12 Dec
2015
12 Dec
'15
5:01 p.m.
I have four 1541's now, two of which seem to init OK, and two of which sit
there with the drive spindle constantly spinning and the access LEDs lit,
the latter behavior which I believe indicates a multitude of possible
microcontroller-related faults...
I've done the obvious, reseating socketed ICs, checking the +12V and +5
rails, and checking the on-board CPU reset line. Does anyone have any tips
for what's best to try next?
I read somewhere that ROM faults aren't unheard of, ditto with 6522 VIA
failures. Also the 74ls14 at UA1 (particularly if someone's unplugged the
drive from the host with power on), but I think that affects CPU reset,
which appears to be working in my units.
(is the firmware interchangeable between different board revisions? i.e.
can I swap in firmware from one of the drives that appear to init OK to
rule that out as a fault, or do I have to worry about matching PCB revisions?)
Also, has there been a worse external drive in the history of floppy
storage? Slow, complicated, expensive and unreliable seems to have most
bases covered ;-)
cheers
Jules
12 Dec
12 Dec
6:27 p.m.
On 13/12/2015 2:01 p.m., Jules Richardson wrote:
I have four 1541's now, two of which seem to init
OK, and two of which
sit there with the drive spindle constantly spinning and the access
LEDs lit, the latter behavior which I believe indicates a multitude of
possible microcontroller-related faults...
I've done the obvious, reseating socketed ICs, checking the +12V and
+5 rails, and checking the on-board CPU reset line. Does anyone have
any tips for what's best to try next?
Locked on LEDs and spindle motor would suggest VIA to me, though it's
been many years since I've looked at one. First thing I'd check is that
there's activity on the address/data buses. If things are going on, and
there are 6522s at hand, I'd try a quick swap.
I read somewhere that ROM faults aren't unheard
of, ditto with 6522
VIA failures. Also the 74ls14 at UA1 (particularly if someone's
unplugged the drive from the host with power on), but I think that
affects CPU reset, which appears to be working in my units.
Can you swap ROMs in from one of the booting drives? Worth a go, though
I've never seen a ROM failure.
(is the firmware interchangeable between different
board revisions?
i.e. can I swap in firmware from one of the drives that appear to init
OK to rule that out as a fault, or do I have to worry about matching
PCB revisions?)
I wouldn't have thought it'd matter, though maybe the 1541 II was
different enough.
Also, has there been a worse external drive in the
history of floppy
storage? Slow, complicated, expensive and unreliable seems to have
most bases covered ;-)
I sort of agree, though I have to say that of my pile of about 30 1541s
and 1541 IIs, all but two ran up fine on first test, after having been
in storage for 25 years. One had a bad 6522, and the other had an
alignment issue. They were all still amazingly slow though!
Cheers,
Mike
13 Dec
13 Dec
7:48 a.m.
On 12/12/2015 08:27 PM, Mike van Bokhoven wrote:
Can do, I just wasn't sure if the ROMs are tied to a particular board
revision (so I need rev A ROMs for a rev A board etc.). Looking at my Rev A
and Rev C boards though it would seem that they have the same part numbers,
and I assume they'd be different if the contents were.
:-) Gotta love 'em, just for how quirky they are - I really don't know
what Commodore were thinking when they designed them (even if they'd worked
as intended with a higher transfer rate).
I do have a drive unit for a Research Machines system designed along
similar lines (on-board CPU and serial connection to host), but at least
that is a dual-drive unit and uses an off-the-shelf FDC.
cheers
Jules
Locked on LEDs and spindle motor would suggest VIA to
me, though it's been
many years since I've looked at one. First thing I'd check is that there's
activity on the address/data buses.
I don't have a lot of test gear, but I suppose poking at A0 with my DMM on
frequency setting should give a sign of whether anything's going on...
If things are going on, and there are 6522s at hand,
I'd try a quick swap.
I can certainly borrow ones from one of the good drives - I suppose given
that the PSU rails are OK, it's unlikely that there's anything in the
faulty drives that will damage them.
I read
somewhere that ROM faults aren't unheard of, ditto with 6522 VIA
failures. Also the 74ls14 at UA1 (particularly if someone's unplugged the
drive from the host with power on), but I think that affects CPU reset,
which appears to be working in my units.
