So you have a colour system? An MFX1 or MFX2?
It's certainly at least rev 6 as you've got a wave super.
Then this pin on the CMI 28 got broke....but you could leave that alone and the CMI would still boot up and go, it just wouldn't do midi or normal keyboard notes. But you could play sequences and stuff.
So you fixed that and now the CMI28 apparently has some other problem.
And you said it would "see" the channel cards, I'm assuming you mean the startup kmon screen where it does a check of what's installed in the system?
Maybe a silly question, but maybe its a good time to just audit what's in which slot in the machine. Are you quite sure everything's back the way it was supposed to be? If it boots, even part-way, I daresay _some_ of the cards must be right. Also, if you're running a WFM32 I'd put it all the way over to the far end of the WRAM sockets. If you're not running a WFM32, go get one from Peter Wielk.
BTW, in future, cleaning an old computer system that isn't broken, may cause it to start playing up, whether by static damage due to disassembly, or due to incorrect reassembly, or just brittleness. I tend only to clean a CMI if the dirt is so bad it's obstructing the airflow. Just flicking a feather-duster or synthetic cloth across a board can be enough to statically charge and damage it, especially in winter or desert conditions with dry air. Sounds obsessive, but it's true.
Humidity is your friend, unless you're a chip that will be wave-soldered!
> I will check the q777 is always the in : I suspect it is fine as the system is booting from SCSI disk.
Erm, there's a bit more to it than that in the wave super system. You'll want to make sure the SCSI cable is plugged into the wave super, rather than the Q777, but that Q777 is in the right-most slot, to the right of the floppy board. For a good primer on why the 777 still needs to be there, read KMI's site, under CMI hardware, where it talks about DMA on a series III.
Anyway, I think you'll get there.
Best of luck,
Joe
On Fri, Feb 18, 2011 at 12:04 AM, arroncx <aclague@synapse-consulting.com> wrote:
Hi Joe :
I will check the q777 is always the in : I suspect it is fine as the system is booting from SCSI disk.
OK : so the history in summery is that the machine was working, and then it was in storage for a couple of years while I was moving about and working out of the country.
When I got it back out of storage, the VGA video card blew a fuse, which I replaced, I was not detecting any voice cards installed, and one of the pins had snapped off the CMI28 ribbon connector. I should none that the first thing that I did when I got it out of storage was disassemble the whole system, partly to check it, but also because I needed to carry the mainframe up three flights of steep stairs, and wanted it to be lighter.
Since then, I have replaced the CMI 32 card, and I can now see all the voice cards.
Yes I typed DI : I don't have any reference for the diagnostic commands so I was just guessing in the dark that the process found in this thread on the fairlight boards would be something worth trying :
I really appreciate the suggestions : tonight I will try the following :
1) Reseat all socketed chips on CMI 28
2) Repet diagnostics with escape D only
3) Check the Card keys are present and the pins are lined up correctly (I have the top off the system and so can confirm visually)
4) Check the Q777 is seated OK.
Again : I really appreciate the suggestions
Best wishes : Arron> On Thu, Feb 17, 2011 at 10:21 PM, arroncx wrote:
--- In Fairlight-CMI@yahoogroups.com, Joe Sleator wrote:
>
> Hey Arron
>
> So, that's Chris Alfred's "new" diagnostics screen for the Waveform Super,
> CMI41. The address error that occurred is happening in the other 68K board,
> the Wave Super, CMI41.
>
> That screen has little to do with the CMI28, AFAIK
>
> Are you saying you typed "DI" at the prompt? Are you sure it's not just "D"?
>
> I'm guessing the D command was diagnostics, and the I command told it to do
> something else, which caused the address error, unless my own memory's
> faulty.
>
> I think that since you said it booted from disk, up to getting mptask
> running, that's a fair way in, and if CMI41 can do that, it's probably
> working ok.
>
> Another possible issue might be the DMA configuration of your machine. Is
> the Q777 plugged in all the way at the right, or is there another card
> there?
>
> And can I ask, from when this machine did work, what all has changed? Did it
> work recently? Have you exchanged any cards already?
>
> Without unpacking my Series III (still packed after recent move) I can't
> remember what the various diag menu options are, but I'll need to set it up
> soon enough for other requirements.
