Trying to revive an 800, no color, low volume, keyboard doesn't work

[StickJock]

So I've been trying to revive an 800 that I found recently.  It was my wife's from many years ago and I just recently came across it while cleaning out a closet.

I've cleaned the case and blown the dust off of the boards.  I just tried to power it up and this is what I have found:

 

One of the ram cards wasn't working.  Swapping each card out, I found the bad card, and then swapping chips, I isolated a single bad 16Kx1 ram chip.  I used the same diagnostic process that I did back in 83 or 84 when I had a ram issue.  At that time, I ordered two chips from Jameco, so that I would have a spare just in case another chip failed.  Amazingly, I was able to find the spare ram 4116, despite 36 years and many moves!  It didn't even take me long to find it!  So, RAM is now OK.

 

Boot to memopad.  Using composite, I only get a black & white image.  Same cable swapped to my original 800 works well, so it is not the display or cable.  I see this burnt resistor (R204) near the video stuff, so I am guessing that it may be contributing to the issue.  Also, the three caps on the vertical board have a white coating on them.  Are they ok? Pic of resistor & caps below.  I get no reading across the burnt resistor.

 

Also, keyboard is not working.  I tried shorting some pins on the keyboard connector, but still no dice.

 

Boot from SIO2SD.  Load up Necromancer.  Game runs, but volume is low.  SIO volume low as well.  Start key, joystick & trigger work.  I didn't try a paddle (I haven't found my paddles yet).

 

Works: Boots, SIO, sound (but low volume)

Doesn't work:  Color, keyboard, low volume

 

Any suggestions?  I'll change out the 9.1K R204, but what caused it to blow?  Is this related to the low volume or lack of color?  As for the keyboard, my first thought was the POKEY chip, but the music in Necromancer sounds good.  Maybe one of the keyboard controller chips (Z103, Z104)?

 

Thanks for the help!

Scott

 

 

no

[xrbrevin]

here are 2x variants of the service manual for the 800

R204 is related to pin 4 of the RF modulator. Is this how you are connecting it to a display? If so, can you try a monitor cable (composite or svhs)?

Sams_Computer_facts_Atari_800.pdf Atari_400_800_Field_Service_Manual_June_82.pdf

[StickJock]

I am using a monitor cable connected to the composite in of the TV.

 

I was guessing on the R204 since it is burned and it is located near the RF modulator.  If it is only used for RF, then maybe I'll just leave it alone.

[Gunstar]

Are you using the same TV/monitor to check color that you use with your other 800? My 800 will show in B/W only plugged into ALL my modern LCD TV's and the only way I get color is through my Ambery video-to-VGA converter hooked up to VGA on those same TV's. I've been told that many modern TV's won't display the color coming from old stuff like the 800.

[StickJock]

Yes.  Same TV, Same cable.  Plug cable into old 800 & everything looks perfect.  Plug cable into new 800 and everything is in B&W and volume is significantly reduced.

 

TV works fine on 320XE, too.

 

Edited March 2, 2020 by StickJock

[xrbrevin]

try swapping over the POKEY chips. POKEY controls the keyboard and the audio so maybe its the common denominator?

[+DrVenkman]

6 hours ago, StickJock said:

  I'll change out the 9.1K R204, but what caused it to blow?  Is this related to the low volume or lack of color?  As for the keyboard, my first thought was the POKEY chip, but the music in Necromancer sounds good.  Maybe one of the keyboard controller chips (Z103, Z104)?

That resistor is toasted! I'd guess a static discharge, maybe through the TV. A lightning strike nearby maybe when it was plugged in? That kind of thing could also explain the dead DRAM chip and your keyboard issues. As for the keyboard, if swapping POKEY doesn't resolve it, the keyboard mux chips (the 4051 chips) would be the next thing to check. 

[StickJock]

So here are a couple more pics showing some suspicious components.

 

 

It looks like C107 has popped, and there is some corrosion on the left side leads of C108 & C111.  I checked if C107 is shorted to ground, but it is not.

On the other pic, there is corrosion on R126, but measuring across the pads on the back of the board still shows about 100 ohms, so it seems to be ok.

I removed the two mux chips (Z103, Z104) and reinserted them in case they weren't making good contact, and did the same with all of the chips on the CPU card, but the behavior did not change.  Still B&W and keyboard doesn't work.

 

 

[+tf_hh]

18 hours ago, StickJock said:

Any suggestions?  I'll change out the 9.1K R204, but what caused it to blow?  Is this related to the low volume or lack of color?  As for the keyboard, my first thought was the POKEY chip, but the music in Necromancer sounds good.  Maybe one of the keyboard controller chips (Z103, Z104)?

no

 

R204 should be 75 ohms. It is the major pull-down resistor for the chroma (color) amplification around transistor A104B. The way it´s burned the only possibility is that somebody has sometime put a power supply to the monitor jack of this Atari 800. Sometimes Atari has built here 130 or 390 ohms in, but this is wrong. The correct impedance for video is 75 ohm.

