D8B Booting Partially

[markd]

Thanks Phil.c
Do you mean disconnect the ribbon cable, spray the pins and the connector holes and re-seat?

[csp]

Markd,

If you happen to have the bottom off the desk (you will need to do this when ribbon cleaning anyway), when you boot up have a look at the three LED lights on the DSP board and write down what they do as the desk loads, also do the same with the LED lights on the Brain Board and then post the results --- these LEDS and the colours they give are quite important in determining what the desk is doing. Again I suggest reading my very long topic as these LEDS are raised quite a bit. You could find that you get the correct indications until the desk freezes and then the LEDS (especially on the DSP board) give an incorrect reading.

Re what Phil has stated about lifting the spikes, I have never been able to do this without damaging the plug, so I generally as part of the cleaning process and while the plug and cable are quite wet with the DEOXIT (or whatever you use --- do not use WD40 or similar as they leave a residue, only use fast drying electrical contact fluid or isopropy) l twist the plug in all directions (without breaking it !!!) so as to move the wires on the pins. BUT if you can do it Phil's way it would definitey be better as it would most certainly move the cables on the pins. Also if you can (a tedious job but I have done it a number of times) before cleaning the cable plugs do a continuity check on the cables (ie pin to pin), where you should get an almost 0 ohm resistance on each wire, if not or you get a high resistance then it indicated that the particular cable in the ribbon is not making proper contact or no contact and therefore definitely requires a clean. Also mark each end of the cable with a code of some sort and where the red stripe goes to be sure that you put the cable back exactly where it should go --- sounds silly and obvious, BUT I goofed one time and had to buy a new (ie 2nd hand) Brain Board because I put a cable back into the wrong socket and with the red stripe on the wrong side of the socket.

Also be sure that every plug is very firmly pushed in and locked in the socket as I have found that they do move a bit in the socket and therefore not all of the wires make contact.

Hope all of this helps a bit ---- now time for bed in my part of the world !!!!!!!!!!!

David

[Phil.c]

Thanks Phil.c
Do you mean disconnect the ribbon cable, spray the pins and the connector holes and re-seat?

Run your thumb nail in the connector join both sides so you can see a small gap, this lifts the ribbon off the spikes as sometimes the cable oxidises onto them causing a bad connection, put a few drops of IPA in and squeeze back the connector tight with a pliers.

[Y-my-R]

Great to hear that you were able to get into the BIOS with the new keyboard - too bad that a new battery and restoring the BIOS settings didn't resolve the issue.

With this out of the way, my guess is that the desks locks up the moment it tries to engage word clock sync from the clock card (because that gets activated pretty much at the point once the desk is fully booted... so, having the mouse respond for a moment and then having everything freeze, sounds like it happens when the desk starts looking for clock (internal or external).

One possible cause, would be what the word clock setting is set to in the startup session. It shouldn't lock up because of this, but if the startup song has this set to "external" clock that isn't present (and something else is wrong that causes the freeze), then I could see the startup song as a potential cause of the issue.

So, with the USB/IDE adapter, I'd just open the rack unit, but keep the spinning harddrive in place, then connect the USB/IDE adapter to it via your every day computer, then find the startup song and delete it.
Then I'd connect the drive back to the D8B computer, and try booting another time.

If you're lucky, it might not freeze anymore after that... but I suspect that it's not just the settings in the startup song, but likely word clock not reaching everywhere it needs to (e.g. either the solder spots on the Apogee clock card I mentioned, or bad contacts, elsewhere.

You do make a good point about the problem likely happening in the rack unit, rather than the console, since the same thing happens with another console connected. However, the clock ALWAYS gets generated in the console (via the stock or Apogee Clock card in the "SYNC" card slot), and then sent to the various components in the console, but also back to the rack unit.
So, if the lockup also happens with the other console (did that also have an Apogee Clock card installed in it), then I'd focus on first cleaning any connections on the way from the clock card to the rack unit.

