Corcomp 9900 disk controller problems

[+Ksarul]

If the order is different, the JEDEC code will be different as well, only because of the positions where that code expresses itself. A lot of stuff back then used 12L6 and 14L4 PALS to make selection circuits work more efficiently. . .as the PAL would replace several 74XXX series chips.

[justacruzr2]

That's what the datasheet said also. It gave a ratio of better than 4:1. Or at least four chips could be eliminated with 1 PAL. The flipside of that though is if the PAL becomes bad you have a tough problem. Especially when factory support is no longer available. I would also add that I think CorComp never expected the TI crowd to still be going after all this time. Hats off to TI.

 

Still haven't found why the Rev A card isn't working. But I did get a strange reaction when I replaced the new WD1773 with the original one. The LED came on when the PEB was turned on but the console off. It's not supposed to do that. No cards light should come on until the console is powered up....Hmmmm...... At least that would seem to confirm that the original WD1773 had a problem. Because there seems to be some other problem with the early card which is preventing it from reading a disk, I'm concentrating on the Rev A card since, when it was working, it was able to access a disk.. It's puzzling when the state of something changes for no apparent reason.

 

Anyway, I'm going to take a break from it today to get some other things done. Sometimes you have to walk away from a frustrating problem and come back to it later fresh.

[justacruzr2]

OK. Found the problem with the Rev A card. A bad solder joint making intermittent contact. It's back online now, with heatsink, and works perfect. Ran it for about an hour with no problems. So this one has good PALs too. It was good to be able to catalog disks again and see that everything is OK. All smiles here!

Now this leaves the matter of the early card which does work also except it doesn't read the disk. The following are the things that I have identified which may be affecting it:

1) 2 unadjusted trim pots.
2) 3 mods (or corrections) on the solder side.

Not sure how to proceed with #1 unless someone knows where on the board to take the readings and what value I should be getting for each one.

As for the mods, does anyone have a clean early version CC9900 card that is working that you could take a nice picture of the solder side of the board and post it in a reply so I can compare it with what I have? By "clean" I mean a board that hasn't been modified by the end user. Any original CorComp mods or corrections it may have is OK. That's what I'm trying to determine. Mine has 3 mods which look like CorComp mods but I can't be sure without comparing it, or hearing from someone who knows what the official CorComp mods were. Frequently their cards did not work exactly right when they went out the door. The Triple Tech card I bought from them had to be sent back for repair because it wasn't working exactly right when I got it and they made some mods on the back to correct it, I guess it was some design flaw.

[RickyDean]

Here are photos of my actual Corcomp units the 9900 PEB version and the 9900 side card disk controller board. I was sent two Miller Graphic eprom replacements, for the 9900 PEB card,but they didn't work( found after I received my TL866A, that the eproms had useless data, burned schmizti's binaries and now getting the same thing as my original MG eproms were getting. I suspect that it is still an issue with one of the two Pal's found onboard, but I am going to check all the solder points and replace the caps and so forth, to rule those out, on the Geneve, it works fine, but just getting a cyan screen on the TI. The side car expansion, still gives me a cyan screen after pulling the disk controller board out, so I believe the Pal on the RS232 card portion is bad. This happened after copying about 70 disks to 3.5 format to image with TI-PC on a laptop that would do single density. The side car was not anchored well to the side of the TI and changing out drives and disks, kept shifting it, and may have blown the pal. Have checked and/or replaced all the logic chips on it as well, but still just a cyan screen. I still see a need for replacement Pal's for these replicated, and I still believe I need them.

Edited August 1, 2016 by RickyDean

[+Schmitzi]

Hi,

 

I slightly remember there was a difference, unmodded Eprom is 8K, MG-Mod is 16K, IIRC (?!?)

 

AND, I burned it some times, did not work, then I recognized that the used Eproms were flaky....

 

AND, if you have a "Rev.A"-board, EPROM#1 is plugged in the left socket (=U1),

BUT if you have "Rev.C"-board (?), EPROM#1 is plugged to the right socket, #2 to the left (=vice versa)

 

AND, I saw many pics where the CC-9900 only has ONE Eprom plugged in (but not the MG-Upgrade, I think)

 

Maybe this is one of the isssues....

