Tempest spinner quit

[RobMcRaf]

So now that I've got a couple good Aux boards and 1 CPU board that works, I took some readings of the other 2 (actually all three) with the monitor disconnected.
Going to make a new post since the voltages being off on the X & Y are not related to this thread.

 

[andrewb]

No need to start a new thread. Just keep it here.

 

[RobMcRaf]

Well Ok then!
3 main boards. 1 working, 2 not sure, probably not. Got 2 Aux boards to swap in. Checked all the interconnect cables, reflowed headers, etc... Checked with same power supply and Aux board.

#23605 (working reference)
Y-out DC ~ 1.5 max
Y-out AC ~ 1.4
X-out DC ~ .5 max
X-out AC ~ 1.2

#00174 (untested, intact, but leg on tantalum cap C56 has been lifted. Think I did that years ago due to it being shorted)

Y-out DC ~ +2.2 max
Y-out AC ~ 4.3
X-out DC ~ -.4 max
X-out AC ~ .1

#10289 (untested, intact, red LED does not light)
Y-out DC ~ -4.6 max
Y-out AC ~ 1.4
X-out DC ~ -6.8 max
X-out AC ~ 1.7

So, on those 2 untested boards, I can see that some of the voltages are too high.

 

[andrewb]

Keep in mind that those voltages are adjustable, via the pots. You can change the DC by several volts, just be turning the XCTR or YCTR pots, and the SIZE pots will control the AC. So those values aren't actually that bad, potentially aside from the last one, depending on if it's controllable or not. You can start by centering all of the pots on the main board, and see where you stand.

You really should have a scope here. But if you want to use your monitor, adjust the DC values to get them as close to zero (or an average of zero) as possible. Note that Tempest will drift up to about 3V briefly (for the Y DC), when it's drawing the scrolling TEMPEST logo during attract. But that's normal. The X DC should be pretty close to zero when adjusted.

However if you want to test them by hooking to a monitor, the best thing would be to boot the board with the test switch already turned on, so it automatically runs self test, and gives you a static screen. As long as the DC is within -1 and +1V, and the AC isn't 10V, you should be safe to connect to a 6100 and see what you're getting.

 

[RobMcRaf]

Yeah, I noticed that when playing with the pots. Some of them look a little sketchy anyways and I have some replacements that I got years ago.

Haven't hooked up scope in years so will do that just to see if I remember how! Still looking for the front clamshell cover to a Tek 454a

BTW, thanks for the help! Be back up in a week or 2.

Keep in mind that those voltages are adjustable, via the pots. You can change the DC by several volts, just be turning the XCTR or YCTR pots, and the SIZE pots will control the AC. So those values aren't actually that bad, potentially aside from the last one, depending on if it's controllable or not. You can start by centering all of the pots on the main board, and see where you stand.

You really should have a scope here. But if you want to use your monitor, adjust the DC values to get them as close to zero (or an average of zero) as possible. Note that Tempest will drift up to about 3V briefly (for the Y DC), when it's drawing the scrolling TEMPEST logo during attract. But that's normal. The X DC should be pretty close to zero when adjusted.

However if you want to test them by hooking to a monitor, the best thing would be to boot the board with the test switch already turned on, so it automatically runs self test, and gives you a static screen. As long as the DC is within -1 and +1V, and the AC isn't 10V, you should be safe to connect to a 6100 and see what you're getting.

 

[andrewb]

Also, do a simple test and just try to boot the boards (with monitor disconnected), and see if they play blind. If they don't, you have other issues, and there's not much value in checking the outputs at this point.

C56 is just a bypass filtering cap for the -22V supply. It isn't critical for operation, as there are other filter caps already (and more after the -15V regulator). You can just clip it and run without it for now. (I've had those short on me as well, as they're just old.)

As for the one with no LED, just test to make sure you have 5.00V on the board. If so, and the LED is off, it's just a bad LED. No problem trying to boot it.

 

[RobMcRaf]

Welp, got the oscope settings figured out and tested it on the good set.

For the bad boards, one of them is spitting out some vectors and rebooting
https://youtu.be/upVmadKgW4M

And the other one barely has a heartbeat.
https://youtu.be/YvgORQEd9fI

I may have to punt on these and send them to someone more skilled in the interest of just getting things done and out of the garage.

Also, do a simple test and just try to boot the boards (with monitor disconnected), and see if they play blind. If they don't, you have other issues, and there's not much value in checking the outputs at this point.

C56 is just a bypass filtering cap for the -22V supply. It isn't critical for operation, as there are other filter caps already (and more after the -15V regulator). You can just clip it and run without it for now. (I've had those short on me as well, as they're just old.)

As for the one with no LED, just test to make sure you have 5.00V on the board. If so, and the LED is off, it's just a bad LED. No problem trying to boot it.