 

I would suggest to check all parts twice if there are any more burned components. Also take a look at the solder side, specially around the connector to the power board, maybe a trace is burned also.

 

[+DrVenkman]

3 hours ago, tf_hh said:

R204 should be 75 ohms. It is the major pull-down resistor for the chroma (color) amplification around transistor A104B. ... Sometimes Atari has built here 130 or 390 ohms in ...

Are you sure about that? I have a late-production 800 with the same 9.1K resistor there. I’m using S-video and composite rather than RF so I guess it doesn’t matter. 

 

 

[+tf_hh]

13 minutes ago, DrVenkman said:

Are you sure about that? I have a late-production 800 with the same 9.1K resistor there. I’m using S-video and composite rather than RF so I guess it doesn’t matter.

 

Yes - and no ? - You´re absolutely right, my power PCB has also 9.1K at this place. But Atari did something weird, R204 two time exists - one at the power PCB (with 9.1K) and one left side of the variable inductor on the mainboard. This one was the one I meaned, and it´s 75R as shown in the FSM, too. From memory I didn´t recall when writing this above, that R204 should be on the mainboard, not on the power PCB.

 

R204 at the power PCB is connected to the modulator mixed video/audio input and ground. I would like to know HOW this could be burned this way...

 

[+DrVenkman]

3 minutes ago, tf_hh said:

R204 at the power PCB is connected to the modulator mixed video/audio input and ground. I would like to know HOW this could be burned this way...

I already suggested a lightning strike while the computer was plugged in and connected to a TV or display device. That might also explain the dead DRAM chip and other random burned components in the system. Induced currents can do some very odd things, and lightning nearby can induce some large currents indeed.

[ChildOfCv]

2 hours ago, tf_hh said:

 

Yes - and no ? - You´re absolutely right, my power PCB has also 9.1K at this place. But Atari did something weird, R204 two time exists - one at the power PCB (with 9.1K) and one left side of the variable inductor on the mainboard. This one was the one I meaned, and it´s 75R as shown in the FSM, too. From memory I didn´t recall when writing this above, that R204 should be on the mainboard, not on the power PCB.

 

R204 at the power PCB is connected to the modulator mixed video/audio input and ground. I would like to know HOW this could be burned this way...

 

Yeah Sams lists them as R204 and R204A.  Strange that Atari reused some of the reference numbers.

[ChildOfCv]

2 hours ago, DrVenkman said:

I already suggested a lightning strike while the computer was plugged in and connected to a TV or display device. That might also explain the dead DRAM chip and other random burned components in the system. Induced currents can do some very odd things, and lightning nearby can induce some large currents indeed.

In order for lightning to destroy the resistor there, though, it must have also toasted the RF modulator.  If so, it's possible that there are some shorted semiconductors in there.

[+DrVenkman]

7 hours ago, ChildOfCv said:

In order for lightning to destroy the resistor there, though, it must have also toasted the RF modulator.  If so, it's possible that there are some shorted semiconductors in there.

Nope, not necessarily. Like I said, induced currents are weird. As in, utterly bizarre. My wife's TV was destroyed as a kid when lightning struck a tree across the street, though nothing else electrical in the house was damaged. The TV actually exploded and the wires in the wall smoldered and had to be torn out and replaced. But the house circuit breaker never popped because, of course, no actual current was flowing through the breaker box. 

 

And it doesn't take lightning to induce a current. I accidentally plugged a 9V AC PSU into my 1088XEL last year and the only things damaged were in the XEL-CF card interface and the the CF card itself, believe it or not. Nothing in the power supply components, nor the logic chips, nor the UAV, Sparkfun board or U1MB was damaged. The CF card stuff is no where close, electrically-speaking, to the power input. But the AC induced some kind of flow where it didn't need to be and fried something. I'm just glad it didn't damage my Atari ICs (especially the two POKEY chips inside!).

[ChildOfCv]

I suppose you have a point.  After all, lightning is where electricity forges a path through thousands of meters of air, so what's a little insulation going to do?

 

But still, it's quite unusual for lightning to be so precise as to target a single resistor.  But it also wouldn't hurt to inspect anything else near the resistor

[StickJock]

So I finally got my 4051 mux chips from AliExpress.  I plugged them in and tried it out, but the keyboard still isn't working.

