Matt's Asteroids Troubleshooting / Repair Log

[matman]

My asteroids machine is exhibiting some issues and thought I would put them all in one thread.

1. Some characters are not being drawn properly. There appears to be some 'pixels' shifted in the vertical direction, but only along the centerish of the screen. (See images where digit 2 is impacted). Need to recheck test crosshatch pattern.

2. Some of the asteroids are missing lines, not sure if it is related to #1.

3. There are times where the picture will jitter in the Y direction. Will try to 'wipe' the y-gain pot and see if the problem persists or not.

4. On the AR board, 1N4001 was burnt to a crisp upon purchase. Have replaced with 1N4007 per expert Andy's advice and seems fine so far.

5. Need to undo the Fuse1 bypass on the power block as previous owner hacked this.

6. Audio hum present but reduced when fiddling with volume control. Either the knob needs to be replaced or I am grounding it when I touch it.

7. Player 1 and Player 2 buttons do not light up

 

[andrewb]

The glitches and jitter are analog issues on the game board. Bad 4016's, TL082/81's, and bad DACs in cases (though in this case I'd suspect the 4016's first.)

Also, for missing lines, make sure your brightness and contrast are properly adjusted.

Hum is usually bad TDA2002's on the AR, or a bad LM324 on the game board (if hitting the volume pot with some DeOxit doesn't fix it).

First step in fixing the LED's is to make sure they're connected properly. People often hook up the connections wrong/backwards. Else the LEDs inside the buttons can go bad.

 

[matman]

Had a little time to look at the machine today.

Turning up the brightness seemed to resolve the missing lines, however i'm not sure if it is normal for some lines of the asteroid to be thicker than others.

Throwing into test mode give me the attached images. The diagonals are not perfectly straight (a little bowed) and there are 'glitches' at almost all the intersections, including through the brightness gradient in center of screen. The upper right corner doesn't quite connect, it flutters up and down.

I will look at the TL081s and 82s.

 

[VectorCollector]

Diagonal bowing can be fixed with the linearity pots on the deflection board.
Fluttering can be caused by dirty pots on main logic board (X/Y) and/or deflection board. Or cold solder joints or poor connections to/inside monitor.

 

[DWJ]

Had a little time to look at the machine today.

Turning up the brightness seemed to resolve the missing lines, however i'm not sure if it is normal for some lines of the asteroid to be thicker than others.

Throwing into test mode give me the attached images. The diagonals are not perfectly straight (a little bowed) and there are 'glitches' at almost all the intersections, including through the brightness gradient in center of screen. The upper right corner doesn't quite connect, it flutters up and down.

I will look at the TL081s and 82s.

Every 256th count is confused... if it's not in the analog section, check the 74LS191 chips in row 9 (depending on board revision?) that turn DVY0-9 into UNMDACY0 through 9. Might not hurt to do the same for Y axis. Basically, start at the analog section, and work backwards through the 374s that turn DACY1-9 into DACY1* to DACY8*, then the UNMDACY outputs from the 191 counters, the 7497 rate multipliers, and possibly the LS273s that turn DDMA0-whatever into DVY0-whatever.

If you have a scope, and if it's a digital problem, you'll see something like an input pin that receives a signal that spends 99% of its time at either 0V or 5V, and an output pin with a non-TTLish output of 0-1.5V, or noise. Or you'll see an input pin that's fluctuating between 0V and 5V, and an output pin that *should* be fluctuating based on its input, but is instead stuck at 5V or 0V.

 

[matman]

Diagonal bowing can be fixed with the linearity pots on the deflection board.
Fluttering can be caused by dirty pots on main logic board (X/Y) and/or deflection board. Or cold solder joints or poor connections to/inside monitor.

I did a fairly comprehensive monitor rebuild but could have missed something. I think my x pot is at risk of breaking off. What is the replacement for this.? I must be missing it in tm-134 doc.

 

[VectorCollector]

KLOV has answers for everything. 3386

https://forums.arcade-museum.com/showthread.php?t=392185

(pic courtesy of andrewb)

 

[andrewb]

The glitching is analog. The 4016's usually fix that. There are other threads here about it.

And you can use 3386 pots, you just need to bend the legs, and put some kapton tape under them, so they don't short to the ground plane. That's what I do on all of the boards I refurb.

 

[matman]

Thanks, I had read that in your other posts about the tape. All good feedback. Will start with the analog section and go from there.

 

[andrewb]

Thanks, I had read that in your other posts about the tape. All good feedback. Will start with the analog section and go from there.

Yeah, you can get other types of pots that fit better, in terms of lead spacing, but there are other tradeoffs. I like the 3386's because 1) they're low-profie, so less likely to get whacked or caught on stuff and broken off, 2) they're sealed, so grime can't get in them, 3) they're cheap, if you buy them in bulk, and 4) you need a screwdriver to adjust them, which is nice because you know they can't get accidentally knocked out of adjustment (which also applies to shipping).

The last point is especially beneficial because if you ever do see the screen change, you know it wasn't due to the pots, so that rules out one variable, and you know you have other issues. And for color vectors, knocking the pots accidentally can actually blow your monitor in some cases. If the centering pots get maxed out, it is possible to fry the deflection board, just from the DC value for either axis getting too high. This isn't as easy with b/w's (which are a little more resilient and tolerant of bad signaling), but the other things still apply.

 

[matman]

All good info from everyone. I'll hopefully have time this weekend to narrow this down more. That said, I'm also in the middle of trying to close out my 10 year robotron restoration (I'm not that slow...more important things in life came up like getting married and having kids).

 

[Alpha-tron]

Since, it hasn't been mentioned yet. I am going to mentioned it now.

Repinning the connectors is an important step to make sure you have a good connection from the power to the pcb and the monitor to the pcb.

Dirty, corrosion, weak pins will cause you issues either now or down the line.. infecting them and checking for tension is a good step in making your machine work or last longer between repairs.

 

[matman]

I swapped out all the 81s, 82s and 4016s (B12 and D12) and that had no effect on the glitching (shifted pixels or fluttering in the upper right corner). Do i need to change the other 4016s (M9, R11, N10)?

I also found that in my attempt to fix the linearity on monitor, one of my pots (600 I think) had absolutely no effect, so I will need to look into that as well.

I have not changed my X/Y gain pots on the main board, after further looking at them, they don't look so bad and seem to operate fine.

finally, a 555 at L9 looks like it has had better days. I may need to look that, but haven't looked into what that does.