AVPC Card went flaky

[atrax27407]

Just a Speech Synthesizer.

Edited February 2, 2020 by atrax27407

[+9640News]

And he should have the PEBox interface card.  I would pull out the speech synthesizer, and also if it is available, try a different PEBox interface card with no other cards present in the system.

 

Beery

 

[atrax27407]

It won't boot at all with just an Interface card and no other cards present. The HSGPL Card contains the system GROMs which have been removed from the console. It appears to be OK. All of the options are present on the MENU and the GROMs look to be intact. I'll probably eventually end up sending both the HRD16 and the BWG to Richard for testing and repair if needed (he is the authorized SNUG repair here in the US). I'll probably need someone to take a look at the AVPC as well.

 

The programs are not being saved correctly (or at least not stored correctly) on the HRD16. I have been getting a variety of error messages using both ARCHIVER and my normal disk copying programs. At this point I am not sure what the problem might be but, since the BWG has a clock in it and it is accessed all of the time, there appears to be some"bleed over" from the HRD16 that is affecting the Disk Controller as well.

 

I might take a break for a day or so before I attack this latest problem. What I really need is a test program that will test a 3 Meg HRD16. That might give me some insight into exactly what is going on with it.   

[atrax27407]

As per Richard's instructions (SWIM) I replaced the 74LS245 on the HRD16 without any change in performance. I am still getting the odd "Klingon" screen and I can't extract files with ARCHIVER and a bunch of them don't copy/load properly.  The BWG just "goes out to lunch" in the middle of a copy/extraction. I'm taking a break. 

Edited February 2, 2020 by atrax27407

[+9640News]

 

16 hours ago, atrax27407 said:

It won't boot at all with just an Interface card and no other cards present. The HSGPL Card contains the system GROMs which have been removed from the console. It appears to be OK. 

OK, Have you tried then just the HSGPL card and a different Interface card if you have one?  Also, disconnect power from the floppy drives.

[atrax27407]

The agenda for tomorrow:

 

Pull the BWG Disk controller and replace it (at least temporarily) with a CorComp DSDD Disk Controller. If it works, the BWG is in need of repair. In which case, I will replace the 74LS245 in that card that has caused some strange behavior (auth. Richard).

 

If that doesn't solve the issue, I will swap out Interface cards.

 

That should eliminate everything except for the AVPC card and I will need to find someone to work on it.

 

I verified that the 74LS245s on the HRD16 are good, the HSGPL appears to be operating normally, and the 32K portion of the SAMS is working since the E/A cartridge boots. That leaves the Interface Card, the BWG, and the AVPC to be eliminated.

[atrax27407]

OK- I pulled the BWG Dist Controller and replaced it with a CorComp DSDD FDC. No change. I cannot extract files from an archive and copy files, etc. is problematic. I tested the 74LS245 in the BWG that is sometimes problematic and it is good. That would indicate that the BWG is probably good. I exchanged the Interface Card. Also no change in performance.

 

OK - the winner is -WHOEVER SUGGESTED THAT IT WAS THE HSGPL CARD AND CORRUPTED GROMs. I have my "Black Max" console installed with the AVPC and stock GROMs

 

At last, the system is up and running running as it was before albeit WITHOUT the HSGPL Card. Now, I just have to figure out how to re-flash the GROMs. Thanks to everyone for their help and suggestions.

 

 

 

 

 

 

 

[+InsaneMultitasker]

Wonder if it corrupt GROMs or a problem with one or more of the logic chips, regulators, or caps.  Bet it feels good to be back in business to a degree!

[atrax27407]

It feels really good to be up and running!

[atrax27407]

It seems the I was premature in thinking that my TI system was back to normal. It crashed again! I can load CFG and "see" the HRD16 but loading ROS is sometimes problematic and I am unable to load any programs into it it. Using ARCHIVER (which I can load from an E/A cart, I can catalog disks (including the HRD16), but attempts to load files anywhere - even between DSK1 and DSK2 "crap out". I spent the better part of the overnight hours alternately dropping the HRD "flat" and then going through the process of formatting it and trying to load files into it.