 

nice pics, thanks

[atrax27407]

There are two 8K EPROMs in the set and they must be in the proper sockets.

[justacruzr2]

Here are photos of my actual Corcomp units the 9900 PEB version and the 9900 side card disk controller board. I was sent two Miller Graphic eprom replacements, for the 9900 PEB card,but they didn't work( found after I received my TL866A, that the eproms had useless data, burned schmizti's binaries and now getting the same thing as my original MG eproms were getting. I suspect that it is still an issue with one of the two Pal's found onboard, but I am going to check all the solder points and replace the caps and so forth, to rule those out, on the Geneve, it works fine, but just getting a cyan screen on the TI. The side car expansion, still gives me a cyan screen after pulling the disk controller board out, so I believe the Pal on the RS232 card portion is bad. This happened after copying about 70 disks to 3.5 format to image with TI-PC on a laptop that would do single density. The side car was not anchored well to the side of the TI and changing out drives and disks, kept shifting it, and may have blown the pal. Have checked and/or replaced all the logic chips on it as well, but still just a cyan screen. I still see a need for replacement Pal's for these replicated, and I still believe I need them.

 

Hi Ricky,

 

Thanks for posting the photos but could you please take a photo of the backside (solder side) of your early version CC9900? This is the side I'm not sure about (for reasons mentioned above). One of the other things I found on my Rev A card was a broken trace. When I repaired that, the card worked. If you have an LED flashlight, shine it through the board from one side while looking at the other side. That was how I found it. It's not a bad idea to make this check due to the age of these boards and you might find something similar on yours. That's all it takes for the board not to work. All circuits must be complete. I also have a complete nearly unused set of the MG EPROMS that I bought from Bud Mills Services. By nearly unused I mean I plugged them in when I first bought them to see what they did for the card but didn't like the MG blurb at the top of the TI master screen and didn't like that I couldn't access the disk manager with one keypress. Also, I used them a couple of times while trying to figure out what was wrong with my cards to eliminate the possibility that the original EPROMS were bad (testing purposes). I would be willing to sell these and they come within the original packaging (blue plastic cube with foam pad) and the original instructions detailing the extra commands contained in those EPROMS. Let me know if your interested and we'll proceed from there. I'm still working on repairing my floppy drives so I'm not completely back up to speed yet. It's very tedious work to remove and replace the 45 pin 4 sided surface mount master floppy read/write IC.

 

PS I also see that your EPROMS are missing the paper tags which identify which bank they are and the revision date. With those tags removed there is always the possibility that they were exposed to ultraviolet light and have been either erased or messed up.

Edited August 2, 2016 by justacruzr2

[RickyDean]

 

Hi Ricky,

 

Thanks for posting the photos but could you please take a photo of the backside (solder side) of your early version CC9900? This is the side I'm not sure about (for reasons mentioned above). One of the other things I found on my Rev A card was a broken trace. When I repaired that, the card worked. If you have an LED flashlight, shine it through the board from one side while looking at the other side. That was how I found it. It's not a bad idea to make this check due to the age of these boards and you might find something similar on yours. That's all it takes for the board not to work. All circuits must be complete. I also have a complete nearly unused set of the MG EPROMS that I bought from Bud Mills Services. By nearly unused I mean I plugged them in when I first bought them to see what they did for the card but didn't like the MG blurb at the top of the TI master screen and didn't like that I couldn't access the disk manager with one keypress. Also, I used them a couple of times while trying to figure out what was wrong with my cards to eliminate the possibility that the original EPROMS were bad (testing purposes). I would be willing to sell these and they come within the original packaging (blue plastic cube with foam pad) and the original instructions detailing the extra commands contained in those EPROMS. Let me know if your interested and we'll proceed from there. I'm still working on repairing my floppy drives so I'm not completely back up to speed yet. It's very tedious work to remove and replace the 45 pin 4 sided surface mount master floppy read/write IC.

 

PS I also see that your EPROMS are missing the paper tags which identify which bank they are and the revision date. With those tags removed there is always the possibility that they were exposed to ultraviolet light and have been either erased or messed up.