Can you swap ROMs in from one of the booting drives? Worth a go, though
I've never seen a ROM failure. Also, has
there been a worse external drive in the history of floppy
storage? Slow, complicated, expensive and unreliable seems to have most
bases covered ;-)
I sort of agree, though I have to say that of my pile of about 30 1541s and
1541 IIs, all but two ran up fine on first test, after having been in
storage for 25 years. One had a bad 6522, and the other had an alignment
issue. They were all still amazingly slow though!
8:22 a.m.
On 12/13/2015 09:22 AM, tony duell wrote:
Yes, I've certainly seen that on the schematics, and was considering them
as a likely failure too - after opening cases though, my drives all seem to
have a single larger SRAM (even the Rev A board, so that's presumably "Rev
A of the type that doesn't use 2114's" :-)
While 2114's could be somewhat unreliable, I think the larger (6116 etc.)
SRAMs are generally OK; while a fault there is a possibility, I'm thinking
that the failure is likely elsewhere.
(Aside: Why on Earth do people scan schematics as bi-level images? I've not
found a good 1541 schematic yet - the ones out there are all high
resolution, but the depth reduction renders component information unreadable)
Jules
I've done the obvious, reseating socketed ICs, checking the +12V and +5
rails, and checking the on-board CPU reset line. Does anyone have any tips
for what's best to try next?
Didn't at least some versions use 2114 RAM chips? If so, then check/change
those first 
10:22 a.m.
Yes, I've certainly seen that on the schematics, and was considering them
as a likely failure too - after opening cases though, my drives all seem to
have a single larger SRAM (even the Rev A board, so that's presumably "Rev
A of the type that doesn't use 2114's" :-)
Yes, later versions used a single 6116 RAM (which as you said is not likely
to be the problem).
I have had a ROM failure in a 1541. I am pretty sure mine now has an EPROM
in a home-made kludgeboard (to move the enables round, etc) in it.
I think at this point you have to do real tests, there is no stock fault. I do
have the Sams book on the 1541 somewhere, but IIIRC it suggests swapping
out the socketed devices (ROMs, RAM(s), VIA, CPU, etc. If that doesn't help
then you pull the CPU and force the logic levels of the address pins to check
the address decoder circuitry (but the address decoder almost never fails).
I am pretty sure the firmware is not changed for different PCBs. I would keep
ROMs in pairs just in case there is a different revision and it matters, but I think
you could put the ROMs from the defective drive into a good one and see what
the LED does.
-tony
10:42 a.m.
Had lots of 6522 and ROM failures here. I'd begin with rom replacement
Enviado do meu Tele-Movel
Em 13/12/2015 16:34, "tony duell" <ard at p850ug1.demon.co.uk> escreveu:
Yes, I've certainly seen that on the schematics, and was considering them
as a likely failure too - after opening cases though, my drives all seem
to
have a single larger SRAM (even the Rev A board,
so that's presumably
"Rev
A of the type that doesn't use
2114's" :-)
Yes, later versions used a single 6116 RAM (which as you said is not likely
to be the problem).
I have had a ROM failure in a 1541. I am pretty sure mine now has an EPROM
in a home-made kludgeboard (to move the enables round, etc) in it.
I think at this point you have to do real tests, there is no stock fault.
I do
have the Sams book on the 1541 somewhere, but IIIRC it suggests swapping
out the socketed devices (ROMs, RAM(s), VIA, CPU, etc. If that doesn't help
then you pull the CPU and force the logic levels of the address pins to
check
the address decoder circuitry (but the address decoder almost never fails).
I am pretty sure the firmware is not changed for different PCBs. I would
keep
ROMs in pairs just in case there is a different revision and it matters,
but I think
you could put the ROMs from the defective drive into a good one and see
what
the LED does.
-tony
8:41 a.m.
On 12/13/2015 7:22 AM, tony duell wrote:
Instead of just randomly replacing stuff, check here:
http://personalpages.tds.net/~rcarlsen/cbm.html
Ray Carlsen's articles are the gold standard in Commodore hardware
diagnostic and repair information.
g.