I popped out the Pokey chip.  Pins on the chip look ok.  Pin 20 on the socket looked bent, but I pried it back out and verified that the ship shoulder to the bottom of the board had continuity.  Tried it out again.  KB still not working.  So I guess that I need to get a new Pokey, huh?  Any other tests I can try first?  As I mentioned in the OP, music & SIO sound work, but are not as loud as they are on the other A8.  Maybe I should try another pair of muxes?  I got 10 in my order.

 

Also, any ideas on why I am not getting color on my composite output?  I know that the cable & TV are good since I get color on the other A8.

 

Thanks

 

[_The Doctor__]

replaced all of the dead resistors and capacitors?

verified voltages under load?

tuner cleaner/deoxit on all interconnects card edges, and color pot?

cards all in the correct locations and order?

made sure the keyboard switches are all still firmly soldered, no cracks, cold joints?

does rf modulator produce a picture or color?

checked transistors near or inline(tracewise) with previously burned components?

checked the ic's inline with previously toasted components?

 

Edited April 16, 2020 by _The Doctor__

[+kheller2]

As has been suggested, try swapping the pokey.  First try the bad keyboard on your old working 800.  In fact try the muxes too.  I would put the suspicious pokey in the working 800 and see if the same keyboard and sound problems exist.  

[_The Doctor__]

make sure nothing is shorting inside card slots or under pcb's...

 

-yup, once all the basic stuff is done, it's chip swapping time...

[StickJock]

So here's the latest on trying to revive this system:

 

I swapped keyboards.  Keyboard not working followed the keyboard, so that's pretty definitive.  My old KB works on new system.  New KB doesn't work on my old system.

 

BTW:  I sprayed an entire can of cleaner on the motherboard & power supply board about a month ago, early in my diagnostics.

 

Now, for the color & volume issue, I tried several things.  I tried the RF cable, after finding my old switchbox.  RF is really noisy, with lots of static, on both systems.  New system does not display color, and volume is even lower.  Back to composite!  I tried turning the color wheel (after marking it's position) to both extremes with no change.  I put the new pokey in the old system and the volume did not change.  Since a working KB works, I'm thinking that the Pokey is probably OK.  Whew!

 

I put the new CPU card in my old 800, and it didn't boot.  Got a click and a green screen.  I put the original old CPU card back in and it still didn't boot.  Looks like I fried some RAM.  I can move all of the new cards over to the old 800 and everything works fine.  BTW:  IIRC, one of the CPU cards said rev 5 and one said rev 6.  Or maybe that was the ROM card?  Anyway, with all three RAM cards, the ROM card & the CPU card moved into my old 800 motherboard, the color & volume are normal.

 

Then I tried swapping Pokey chips, and there was still no change.  Then I tried swapping power supply boards, and it didn't boot.  Oh Crap!  What did I break now?  Oh, I forgot to put the original Pokey back.  Feeling pretty dumb, I put it back and tried again.  It boots up fine, color & volume are fine using all new cards & new power supply board.  So it looks like the color issue and volume issue are somewhere on the new 800 motherboard.

 

So that's where I am at.  All 5 cards appear to be good.  Power supply board appears to be good (using composite).  Keyboard itself is bad, explaining why KB doesn't work.  No color and volume issues appear to be isolated to the motherboard.

 

Any more ideas?

 

Time to go figure out what I broke on my old 800 now.

 

 

 

[StickJock]

WooHoo!  I found it!  A104 is bad.  The schematic shows it in both the video & audio path.  I put the new one in my old 800 and the color went away & the volume went down.  I wasn't brave enough to put the good chip in the suspect mobo, tho.  Anyway, it looks like it is an LM3086N transistor array.  Oddly, the working one is labeled LM3086N and the nonworking one is labeled CA3086.

 

Anyway, time to order some more parts and then wait, wait, wait.

 

Thanks for all of the help & encouragement!

 

[StickJock]

BTW:  The old 800 is working again.  Phase of the moon?

 

[StickJock]

I was able to get the keyboard working again (I think).  It's a Mitsumi keyboard.  After unscrewing the many tiny screws and pulling the PCB off of the mechanical part, I carefully peeled the plastic silver ink piece off of the PCB.  It was stuck on pretty good, so I used a heat gun on low to help soften it.  I peeled it off everywhere except where the traces were bonded down to the PCB.  I then let it go back in place and reassembled the KB.  Using a multimeter, I was able to beep out the entire keyboard matrix (pins 1-8 against pins 9-17).  I haven't tested it out in the actual 800 yet, but since I was able to get every key to beep (some took a bit of "banging" on), I anticipate it all working fine.

Now I just have to wait for my 3086 chips to get in and I should have the entire 800 back in shape!

 

Then it is a matter of waiting for the Incognito to ship....

[Sugarland]

Thanks for the update. Enjoy reading your progress and learning what was the cause.