>
> But not long ago, we made some more WFM32s, and I seem to recall there being
> a memory test for CMI41. Maybe just "M".
>
> If you do find the right key for the memory test, and I'm pretty sure it's
> just one letter for each thing, you may as well verify the memory for the
> CMI41, even though I don't think that's your problem.
>
> Where did you get your diagnostic key code info? Maybe double-check?
>
> Another thing to look for whenever you have cards out, is there should be a
> key near the bottom of every backplane slot, that goes in the notches in the
> boards. When those are missing, all sorts of random things go wrong, as the
> pins don't line up!
>
> What happens if you just type escape and D?
>
> Good luck,
>
> Joe
>
>
>
>
>
> >
> >
> > Hi there : Ive posted the error screen i get from the diagnostics in the
> > group photos page if anyone can shed any light on what it means :)
> >
> > Maybe one day there will be an upgrade where you can get a PC on a
> > fairlight sized card, and a Crystal Core board in the same format, and then
> > just take all the old cards out of your fairlight, and wire up the old music
> > keyboard via midi , and the outputs round the back as straight thru sockets.
> > :)
> >
> > Photo at :
> >
> >
> > http://groups.yahoo.com/group/Fairlight-CMI/photos/album/1576007894/pic/66577237/view?picmode=&mode=tn&order=ordinal&start=1&count=20&dir=asc
> >
> > Cheers,
> >
> > Arron
> >
> >
> > --- In Fairlight-CMI@yahoogroups.com, James Thomson
> > wrote:
> > >
> > > Hi Joe,
> > >
> > > I would think that CMI-3 board repair could be a fairly lucrative
> > sideline to compliment someone's main job. There are many of us who would
> > pay a good price to have a board repaired as it is imperative to keeping the
> > system running. I would pay £ 100.00 - £ 150.00 to have a faulty CMI-28
> > board repaired. This board seems to fail a lot and I think I have about 3
> > faulty CMI-28 boards ( rev 5.1 I believe ). The CMI 2X's will last forever I
> > am sure but the CMI series 3 is a different matter. There is a market there.
> > Jo Britt recently produced a run of CMI-2X USB boxes so that you can use a
> > USB mouse and TFT screen instead of the lightpen and old green screen
> > monitor. They sold out immediately and there is demand for more. The speed
> > improvement using a USB mouse as compared with the lightpen is brilliant.
> > Using the whole CMI 2X is so much fluid now. The lightpen looks great and
> > this can still be used as well as the USB mouse. Perhaps the recently
> > established Fairlight Instruments could provide a CMI-3 board repair service
> > alongside their i-Pad and CMI-30A products ?
> > >
> > > Cheers,
> > >
> > > James
> > >
> > > Magic Audio Co
> > > Any Sound in Theory and in Practice
> > > Studio 26
> > > The Coach House
> > > 2 Upper York Street
> > > Bristol BS2 8QN
> > > U.K.
> > > studio direct line: +44 (0)117 370 5810
> > > mobile: +44 (0)7711 031 615
> > > email: jamesthomson@
> >
> > >
> > >
> > >
> > >
> > >
> > > On 17 Feb 2011, at 09:45, Joe Sleator wrote:
> > >
> > > >
> > > >
> > > > > I haven't found anyone who can repair the series 3 cards.
> > > > > It's really a shame that these amazing instruments are not supported.
> >
> > > >
> > > > OK, let's assume they were supported, and very well, better than even a
> > 2 year old Nokia phone, or a 69 MG Midget, or that old Russian space suit I
> > can't get parts for any longer. :-D
> > > >
> > > > Lets say you could fedex the card to a company, and they would either
> > repair or replace the card, to good as new, with 90 day warranty, or return
> > it to you and refund your shipping if it could not be repaired or replaced
> > for whatever reason. Zero risk to you. And that company would even pay you
> > for your broken boards if you decided not to get them shipped back.
> > > >
> > > > What would that be worth, lets say, for a CMI28? Go on and quote me a
> > price in any currency you'd care to. And how many paid Series III board
> > repairs do you reckon would come in per year? Not the sum-total of broken
> > CMI boards, mind you, but just the ones people wanted repaired enough to
> > pay.