Yes I'll take a picture tonight and post it tomorrow. As for the stickers being off the eproms, this was done Sunday as I removed them to erase the bad data off the two eproms that I purchased that were supposed to have the MG binaries burned on them. I had just burned them with Schmitzi's binaries and haven't covered them yet. They, with the freshly burned images still do the same thing as the original MG eproms, getting a cyan screen on the TI, but the Geneve bypasses the eproms and works with the card, so that is why I am still leaning toward bad Pal(s).

Edited August 2, 2016 by RickyDean

[RickyDean]

Here is the 9900 with the original MG eproms reinstalled and also a backside view. I looked through it last night with a LED flashlight, but could not see an issue there.

Edited August 3, 2016 by RickyDean

[atrax27407]

I can burn some new EPROMs should you need them.

[RickyDean]

I can burn some new EPROMs should you need them.

Thanks for the offer, but if you look at post 79, you will see the set, I burned myself, the problem is not in the Eproms(I do not believe). I think the problem is in one of the two PAL's. That is why I am interested in finding out if someone has or can copy two from a known good one. I understand the PAl fuses could be blown to prevent copying, and if there is not a PAL solution, I will have to try and reverse engineer the card to see if I can come up with an equivalent that will plug into the sockets. I have two HFDC's one for each Geneve and I want a DSDD controller for one of my TI's. Also I want a copy of all the PAL's in the 9900 CES sidecar, especially the RS232 board portion. Thanks again, though.

[atrax27407]

You are, of course assuming that the code that you burned your EPROMs from was good. Corcomp was very efficient in protecting their PAL content by locking them. You can check with Jim Fetzner to see if he might be able to unlock them and read the code. If he doesn't have them, the only recourse would be to trace the circuits and re-create the PAL equatiuons. That would be a good project - a repository for the PAL code from TI devices.

[RickyDean]

You are, of course assuming that the code that you burned your EPROMs from was good. Corcomp was very efficient in protecting their PAL content by locking them. You can check with Jim Fetzner to see if he might be able to unlock them and read the code. If he doesn't have them, the only recourse would be to trace the circuits and re-create the PAL equatiuons. That would be a good project - a repository for the PAL code from TI devices.

The code I used for the eproms was from schmitzi, He always has good code right? Also there has been some other discussion in another thread about the PAL issue. http://atariage.com/forums/topic/253808-corcomp-9900-fdc-pals/?hl=%2Bpal&do=findComment&comment=3532884

Edited August 3, 2016 by RickyDean

[atrax27407]

I have some known working EPROM code. The EPROMs in both revisions are the same, I believe. However, if the circuitry is different, the PAL equations will also be different.

[RickyDean]

I have some known working EPROM code. The EPROMs in both revisions are the same, I believe. However, if the circuitry is different, the PAL equations will also be different.

Yes, I agree. I believe that burning another eprom set will not fix the problem. I have tested all the logic chips, changed out the 9901 and the 2793 FDC and examined my traces with a flashlight.

That only leaves bad solder joints, caps, resistors,a transistor( to be honest, I haven't checked it yet, will do tonight),crystal, and variable resistors(did I already say resistors?) .

Edited August 3, 2016 by RickyDean

[justacruzr2]

Here is the 9900 with the original MG eproms reinstalled and also a backside view. I looked through it last night with a LED flashlight, but could not see an issue there.

 

Thanks for the backside photo. Very helpful and also a bit puzzling. Only one modification (or correction) is the same on both our cards. That being the wire jumpering pins 2 and 5 on the right side of the 74LS74 together. 2 other modifications on mine are not present on yours. Also mine has the first 4 pins of the MCM2114P jumpered together. I think this is OK though as my Rev A card does the same thing but it has a circuit trace doing the job instead of a wire. Probably one of the CC revisions made on that later card. You might want to consider doing that also. I think it's necessary. Before I undo those other 2 mods and see what happens just a question....did your card ever work to start with? Trying to determine if those were user mods that didn't pan out. My card does power up and present the CC menu screen but it doesn't recognize a floppy disk in the drive. Weird. Gotta be a reason for that.