Did you try what happens if you try booting the rack unit without any console connected at all? At least in that scenario, I'd expect it to boot to the same place it went for you, so far, but then instead of locking up, it should give an error, such as "Last Status: DSP card not ready" (although I'm not sure if OS3 typically does that... OS5.1 would definitely show this message)... but I think the mouse should then still move, instead of freezing.

If this is what happens with no console connected (Error message, but mouse still can be moved, although you can't "do" anything with it), then it would once again point to the word clock not being distributed everywhere it needs to go, IMO.

If it locks up all the same, then I'd have a close look at the power supply components inside the rack unit, to see if anything seems odd, and also reseat connectors and clean them.

...at least that's the stuff I'd do first, before doing the full ribbon cable reset in the console, etc. It's probably a good idea to do that, anyway... but I usually like to take one step at a time and then check for changes in behavior, to get a better clue where the problem actually came from (...and know better if something similar happens, again).

As for what Phil mentioned about the ribbon cables... (I hope I'm not beating a dead horse) - the way the connectors are attached to ribbon cables is, that within the connectors themselves, there's these metal spikes that pierce through the ribbon cable at the right distances, to make contact with the wires inside.
...and sometimes there's oxidation, where those spikes touch the wires inside the ribbon cables. So, short of replacing all the ribbon cables with new ones, it can help to clean the spikes/wires inside the connector.

I mean, from my understanding, Mackie had issues with these ribbon cables from early on with the D8B. So, not the best contact in the first place, I guess, and then maybe not the highest quality materials in the connectors (i.e. probably went for the cheapest vendor for these cables). I think that's why the ribbon cables in the D8B need more maintenance than ribbon cables in various other equipment of similar age, usually do.

So, to summarize, I'd first delete the setup song to make sure that doesn't cause the lockup. Then I'd take a good look at the components on the Apogee Clock card (i.e. SYNC slot - if it says Apogee but isn't in the SYNC slot, then that's likely not a clock card. DIO8 I/O cards also say "Apogee" on them, for example), to make sure no components are trying to fall off of the card (i.e. re-float or re-do the solder points, if so).
Then I'd look in the rack unit's power supply, to see if anything isn't right, and clean a few of the contacts while at it... (and also clean the BFC connector contacts, while at it).

Once I checked all of this "quick and simple" stuff, that's where I'd look at doing the larger tasks, like reinstalling or cloning the Mackie OS back onto the spinning drive (or a new CF card), or doing the full ribbon cable reset.

I think if you go about this with the expectation that you might well have to do ALL of these tasks before having a fully working desk, then you shouldn't get too disappointed... of course it can happen that such a desk doesn't want to be brought back, somehow (i.e. see csp/David's thread)... but most of the time, when patiently working through all of these "typical" D8B maintenance/ressurection steps, it should eventually "come back."

Just keep your patience up, and your expectations low to avoid frustration, haha

Again, best of luck in the meantime!

[csp]

Y-my-R,

I am glad that you mentioned the possibility of the 'clock card' actually being a DIO8 rather than a real Apogee clock card and it not being in the Sync slot as I hadn't thought about that possibility. Over the years (especially more recent times) I have seen items on ebay listed as an Apogee clock card, when in fact they are actually the DIO8 card.

Markd --- The DIO8 card does NOT have the BNC connectors but rather a coaxial, ADAT optical and DB25 plugs, whereas the real Apogee clock card only has the two BNC connectors.

David

[Y-my-R]

...just to avoid confusion:

The Apogee Clock card has 2 BNC connectors on it, as David said (and no other connectors... just a big Apogee logo).

The DIO8 card has a single BNC connector, two ADAT (optical) connectors, and a single DB25 connector for TDIF (no RCA/coax).

It's definitely an important thing to double-check, though. I also recall multiple occasions on this forum, where when asked what clock card, the response was "Apogee" because they read it off of a DIO8 card in the back (...and often with the Apogee clock card missing... since they're the best seller of any used D8B parts there is).