I've done the obvious, reseating socketed ICs, checking the +12V and +5
rails, and checking the on-board CPU reset line. Does anyone have any tips
for what's best to try next?
Didn't at least some versions use 2114 RAM chips? If so, then check/change
those first
14 Dec
14 Dec
9:47 a.m.
On Sun, Dec 13, 2015 at 10:22 AM, tony duell <ard at p850ug1.demon.co.uk> wrote:
I don't think any of the single-disk CBM drives used 2114s (but I
agree - those are favorite suspects when present). I'm pretty sure
all the units I've worked on have 6116-type 2K SRAMs or perhaps 6264s.
I'd have to go back and check parts lists, but the older dual-drive
units might have had 2114s as the shared memory between the two
processors.
-ethan
I've done the obvious, reseating socketed ICs, checking the +12V and +5
rails, and checking the on-board CPU reset line. Does anyone have any tips
for what's best to try next?
Didn't at least some versions use 2114 RAM chips? If so, then check/change
those first
9:51 a.m.
I don't think any of the single-disk CBM drives used 2114s (but I
agree - those are favorite suspects when present). I'm pretty sure
all the units I've worked on have 6116-type 2K SRAMs or perhaps 6264s.
I'd have to go back and check parts lists, but the older dual-drive
units might have had 2114s as the shared memory between the two
processors.
At least one version of the 8050 uses 8 off 2114s as the shared memory
I had to replace the lot in mine....
-tony
12:38 p.m.
They can be easily replaced with newer chips. But any work with it should
make you think about building a sd2iec hehehehe
(I gotta build one for me...)
Enviado do meu Tele-Movel
Em 14/12/2015 16:09, "tony duell" <ard at p850ug1.demon.co.uk> escreveu:
I don't think any of the single-disk CBM drives used 2114s (but I
agree - those are favorite suspects when present). I'm pretty sure
all the units I've worked on have 6116-type 2K SRAMs or perhaps 6264s.
I'd have to go back and check parts lists, but the older dual-drive
units might have had 2114s as the shared memory between the two
processors.
At least one version of the 8050 uses 8 off 2114s as the shared memory
I had to replace the lot in mine....
-tony
3:32 p.m.
Do remember when ordering 2114's that these are all NOS units
and just about as likely to be bad as the ones in your unit.
I don't know of any surplus place that has the ability to test them.
Most any of the places that I've dealt with will replace bad ones
but if dealing by mail order, it may not be worth the hassle.
Order a few extras.
Dwight
5:48 p.m.
If you need 6522 chip(s) please contact me off-list - I have more than I
need, would be happy to pass one along to a needy drive..
On Mon, Dec 14, 2015 at 5:32 PM, dwight <dkelvey at hotmail.com> wrote:
Do remember when ordering 2114's that these are all NOS units
and just about as likely to be bad as the ones in your unit.
I don't know of any surplus place that has the ability to test them.
Most any of the places that I've dealt with will replace bad ones
but if dealing by mail order, it may not be worth the hassle.
Order a few extras.
Dwight
6:17 p.m.
On Mon, Dec 14, 2015 at 6:32 PM, dwight <dkelvey at hotmail.com> wrote:
Do remember when ordering 2114's that these are
all NOS units
and just about as likely to be bad as the ones in your unit.
Too true. NOS doesn't mean it's working.
I don't know of any surplus place that has the
ability to test them.
Or the interestg.
Most any of the places that I've dealt with will
replace bad ones
but if dealing by mail order, it may not be worth the hassle.
Order a few extras.
That reminds me, I need to make a quickie little microcontroller-based
2114 tester. I have tubes of the chips (NOS from 1982 - we used them
in the COMBOARD 1) and no confidence of which ones work and which ones
do not. I'm sure 98% or more work. Which 98% is the question.
It's not hard to make one with a <$10 microcontroller. It's just a
question of sitting down and building it.
-ethan
3295
days inactive
3297
days old
14 comments
8 participants
participants (8)
-
alexandre.tabajara＠gmail.com -
ard＠p850ug1.demon.co.uk -
dkelvey＠hotmail.com -
drlegendre＠gmail.com -
ethan.dicks＠gmail.com -
geneb＠deltasoft.com -
jules.richardson99＠gmail.com -
mike＠fenz.net