 

I have switched and altered consoles, moved PEB cards in and out, replaced chips on cards, and, quite frankly am really weary of delving into the innards of consoles and changing out all of the cards that I have in my inventory. The end result is that I am out of TI business for the forseeable future. I have run out of ideas and will likely NOT be spending more time on a regular basis trying to fix my system for some weeks (months). It is simply getting to the point where the loss of sleep and antagonizing my "better half" are providing diminishing returns and certainly no longer considered "a fun hobby". I am NOT having any lasting success trying to get my TI gear up and running and it is disheartening and frustrating. I'm going to take a break from the TI for awhile but I will remain on AtariAge and continue to provide replacement/enhancement EPROMs to the community. So, that is where things stand at the present time. Who knows, I might even be able to get a copy of MESS running on my laptop.

[atrax27407]

Still working on the problem - I guess after 30 years, I'm just to stubborn to completely give up on my TI gear. I' still soliciting ideas.

[atrax27407]

I'm thinking that it might be a flaky VDP chip or two on the AVPC. Suggestions?

[+InsaneMultitasker]

It seems that you are experiencing a lot of ramdisk-related problems so my advice is to remove the ramdisk(s) until you can confirm the rest of the system is stable.   That little switch on the ramdisk does not take the card out of the equation! If there are bad buffer chips or other circuitry on the card it can still wreak havoc. 

[atrax27407]

I think (hope) the problem is FINALLY SOLVED. I found a working, unmodified console in the dark recesses of my TI closet. I even found an old audio/video cable to go with it. I pulled the AVPC from the PEB, changed my RGB monitor to CVBS and fired the system up. Everything worked as it was supposed to work. I could copy and extract file to wherever I wanted. The AVPC had been modified in 2009 and there were apparently two 4464 dynamic RAM chips that were replaced (they were socketed) while all of the others soldered directly to the board. 

 

My thought is that, since everything passes through VDP, one or more of the 4464s went flaky. The 192K VDP in the AVPC acts sort of like a cache when copying and extracting files. The Disk Controller copies them and they pass through the VDP and are then passed to the RAMdisk (or physical disk) in "chunks". With the flaky VDP, everything just sort of got stuffed up and transfers came to a screeching halt. That was very likely the problem all along since disk copying worked fine with the AVPC out of the system. 

 

There is no way to tell for certain which of the 4464's is bad. It might be one of the socketed replacement chips or one of the four remaining originals that had soldered chips. I'm going to have the AVPC repaired. I have ordered new 4464s, and sockets for the other four chips and that should completely solve the problem. Now, I'm just waiting for the parts to arrive. I probably should have trusted my instincts in the beginning - I always thought the AVPC was suspect. 

 

I might even get my HSGPL Card back in the system.

Edited February 5, 2020 by atrax27407

[Swim]

If it's just the 4464's you think are bad you might be able to test the two socketted chips in your programmer and see if they are good. If they test good and you have to desolder the other 4464's don't forget to heatsink the chip temporarily while desoldering. Then once you have the chip out you can install sockets to make things easier in the future if problems return.

[+arcadeshopper]

i have found the minipro test to test chips for basic function only.. ram that passed in mine fails in Matt's memory test in a SAMS board  ymmv

 

 

[atrax27407]

I have decided to take the cautious approach - have sockets put in for the other four chips and replace all six of them with good chips.

[Ed in SoDak]

If you plan to discard the desoldered chips, you can snip the legs off and remove each pin individually. Probably much safer to the board traces as well, since you can observe them as you work.

 

Much easier than trying to get all of the pins on a chip clear and free, then pull it intact. The only advantage of this method is you could possibly reuse that chip if it was still good. Or to be able to test it by substitution in a socket with a known good chip, and perhaps discover if it was actually bad or not.

 

You could do one chip/socket at a time and retest. Starting with the socketed chips. Maybe you get lucky and find the bad chip right away. And if your work creates a new problem, you catch that in time to correct your last couple of steps. If you do all before testing any of it, you may not have a clue where things went awry.

-Ed

[atrax27407]

The absolute "master" is going to do the de-soldering and putting in sockets. I have the utmost confidence is his ability (as should the community at large). Even with my cataract surgery and new glasses, I am not confident in my own abilities to complete such a task ?.