
the look of. If I wanted a bench instrument, it would be easy (although
paying for it wouldn't be..), but a nice handheld one doesn't seem to exist.
Seems you can get a cheap junky DMM or an expensive, all singing all
dancing DMM. There doesn't seem to be much in the middle these days.
It's as if
the ASIC were oscillating internally...
It may well do this. There's probablhy an AGC circuit that cranks up the
gain if there's no signal from the head.
I asusme the head windings are continuous, and that the heads are
actually touching the disk...
But that head CT signal sounds totally wrong.
8 Nov
8 Nov
12:12 p.m.
On 07/11/10 20:04, Tony Duell wrote:
I noticed another ex-PCW drive on Ebay where the seller said he'd tested
in in a Tatung Einstein. Now, it's probably OK, but my Einstein uses the
Hitachi-style drives with a 34 pin edge connector and a 4 pin power
connecotr like the ones used on most 5.25" drives (and with the normal
pinouts). If he's used a standard PC power adapter (the one you use for
3.5" drives) then ...
I suspect, and suspected at the time, that you
areperfectly correct :-)
Well as I said, I'm bidding on another set of drives from a 9512.
Hopefully this seller is a bit more honest.
I'll be fishing the drivebelts out of the two drives once I can prove
they're shot, but I'll probably keep them for mechanical spares (read:
spindle motor).
Etepper motor, maybe the head carriage, stepper driver IC, etc?
Before we get another flamewar, I will point out
that just about every
seller I've dealt with at radio rallies has been extremely honest. If
something doesn't work, they say so. If it's missing important parts,
they say that too.
I can say the same -- I've yet to meet a dishonest seller at a radio
rally. A few traders have an over-inflated idea of the value of their
stock, but you see that *everywhere*... I've met plenty of sellers who are happy to
let you remove covers and
have a look inside.
Seen that, too.
"Interested in that, sir? It won't power up, but it's full of mixers,
filters and other bits. Bound to be something good in there -- just let
me get my screwdriver and I'll show you!" And I rememebr a chap with a box of assorted
valves
who provided a multimeter so you could at least check if the heater was
OK.
One of these days I'm going to have to buy a couple of triodes to play around with... "valve amplifier" is on the
long-term projects list, as
Any particualr design/specification?
is "fix that X-Y display you got half way through
building". I was going
to build a few front-ends for that -- NBTV, CVBS, variable amplitude X-Y
and maybe an FPGA version of Asteroids and Battlezone.
You should try building a clone of the HP1350 'Grpahics translator'. It
takes in a subset of HPGL (over an HPIN interface, I think RS232 was an
option) and outputs XYZ signals to drive a non-strage display. The
original was a lot of random logic (amazingly no processor or
processor-like circuitry). I suppose ti could go in an FPGA, much as I
abhor such things...
Odd that the
schematics in the PCW8256 manaul show an optical write
protect sensor (look at the circuitry connected to pin 42 of the ASIC).
Huh. This one has a mechanical switch. Maybe the SS drive with the mech switch is out of a Spectrum +3 or CPC, and the DS
drive is out of a
later PCW?
The DS drive shines a light at the disc if memory serves. Black for
"protected" (i.e. tab pushed in), reflected white for "unprotected".
I have a flippy 3" drive (single head) which optically detects the notch
in the front edge of the disk hosing and changes the colour of the in-use
LED (red for one side, green for the other). Alas AFAIK that signal is
not avaialble on the itnerface connector.
Ah, my multimeter lives on the workbench.
I haven't got a workbench, but I've appropriated the kitchen table for
DiscFerret manufacturing :)
Though it's currently taking me a day to make and debug each DF. I At the moment,
it's a horrible
analogue thi9ng becuase I still can't find a digitial meter that I like
the look of. If I wanted a bench instrument, it would be easy (although
paying for it wouldn't be..), but a nice handheld one doesn't seem to exist.
Seems you can get a cheap junky DMM or an expensive, all singing all
dancing DMM. There doesn't seem to be much in the middle these days. 
7 Nov
7 Nov
2:19 p.m.
Tony Duell wrote:
I get a bit fanatical about testing anything that I'm parting with - I don't
like the idea of the recipient finding a problem that I could have found
myself and told them about. That's always meant I've had a pile of
"extensively tested" stuff, and a pile of "untested because I haven't
had the
time".
I suppose it's not impossible to encounter that situation at a radio rally,
car boot sale etc. although I must admit that it's tempting (as a buyer) to
merge the sold-as-seen and I'm-going-to-get-screwed categories :-)
The behavior should presumably be pretty much the same for a 3.5" or 5.25"
drive (depending on media coercivity - I don't recall what 3" is now); maybe
it's worthwhile Phil sanity-checking observed readings against a drive that he
does have around, if he doesn't have a known-good 3" drive handy?
cheers
Jules
On 06/11/10
20:03, Tony Duell wrote:
I suspect, and suspected at the time, that you areperfectly correct :-)
Before we get another flamewar, I will point out that just about every
seller I've dealt with at radio rallies has been extremely honest. If
something doesn't work, they say so. If it's missing important parts,
they say that too. Aha...This remionds me ao a character I saw at a
couple of radio rallies
many yearsa ago. He had a pile of working hard drives. And another pile
of ''untensted' hard drives. Hmmm...
Read as: "I've tested
all of these, but I'm selling the broken ones as
'untested'." Ah, my multimeter lives on the workbench.
Mine's always bouncing between house, workshop and garage - I'm rapidly coming
to the conclusion that I need three of everything :-(
I'm seeing
around 2V on all three pins of the head -- 'top', 'bottom'
and the centre tap. If I switch to AC coupling, 200mV/div, I see what
looks like a very noisy sine wave on the centre tap, and then the coils
Trhis is for the selected head? That can't be right!~~~~~~~~~~~~~~~~~~~~
8 Nov
8 Nov
12:23 p.m.
I suppose it's not impossible to encounter that
situation at a radio rally,
car boot sale etc. although I must admit that it's tempting (as a buyer) to
merge the sold-as-seen and I'm-going-to-get-screwed categories :-)
Well, if th sell has 2 piles of similar devices (in the case I mentioned,
IDE hard drives), it's very hard to avoid assuming that 'untested' realy
means 'not working'...
I haev no problem with 'sold as seen' or 'not working' or whatever. And I
do rememebr that in most cases I am buying something that's 30 years old
or more, so I don't expect it to look and work like it's just left the
factory. Howeever, it should not have parts missing (particularly not
complete PCBs or subassemblies) unless I am told about it before I buy
it.
For example, some years ago I bought a non-working HP9820 calcualtor on
E-bay., Indeed it wasn't working, but it was complete (the fault turned
out to be a 7495 chip on the memroy address PCB). I was therefore happy
with my purchase. Had there been PCBs missing, while the description
would still I suppose have been correct, I would have been unhappy.
Ah, my multimeter lives on the workbench.
Mine's always bouncing between house, workshop and garage - I'm rapidly coming
to the conclusion that I need three of everything :-( I'm
seeing around 2V on all three pins of the head -- 'top', 'bottom'
and the centre tap. If I switch to AC coupling, 200mV/div, I see what
looks like a very noisy sine wave on the centre tap, and then the coils
This is for the selected head? That can't be right!~~~~~~~~~~~~~~~~~~~~
5119
days inactive
5123
days old
22 comments
6 participants
participants (6)
-
ard＠p850ug1.demon.co.uk -
cclist＠sydex.com -
cisin＠xenosoft.com -
classiccmp＠philpem.me.uk -
jules.richardson99＠gmail.com -
jws＠jwsss.com