If you do have an Apogee Clock card (with 2 BNC connectors but nothing else) in the SYNC card slot, I'd still take it out (loosen the 2 thumb screws and pull) to look at the solder spots/components... and maybe try to clean the card-slot the clock card is installed to, while at it.

Anyway... David has a good point, though. The Apogee Clock cards and the DIO8 cards get mistaken for one another often enough.

[markd]

Thanks to all who offered assistance and information relevant to my situation.
Hope you all had a happy new year.
Update:
I checked the Wordclock card and I have genuine Apogee cards in both consoles.
2 x BNC connectors only on each card. Thanks for the clarification provided on this.

I have now purchased an IDE to USB interface and made up a DC supply for the Mackie HDD using an old PSU and made up a DC converter cable.
I used this rig to image the Mackie 3 drive using Macrium Reflect Free (under trial).
I then deleted a file called 'Mix#1.d8b' on the Mackie drive as this was showing as the filename at the top of the screen after D8B boot. Is this the session file that was suggested to be deleted? It did not change (get rid of) the problem.

I also purchased a CF to IDE card and a CF card in order to be able to replace the internal Mackie HDD for further testing.
I also have a USB CF Card Reader/Writer. I am thinking of trying the CF replacement drive sooner rather than later as I don't know how solid the Mackie HDD is. I was able to repair a cluster using chkdsk while the laptop was connected to the drive, but this did not clear the problem either.

I pulled the Apogee card and cleaned the contacts - no change.
My next step is to get the Mackie OS written to the CF card and get that working for further testing / fault finding. My other D8B (parts donor) CPU Rack unit has a burnt interface board on the HDD. Are there any tricks when formatting the CF card and installing the Mackie OS for boot? I do not have the Mackie floppy disks, I intend getting the OS from the forum. I think I should get V3 working first on the CF card then move on to V5.1 once my other problems are sorted.
Thanks for your continued support.

[markd]

Btw, I disconnected the BFC and rebooted (as suggested by Y-my-R) and boot was same - stopped on the same screen with error message "Error: DSP Control Card not ready. State: Reset". Mouse not moving the cursor and keyboard ignored. Caps Lock turns on and off (normally keyboard does not do this here when BFC connected).

[Y-my-R]

Hi markd,

That Mix#1.d8b file wasn’t the right file to delete. That’s a regular user-generated session.

The D8B has a feature, where you can have it start up with your own “template session” every time - but this only works if that project is named “Startup.d8b” and if that project is located in the “Session5” folder (or probably Session3 or similar on OS3), and if “Load Last Session at Power-up” is NOT enabled under settings.

Where this Startup.d8b file can make problems, is if either the file has gotten corrupted, or if that session contains some settings that just don’t want to get along with the environment it’s in (e.g. set to external clock with no external clock present... that shouldn't cause a lockup by itself... but possibly combined with some other issue).

It’s entirely possible that no such “Startup” session exists. But if it’s there, try deleting it and then check if the D8B gets past the point where it currently locks up. The D8B should then create a new Startup session with the most basic settings, when successfully starting the next time.

Good that you have an image of your OS3 drive, now. As long as the image was made to be of the entire disk (and not just the active partition), then you should now have a clone of the all-important boot sector on your D8B drive. I only know of two ways how to get that boot sector onto a new/different drive… and that’s either by formatting the new drive during a re-installation from the floppy drive (and the “format:on” flag set on the first floppy disk), OR by restoring an image of such a drive to a new drive (e.g. the Macrium Reflect image you now have).

You could of course just restore arjepsen’s OS 5.1 image that he shared on this thread, since this also includes that important boot sector:

https://d8bforum.com/phpBB3/viewtopic.php?f=2&t=1634&p=14186&hilit=macrium#p14186

However, as long as the boot sector has been restored to a new drive (e.g. from your own clone), the rest of the files can simply be copied onto the drive, without any proper installation procedure being necessary.

