7800 Chip replacements

[WicoKid]

I have a nonworking 7800 that I am overly determined to fix. Have replaced the RAM/ROM/Maria. If I have a 2600 for parts, are the Sally and TIA direct replacements for the 7800 board?

[+DrVenkman]

10 minutes ago, WicoKid said:

I have a nonworking 7800 that I am overly determined to fix. Have replaced the RAM/ROM/Maria. If I have a 2600 for parts, are the Sally and TIA direct replacements for the 7800 board?

Yes for TIA (and RIOT if you need to replace that). But 2600's don't use SALLY. They use a reduced-pin count version of the base 6502 design called a 6507. The 7800 uses a SALLY, the Atari-customized version of the basic 6502 that adds a hardware HALT line to the pinout. 

[WicoKid]

Thank you

[+Stephen Moss]

8 hours ago, WicoKid said:

I have a nonworking 7800 that I am overly determined to fix. Have replaced the RAM/ROM/Maria. If I have a 2600 for parts, are the Sally and TIA direct replacements for the 7800 board?

Randomly replacing parts that may be perfectly OK could just be a lot of wasted time and effort as you could still be left with a non functioning unit, perhaps try searching the forum to see if anyone has had the same problem for which a solution was found.

 

If that turns up nothing try describing what the problem is, that would potentially enable some of those here to narrow down the problem to a specific part of the circuit.

[juansolo]

^ This.

 

If it's completely dead the first thing to do, if you haven't already, test the PSU. Make sure what's coming out of that is correct. Then check that it's getting to the regulator ok, then out the other side. Make sure you've got power going to each of your ICs. In theory at that point, you've got good power. Then it all depends on what it's doing/not doing as to where to go next.

Edited December 9, 2021 by juansolo

[WicoKid]

I’ve searched the forum quiet a bit. Initially it was inconsistently booting only 2600 games. The cartridge slot wasn’t positionally functional but looked awful even after aggressive cleaning  (rust and green) so replaced without change in function. Then put on sockets and replaced RAM and ROM and worked great initially....put original ROM  in and worked, then got goofy with graphic artifacts, progressive worsening to only booting 7800 robotron then only graphical glitched 7800 startup screen then where I am now, generally black screen sometimes vertical solids in various colors...it changes but not convinced it behaves differently whether a cartridge is inserted or not. Then socketed Maria with new chip, no change.

The power supply is a replacement from Best. The large trace is correct voltage....early on each chip had correct voltage but will recheck.

 

Thank you.

 

 

 

Edited December 9, 2021 by WicoKid

[+x=usr(1536)]

6 hours ago, WicoKid said:

I’ve searched the forum quiet a bit. Initially it was inconsistently booting only 2600 games. The cartridge slot wasn’t positionally functional but looked awful even after aggressive cleaning  (rust and green) so replaced without change in function. Then put on sockets and replaced RAM and ROM and worked great initially....put original ROM  in and worked, then got goofy with graphic artifacts, progressive worsening to only booting 7800 robotron then only graphical glitched 7800 startup screen then where I am now, generally black screen sometimes vertical solids in various colors...it changes but not convinced it behaves differently whether a cartridge is inserted or not. Then socketed Maria with new chip, no change.

The power supply is a replacement from Best. The large trace is correct voltage....early on each chip had correct voltage but will recheck.

 

Thank you.

 

 

 

Here's what I'd suggest:

 

Check power after the machine has been running for 10-15 minutes.  Do this both at the power jack as well as at a component level.  Give the ICs the finger heat test at this time, or when the glitching starts to occur through when it becomes unusable.

 

Since you mentioned the cartridge slot being oxidised, it would probably be worth going around all socketed ICs and cleaning both the legs and sockets (even if newly-installed) just to eliminate them as a possibility.

 

Look for dry solder joins, including at the cartridge connector.

 

While it's entirely possible that there is an IC or passive component(s) causing the problems you're seeing, definitely eliminate the basics first.

[juansolo]

Yeah, the fact you have corrosion on your cart pins would suggest it's been somewhere humid for some time. As suggested a thorough inspection of all your pins and the board in general. Remove, clean and reseat anything that's socketted. You may need to replace the sockets if they're corroded. I'd also as a matter of course re-flow all the pads for any mechanical connections (cart slot, switches, joystick ports and power jack), if there are any cold joints, that's where they'll be.

[+Stephen Moss]

On 12/10/2021 at 3:49 AM, x=usr(1536) said:

Check power after the machine has been running for 10-15 minutes.  Do this both at the power jack as well as at a component level.  Give the ICs the finger heat test at this time, or when the glitching starts to occur through when it becomes unusable.

Particularly the output pf eth 5V regulator to check it has not gone into thermal shutdown.

 

I would also check the pins of U3, it is an OR gate so if any input of a specific gate is High (5V) the corresponding output pin (, 8, ,6 or 11) will also be high. This would be easier if it was fox black video rather than the occasional flicker of colour but should indicate if the problem is with the RF modualtor. If you have the schematic available on AA the pins that have text beginning with T = TIA video, those beginning with M = Maria video, for 7800 games I would expect the TIA related U3 input pins to be 0V and the Maria pins to be constantly changing if video is being produced resulting a voltage (read on a multi-meter)  somewhere between 0 & 5 and the same on the output pins, for 2600 games Maria pins should be 0V and the TIA pins changing.