 

[andrewb]

Try booting them with the test switch already turned on (if those vids are with it off). Do they still do the same thing?

 

[RobMcRaf]

Try booting them with the test switch already turned on (if those vids are with it off). Do they still do the same thing?

Duh! I should have done that. Thanks.

First pic is #00174

second pic is #10289 (it also put out a single tone upon power-up)

Had to change some of the scales on the scope to get them to fit.

 

[andrewb]

Bingo. Those are both showing ROM errors. (And the second one looks to have a Pokey error as well (if I'm reading it right), which again could be the Pokey, or the connection to the CPU.)

You'll need to connect to a monitor to get the specifics (and test the XY voltages before doing so, and DON'T take it out of test mode while it's connected. Just get the info and power down ASAP.) But start by cleaning and testing your ROMs, once you get the specifics.

This is why I single-ROM mod all Tempests, as many of them have ROM issues.

 

[RobMcRaf]

Cool. Yeah, I'll do that tomorrow.

Single-rom mod? Never thought to look that up. Will research that too.

 

[andrewb]

Cool. Yeah, I'll do that tomorrow.

Single-rom mod? Never thought to look that up. Will research that too.

http://www.arcade-cabinets.com/board_hacks/tempest.html

 

[RobMcRaf]

Well, that's not good.... stupid screwdrivers.

http://www.arcade-cabinets.com/board_hacks/tempest.html

Thanks for the link.

THIS could be a problem.. (PCB#00174)

FWIW, would I need to repair this before doing the single rom mod?

Also, I think there is a typo in the instructions:
"Desolder the chip socket at ROM slot D1, we will need to cut some traces underneath the socket.
In ROM chip slot D1, on the PCB board cut the thin trace going from pin 20 on the top side of the board) This is the "Output Enable" pin. The new chip will have an output enable pin in the same location, but we'll need to tie it to the the inverted output of address line 15. (the reason we cannot use the original trace is because since there were 10 chips, the output enable on ROM slot D1 will only be correctly set 1/10 of the time.
There is a trace (on the underside of the board) from pin 21 to 24 in ROM slot 0, Cut this trace, on the 2516 chip, this was Vpp which was tied to put 24 (Vcc) but the new chip in there this pin will be address line 11.
There is a trace connecting pin 24 to +5V (on the top side (parts side)) of the PCB board, cut this trace, as in the new chip the pin in this location is actually Address Bus bit 13
solder back the socket... or better solder in a new 24 pin IC socket."

Looking at my PCB, ROM @D1 = ROM slot X, and ROM @E1 = ROM slot 0.
Do you ever cut the pin 21-24 trace under chip @ E1?

 

[andrewb]

You'll want to fix that broken trace. A piece of stripped kynar wire and a fine iron and you can patch the trace directly, no jumper wires.

I don't cut any traces when I do these. I double stack a dual wipe socket onto the ROM X location, and bend the pins out that I need to do the mod, then solder wires to those, so I don't have to cut traces.

 

[RobMcRaf]

You'll want to fix that broken trace. A piece of stripped kynar wire and a fine iron and you can patch the trace directly, no jumper wires.

I don't cut any traces when I do these. I double stack a dual wipe socket onto the ROM X location, and bend the pins out that I need to do the mod, then solder wires to those, so I don't have to cut traces.

PCB 00174: 256K mod done last night, double checked wiring this morning with fresh eyes and plugged it in (with monitor disconnected). Didn't have time to set up scope before going to work.

Voltages now for X & Y's
(EDIT: I may have mixed up the X&Y values when writing down. Will have to check again after work....)
Y-DC .3
X-DC -.4
Y-AC 4.8-7.2
X-AC .9

 

[andrewb]

If it isn't playing blind, you have a problem.

 

[RobMcRaf]

Yeah, after work will sit down and start from zero going thru the t/s steps.

 

[RobMcRaf]

Yeahhhh buddy.

Started from scratch after work, set up switches, dials, etc....

and it works! PCB #00174 is good.

Thanks for the help!

 

[andrewb]

Cool.

Clean up that spaghetti wiring. For each of those signals, if you follow the traces, there's a much closer spot to the ROM that you can connect a much shorter wire, to, and not need to run long wires all the way over to the TTL chips. The result will be much cleaner (and less prone to damage).

 

[RobMcRaf]

Cool.

Clean up that spaghetti wiring. For each of those signals, if you follow the traces, there's a much closer spot to the ROM that you can connect a much shorter wire, to, and not need to run long wires all the way over to the TTL chips. The result will be much cleaner (and less prone to damage).

Yeah, it was pretty late and my hands aren't as steady as they used to be. My daughter has many shades of nail polish to tack those wires down