So, if you were to restore your OS3 image to your CF card, for example, and it still freezes… if you had access to a functioning OS3 installation, you could simply delete all the files off of your own restored OS3 image (but don’t format, since the boot sector needs to be retained), and then just copy the whole set of files from that other OS3 installation over to your new CF card.

You’ll then likely get some errors about plug-in licenses, and potentially about the OS and hardware serials not matching… but then you could copy the files that contain those licenses back over from your cloned drive to your new drive (or try to extract the serial codes from those files, and re-enter them to unlock the plug-ins the original owner of the console purchased).

Hopefully, with the original boot sector in place, this will give you the flexibility to experiment with OS3 and troubleshoot… and then maybe install OS 5.1 on another CF card when you got that working, then you could easily revert to OS3 if, for example, your setup turns out to not be happy with the amount of RAM in your system, that is listed as the bare minimum for OS 5.1 (at least if you want to use automation via the D8B, you’d probably want to install more RAM - those sticks are still available, luckily, but you have to be really careful to get the right type… I have some thoughts on that, too… but let’s cross that bridge when we get there).

As for booting the D8B rack unit without a console connected:

I thought you said that you do NOT get the “Error: DSP Control Card not ready” error when booting WITH a console connected?
You only get that error if NO console is connected, correct? But the system locks up WITHOUT that error, if a console IS connected?

If so, that’s the EXACT difference in behavior I was looking for, that IMO would still point to a problem with word clock (or the word clock signal reaching everywhere it needs to…, so a potential clock card, ribbon cable or other cable/connectivity issue).

From what I understand, the rack unit boots up and once ready, attempts to communicate with the console (aka the “DSP card” it complains about in that message). So, if the console is not connected, it can’t find it and gives that error.
If a console IS connected and that error doesn’t pop up, it would suggest that the rack unit successfully “finds” the console and starts communicating with it… but that the failure happens at the next step after that.
…and since the next step can be assumed to be that the D8B system now tries to establish clock (i.e. rack: ready, console: ready… start clock), that’s what I think where your D8B locks up.

Good that you have two Apogee Clock cards. So, the first thing I’d try in that case, is with the “other” clock card from your parts system, in your main system, to check if that makes any difference. If not, the clock card(s) are less likely to be the culprit, but I’d still take a good look if the solder connections of the components I highlighted in the thread below look ok, on your clock cards:

https://d8bforum.com/phpBB3/viewtopic.php?f=2&t=2583&p=19786&hilit=solder+spot#p19786

If none of that helps (i.e. fresh OS on CF, different Apogee Clock card in main system and after checking that the components on the card aren’t separating), then I think it would be time to do the dreaded ribbon-cable cleaning.

Alright, curious to hear how this will progress for you! Best of luck in the meantime!

[markd]

Thanks for your detailed response (again) Y-my-R.
I got the CF / IDE card installed in the Mackie rack unit and used Marium Reflect Trial to create an image backup of the original V3 HDD.

I used Restore in Macrium to restore the image to a 2GB CF card. After several tries (deleting the non-working partition each time on the CF card), I got the v3 image written and booted the Mixer with the CF card installed. I got it to the same point as the regular HDD boot - i.e. it booted and ran to the same point as it did on the HDD. Thinking this was good, I switched off and restored another CF card with the same image and then downloaded arjepsen’s OS 5.1 image from the link provided by Y-my-R in his last post.

I then tried re-booting the working v3 CF and got an error message "BOOT FAIL". Every time I try to reboot that card I now get asked to install a bootable disk. The last series of tests I completed - three separate CF cards (2 x OS3 and 1 x OS5.1) give the same error and stops the boot process. I am not sure if all the cards have now gone faulty, or what is happening, but I will try again tomorrow. The first CF card worked once (at least to the same place as the HDD boot), but now everything is completely flakey and I am not sure what is happening. Will keep going.....

PS: I must learn how to snip quotes when responding to make it clear what I am answering.