 

As the issue occurs with both 2600 and 7800 games the problem then is must relate to a common component, in addition the the final video stage/RF moduator and power, other common things to check would be...

Clock, difficult to measure with a multi-meter but anything other than 5V or 0V would suggest it is running, check it is getting to all relevant pins 

Reset pins, when the unit is on all reset pins should be pulled high to 5V, and down to 0V when off.   

[WicoKid]

Got a chance to try some things. On power up with either a 2600 or 7800 game I typically am getting a signal, it varies, usually either a black screen )signal detected) or a horizontal noise on black screen rolling upward intementently with TV frequently going from signal detected to no signal and less commonly a solid white or red or white even less vertical stripes. I can tell no difference whether a 7800 cart or 2600 cart is inserted and even not real different without a cartridge. 

 

I have an oscillocope but its abilities are beyond my knowledge currently. However, the chips seem to be getting a steady 5 volts. U3 (74ls32) I think is a mess. For anyone who knows this on the fly, examples with 2600 cart for U3 (74LS32)Pin 1 pulses between 1 and 4 volts, pin 2 pulses between 1 and 4 volts, pin 3 pulses between 2 and 5 volts, pin 4 pulses  between 2.5 and 5 volts, pin 5 pulses between 1 and 5 volts, pin 6 pulses between 1 and 5 volts. With a 7800 cart pin s 1 and 2 pulse between 1 and 4 volts, pin 3 fpulses between 1 and 5 volts; pin 4 a noisy 1.3 volt, pin 5 pulses between 1 and 4 volts and pin 6 pulses between 2 and 5 volts. As far as the clock on U8 (TIA) pin 11 I seem to be getting a good regular wave. On U2 (6502C) pin 40 (reset) with a 2600 cart reset is at a noisy 5 volts.  U9 (RIOT) pin 34 (reset) at 4 volts both 7800 and 2600 cart.

 

I have a working 7800 and have started trying the chips from the nonworking board on the working board.  I socketed the BIOS, U110, U11, and behavior did not change at all. I confirmed U12 from the non working board does work on the working board. 

 

Currently I am socketing the 6502C on the bad board but have mangled some traces so stoped for now.....I have to clean up the pins on the 6502 and repair one of them, but once that is done will try it out in a 800xl. 

[WicoKid]

Pretty sure I'm side tracked with a start up problem on or about U10. When I check Pin 14 and 3 on a working 7800 I get 2.5 volts prior to turning on then after turning on I measure  on Pin 14 and 3 4.79 volts. On the still not working 7800, it no longer turns on (no LED, no sound, no video), when I measure Pin 14 and 3 on U10 prior to turning I get 4.76 volts on both, when I turn on, nothing changes (same voltage, no LED, no sound, no video). The momentary switch on the nonworking does close when I click it.

 

I am trying to figure out how the unregulated 5 volt line arrives at U10 at 2.5 volts on the working unit and at 4.76 volts on the non working unit.

 

U10 is socketed and same behavior on switching.

Edited December 20, 2021 by WicoKid

[+DrVenkman]

If you haven’t already, you REALLY need to download the 7800 schematics and trace things out methodically. Just measuring voltages randomly, without knowing what’s connected to what, or rather, what is SUPPOSED to be connected to what, will not get you very far. 

[WicoKid]

I have schematics and reading those as well as measuring voltages measuring trace continuity and using this information (7800_Theory_of_ops.rtf, https://atariage.com/forums/index.php?app=core&module=attach&section=attach&attach_id=307005). Since the LED isn't even on, I am on "Page One" trying to sort out U10, On the working unit when plugged in but not "on"  Pin 3 and 14 are "low" or 0 when I'm measuring the 2.5 volts and when I turn the 7800 "on" they both measure 4.76 volts which I am assuming is "high" or 1. 

 

Not quite sure what to make of it all when the "Theory of Op" document page one says  "The regulated +5 supply slowly charges up C54 through the 2.2 megohm resistor R48." but I'm looking directly at the schematic and R48 is 470K (R46 is 2.2 megohm but not in series with C54).

[+DrVenkman]

Forget "Theory of Operation" and stick to the schematics and what you see in front of you. 

 

 

U10 is a 4013 - is the chip good? Do you have a chip tester you can use to test it? Second question - have you studied all the discrete components that tie into the various pins? This flip-flop is the chip that turns on the entire system. There are a lot of caps, resistors and even a Zener diode there that, if any of them are partly or fully shorted will prevent the 4013 from operating correctly and stop the unit from powering on. Also connected to U10 is a transistor which is connected to both ground and another couple resistors and another transistor. 

 

If it was my 7800, first thing I'd do is order some spare 4013's and try a swap. If that doesn't get things going, remove and check those transistors. Then start looking for shorted caps. Etc. 

[WicoKid]

Thanks Doc. It was the 1n3904 at Q9; replaced it and the power circuit is working again. 