[justacruzr2]

Yes, I agree. I believe that burning another eprom set will not fix the problem. I have tested all the logic chips, changed out the 9901 and the 2793 FDC and examined my traces with a flashlight.

That only leaves bad solder joints, caps, resistors,a transistor( to be honest, I haven't checked it yet, will do tonight),crystal, and variable resistors(did I already say resistors?) .

 

Transistors are susceptible but I think the only transistor on that board runs the LED. Also look for any brown discoloration on the caps. I had several on mine that had brown spots on them and that indicates overheating. Also had 2 bad resistor packs (SIPS). Very doubtful that a crystal would go bad.

[RickyDean]

 

Thanks for the backside photo. Very helpful and also a bit puzzling. Only one modification (or correction) is the same on both our cards. That being the wire jumpering pins 2 and 5 on the right side of the 74LS74 together. 2 other modifications on mine are not present on yours. Also mine has the first 4 pins of the MCM2114P jumpered together. I think this is OK though as my Rev A card does the same thing but it has a circuit trace doing the job instead of a wire. Probably one of the CC revisions made on that later card. You might want to consider doing that also. I think it's necessary. Before I undo those other 2 mods and see what happens just a question....did your card ever work to start with? Trying to determine if those were user mods that didn't pan out. My card does power up and present the CC menu screen but it doesn't recognize a floppy disk in the drive. Weird. Gotta be a reason for that.

Oh yes, it worked fine in both the TI and Geneve until one day it just stopped working on the TI, I can't remember if it locked the system up or just gave the cyan screen that it does now, that was 18 years ago. It still works fine on the Geneve, I just booted the night before last on my Gotek floppy emulator and a 16GB thumb drive and have booted in the last few months from it with 3.5 floppies. As for your floppy drive not being seen, I experienced this with my HFDC and some of my old drives, maybe not the same problem, but I think it was the drive's' fault. Also a similar experience with the 9900 CES till it gave me the same cyan screen I have described earlier.

Edited August 3, 2016 by RickyDean

[+Schmitzi]

just because I found this:

 

 

(swapped Eproms on some boards)

 

[justacruzr2]

Oh yes, it worked fine in both the TI and Geneve until one day it just stopped working on the TI, I can't remember if it locked the system up or just gave the cyan screen that it does now, that was 18 years ago. It still works fine on the Geneve, I just booted the night before last on my Gotek floppy emulator and a 16GB thumb drive and have booted in the last few months from it with 3.5 floppies. As for your floppy drive not being seen, I experienced this with my HFDC and some of my old drives, maybe not the same problem, but I think it was the drive's' fault. Also a similar experience with the 9900 CES till it gave me the same cyan screen I have described earlier.

 

OK then. I'm going to undo those 2 mods and see what happens. My only other thought on it is maybe the Rev A card is a better design than the early model card. Currently my PEB has 1 bad drive attached and 1 good one. The bad one is one of the originals I bought that took a hit when I used that WD2793A by mistake. The other one is one I bought off eBay so I would have one good one for testing while repairing the CC cards. I also have a 3rd one that was actually the 2nd drive in the PEB and it took a hit at the same time as the other one. That is the one I'm currently trying to repair. If successful I will then repair the one in the PEB too. Anyway, the Rev A card does recognize a disk in drive 2 with the one bad drive attached while the early CC card does not.

 

Funny about your card working in the Geneve but not the TI. Makes me think that maybe your PALs aren't bad. Don't see how the Geneve could bypass the PALs which are essential to the cards proper functioning. But I could be wrong.

[RickyDean]

 

OK then. I'm going to undo those 2 mods and see what happens. My only other thought on it is maybe the Rev A card is a better design than the early model card. Currently my PEB has 1 bad drive attached and 1 good one. The bad one is one of the originals I bought that took a hit when I used that WD2793A by mistake. The other one is one I bought off eBay so I would have one good one for testing while repairing the CC cards. I also have a 3rd one that was actually the 2nd drive in the PEB and it took a hit at the same time as the other one. That is the one I'm currently trying to repair. If successful I will then repair the one in the PEB too. Anyway, the Rev A card does recognize a disk in drive 2 with the one bad drive attached while the early CC card does not.