> > > >
> > > > Joe
> > > >
> > > > On Thu, Feb 17, 2011 at 7:52 AM, Gmail wrote:
> > > >
> > > > Hi Arron,
> > > >
> > > >
> > > > This error message that you are getting is caused by a faulty CMI-28
> > card. I had the exact same error but was fortunate to have a spare card. It
> > would be great if there was someone with the series 3 knowledge to repair
> > these cards. There are people who can repair the CMI-2X cards but I haven't
> > found anyone who can repair the series 3 cards. It's really a shame that
> > these amazing instruments are not supported.
> > > >
> > > > Cheers,
> > > >
> > > > James
> > > >
> > > > On 16 Feb 2011, at 20:12, Joe Sleator wrote:
> > > >
> > > >>
> > > >> Hi, Arron
> > > >>
> > > >>
> > > >> First guess is it's falling over where it asks if the Midi Processor
> > (CMI28) is present, by poking and reading from its peripheral address, and
> > the midi processor either does not reply, or says something stoopid.
> > > >>
> > > >> Whilst the IIx would boot happily in this configuration, i.e with no
> > CMI28, all the III software releases I'm aware of are fairly tightly bound
> > to the CMI28, as it does most of the work triggering the sounds via the
> > channel cards. See if you can get your hands on a spare CMI28 would be my
> > next goal. You might just be able to re-seat the ROMS and RAMS and the 68K
> > in their sockets, being careful of the pins. Usually it's enough to pop a
> > tiny screwdriver under one end, lift the chip slightly, and then squish it
> > back down. There must be someone else in your neck of the woods with either
> > a IIx or a series III.
> > > >>
> > > >> IF YOU TAKE CMI28 OUT BE SURE YOU ARE EARTHED ESPECIALLY IF IT's
> > WINTERTIME AND THE HEAT IS ON IN YOUR PLACE BEFORE TOUCHING THE CHIPS
> > DIRECTLY. I can't stress this enough, many times boards are degraded or die
> > outright from miniscule ~5KV or less, static discharges and it9;s totally
> > avoidable.
> > > >>
> > > >> If the ROMS have suffered bit-rot (you can tell by reading them out
> > and comparing to known good roms) they can fairly easily be replaced, or
> > erased and re-programmed.
> > > >>
> > > >> The other way to tell is to slot roms from a working CMI28 into the
> > suss board and re-try.
> > > >>
> > > >> The roms, rams, and 68K chip are all socketed. A resourceful soul can
> > find replacements for all these chips by poking around. Virtually all the
> > other chips on the CMI28 are common TTL or early motorola LSI, so also
> > should be obtainable, however finding what's gone bad is the challenge.
> > > >>
> > > >> Since the Series III diagnostics do virtually the same thing as
> > mptask, I don't think they'd shed much more light, but probably worth a try.
> > > >>
> > > >> Good luck.
> > > >> Joe
> > > >>
> > > >> On Thu, Feb 17, 2011 at 5:14 AM, arroncx wrote:
> > > >>
> > > >> Hi all,
> > > >>
> > > >> Thanks for all you assistance so far. Just thought Id report back :)
> > > >>
> > > >> So : I got a replacement CMI32 card from the ever helpful Peter Wielk
> > and great now when I boot, i can see all my channel cards, memory etc.
> > > >>
> > > >> The system runs right the way though its boot-up sequence, but just as
> > its about to start, the system exits to the command prompt with
> > > >>
> > > >> mptask handshake timeout : aborting process.
> > > >>
> > > >> does anyone know where I could start looking next to get a bit further
> > ?
> > > >>
> > > >> I looked on the KMI website, and it says that mptask controls all the
> > realtime events in the CMI. (if Im understanding it right)
> > > >>
> > > >> Also : when I boot up : the status display says that no sync card is
> > present : what does this refer to, and is this part of my my problem ?
> > > >>
> > > >> Any hints or clues appreciated as normal :)
> > > >>
> > > >> Best wishes : Arron
> > > >>
> > > >>
> > > >>
> > > >>
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > >
> >
> >
> >
>