 

Now booting to solid white with 2600 Carts and black with 7800 carts. I did socket the 6502C and tried it in the 800XL...it works fine. So planning to replace all of the transistors and see. 

[+DrVenkman]

Do you have a spare TIA from a working 2600? That would be the next thing I’d check, as TIA is used for all 2600 games and almost all 7800 games (Ballblazer doesn’t use it, so far as I can remember).

[WicoKid]

Socketed the TIA and switched with an Atari 2600 TIA and both chips work in the 2600, essentially  same behavior in 7800....solid stable white screen with 2600 carts, white screen with 7800 carts, no sound, no effect with reset.

 

Only leaves the RIOT for the major chips.....will try when able.

[WicoKid]

Socketed the RIOT and still the the same. Now working through schematics and double triple checking traces and with a significant amount of rust as well as traumatic socketing of chips, I'm gathering jumper wires. Now occasionally able to boot a 2600 cart after multiple on/off cycles but have yet to boot a 7800 cart.

 

One inconsistency I THINK there is between my board and the schematic: On the schematic, Q13 has PIN 2 for the cartridge slot (J1) coming off of the emitter whereas my board clearly has PIN 2 of the slot connector on the Q13 collector. On my Board the emitter has direct connection to PIN 40 of Maria and that is it, no direct connection oto PIN 2 of cartridge connector. So on schematic Q13 collector  has direct connections to Sally 35 and R63 BUT on the board Q13 collector is connected to Sally 35 and R63 AND cartridge slot Pin2 (and traces that seem to verify this)

[WicoKid]

All chips are socketed now.....replaced U3 and U4; usually no video, occasionally black with flicker. I am measuring 5 volts to chips when powered on. On U10 pin 14 there is a connection to UNREGULATED 5 volts that only reads 4.83 V  before and after powered on. I know the unregulated 9 volt comes from the power supply but from where dose the UNREGULATED 5 volt arrive?....VR1 isn't awake until powered on. 

 

Also, can anyone tell me how A12 is supposed to connect to the BIOS?...on the schematics it looks like pin 2 but when I look at the board, I just cannot find a trace that makes this connection

 

 

[+Stephen Moss]

9 hours ago, WicoKid said:

On U10 pin 14 there is a connection to UNREGULATED 5 volts that only reads 4.83 V  before and after powered on. I know the unregulated 9 volt comes from the power supply but from where dose the UNREGULATED 5 volt arrive?....VR1 isn't awake until powered on. 

It is derived by the 5V1 Zenner diode CR5 and R53, it is called unregulated because the voltage derived using Zenners is generally less precise than that of  a voltage regulator. Its only function is to power U10 (pin14) and send it a reset pulse. When the power unit it switched on a reset pulse is sent to pin 4 via C55 and then pin 4 is held low by R47.

At that point the Q output (Pin1) should be 0, turning Q9 off which in turn keeps Q10 off and no power reaches the input of the voltage regulator and the not Q output (Pin2) is high (5V) driving the D input (Pin5), so the output toggles between 0 and 5 with each press of the power switch, and driving R input of the 2nd D-type (Pin10) which when 5V holds the system in reset.

 

When the On/Off switch is pressed the state of those pins should just toggle from 0 to 5, and 5 to 0 respectively and back again. When in the on state Q (Pin1) is 5V turning on Q9 which pulls the base of Q10 to ground turning it on and allowing the 9V supply to reach the input of the Regulator.

On 12/21/2021 at 7:32 PM, WicoKid said:

Not quite sure what to make of it all when the "Theory of Op" document page one says  "The regulated +5 supply slowly charges up C54 through the 2.2 megohm resistor R48." but I'm looking directly at the schematic and R48 is 470K (R46 is 2.2 megohm but not in series with C54).

Either that is a mistake or refers to a different version of the schematic, perhaps they decided a longer delay was required to allow the regulated 5V more time to settle before C54 charges to a high enough voltage to for the S input (Pin8), to set the Q output (Pin13) high taking the system out of reset.    

[WicoKid]

I think the on/off circuit is working OK, at least as far as coming on and delivering 5 volts to the chips and the toggling of those pins on U10. The voltages I'm measuring are 4.6 to 4.8 however.....is there a reset signal that may not be making to the chips because of the low voltages I am measuring or is 4.68 OK for an unregulated 5 volt?

[+Stephen Moss]

4.68 volts is a little lower than I, personally would like but volts should be fine, the schematic indicate that a 4v7 or 5v1 diode could be used, from that voltage I would presume your unit has the former as that is a long way off for a 5v1 diode.

As I recall the supply voltage tolerance on logic devices is +/-10%, so anything above 4.5 should be OK, less than that and things could get a little dicey. The voltage thresholds for a logic Low is generally 0 - 0.4V, and for a logic High generally 2.7V and above, anything between those is referred to as the forbidden voltage because it could be interpreted as either a logic High or a logic Low or alternate randomly between the two.

 

So a 4.68V supply voltage is within the supply specification for logic devices and should produce a logic high output of about 3.8-4V which is well above the aforementioned input threshold, if you would prefer the supply voltage to be closer to 5V then you could try replacing the Zenner with another one.