 

Funny about your card working in the Geneve but not the TI. Makes me think that maybe your PALs aren't bad. Don't see how the Geneve could bypass the PALs which are essential to the cards proper functioning. But I could be wrong.

Yeh, that may be the case, My thinking is that the PAL's(at least the one directly under the eproms) are used in the the process of the TI 'seeing them', but I don't know enough about it for sure. And it could be a bad solder point, blah,blah,blah, etc.

 

I remember doing some work on it back then, but then abruptly packed away everything, as my dad and I were at odds at the machine shop he owned, about a bad family member issue, and I moved back to S. Georgia, where I am today.

 

So after all this time, I'm trying to get my stuff back into useable shape, but I have a lot of irons in the fire, projects(TI, PCs, cars), work, and my wife had been a unwilling participant in a car accident 3 months ago, so trying to plow through it all.

 

I will work on something for awhile then put it down and go to something else another day, because what I had been trying to do wasn't working at that point.

 

Case in point, my IDE cards, I have built them, but having issues with them( used the hot plate soldering technique, so trying to debug).

[RickyDean]

I decided last night to strip my 9900 down to the bare board and rebuild it with all new components to include machined sockets. I will be scanning both sides and using one of the electronic CAD packages to produce gerbers down the line. The only used stuff will be the hard to replace chips. This should tell us for sure if it is a PAL issue or not.

[justacruzr2]

Yeh, that may be the case, My thinking is that the PAL's(at least the one directly under the eproms) are used in the the process of the TI 'seeing them', but I don't know enough about it for sure. And it could be a bad solder point, blah,blah,blah, etc.

 

I remember doing some work on it back then, but then abruptly packed away everything, as my dad and I were at odds at the machine shop he owned, about a bad family member issue, and I moved back to S. Georgia, where I am today.

 

So after all this time, I'm trying to get my stuff back into useable shape, but I have a lot of irons in the fire, projects(TI, PCs, cars), work, and my wife had been a unwilling participant in a car accident 3 months ago, so trying to plow through it all.

 

I will work on something for awhile then put it down and go to something else another day, because what I had been trying to do wasn't working at that point.

 

Case in point, my IDE cards, I have built them, but having issues with them( used the hot plate soldering technique, so trying to debug).

 

Do the same thing myself. When something isn't going exactly right I'll take a break from it. Found that coming back to it refreshed usually makes the difference.

 

I'm curious....you mention cars as one of your projects. Are you talking about restoring older cars like from the 50's, 60's and 70's? Have 4 that I'm doing myself. Currently working on my 70 Plymouth Belvedere (the cheap RoadRunner!). Have been enjoying this hobby since I was 15. Hard to find the time sometimes though. Patience, patience, patience.

[RickyDean]

 

Do the same thing myself. When something isn't going exactly right I'll take a break from it. Found that coming back to it refreshed usually makes the difference.

 

I'm curious....you mention cars as one of your projects. Are you talking about restoring older cars like from the 50's, 60's and 70's? Have 4 that I'm doing myself. Currently working on my 70 Plymouth Belvedere (the cheap RoadRunner!). Have been enjoying this hobby since I was 15. Hard to find the time sometimes though. Patience, patience, patience.

Well I can work on and fix these, yes I am that old, but I am currently working on 4 GMC and Chevy diesel pickups, and some jeeps and a couple of Suburbans, all the GM vehicles are in the 2000 year range.

[justacruzr2]

I decided last night to strip my 9900 down to the bare board and rebuild it with all new components to include machined sockets. I will be scanning both sides and using one of the electronic CAD packages to produce gerbers down the line. The only used stuff will be the hard to replace chips. This should tell us for sure if it is a PAL issue or not.

 

That's basically what I did too. Everything on my board is new except for the PAL's, the crystal, the ribbon cable socket and the 8 position DIP selector. I was able to find everything either from DigiKey or on eBay. One suggestion....before you remove the trim pots take a reading on them so you can adjust the new ones (if you plan to replace them too) to the same value. I neglected to do this with mine and now don't know what they should be set at. This could possibility be part of the problem I'm having with the card not recognizing a disk in the drive. If you do take the reading would you let me know what it was in a post and I will set mine to the same value. Thanks.