Rock on! Asteroids Cocktail Restoration

opt2not

Well-known member
Joined
May 24, 2009
Messages
1,580
Reaction score
241
Location
Orange County, California
I originally posted this to BYOAC, but figured I should get more eyes on it now that I'm at the "meaty" part of the restoration. I'll give you a quicker-run-down of what this project has been through so far (I like pictures, so I'm posting a bunch):

I got a hold of an original Asteroids Cocktail that was left out in an alley for a few months, till it was rescued and stored in a basement for 6 years.

index.php

The cabinet was in rough physical shape, but in remarkable condition considering being exposed to nature for so long. No wood swelling, no critters living in it...just rust, dust and grime...and of course not working.

index.php
index.php


I emptied the cabinet out, it had all of it's internals intact. Did a thorough cleaning inside, did some minor structural stabilizing (tightening screws, re-gluing, etc..). Removed the legs, stripped and repainted them, added new leg glides:
index.php
index.php
 
Stripped the flaking screened-artwork under the glass, cleaned and polished it as best I could. Removed the black electrical tape the previous owner used as the black outline art on the wood grain. Added new t-molding, new artwork underlay as well as a new monitor bezel. Replaced the midway glass clips that were on it, and got NOS Atari glass clips...all this from Arcadeshop. Still need to apply the Atari sticker decals (in that little baggy)
index.php


Took apart the coin panel, stripped it down with a wire-wheel, repainted with Metallic Silver for the main panel(yeah I know, it's not the original scraped metal look, but I like this look better), and repainted the black coin return flap.
Here's some before and after pics:
index.php
index.php


The control panels had minor paint damage on their front sections, but the screened overlay art was in great condition. So I covered the artwork up, stripped off the chipping paint on the front, then re-painted them. Added new Buttons, re-assembled, done!:
index.php



index.php
 
Cleaned up the cabinet base with a magic eraser, like I said, the cabinet was in remarkable shape based on the conditions it sat in. There were no chips, no gouges, just lots of dirt and minor scuffs in the vinyl that rubbed out with the eraser. Cleaned and re-painted the grills, added the new t-molding.

Aesthetic components complete!
index.php
index.php


Onto the electrical portion!

Like I said, the internals were still intact. When it was originally powered up, there was a high pitched noise coming from the speakers, no lights were on except for the PCB LED and some neck glow in the monitor (the monitor itself doesn't have much burn-in). I'm told having neck glow is a good sign of being able to save it.

index.php
index.php


I re-capped the AC PS, AR board, and PCB (kits from bob roberts) I also replaced a burnt diode on the AR (CR2 I believe)

Using dokert's trusty PS testing tips here...

Dokert said:
Power Brick
Remove the big 15 pin connector P5

***V***V***
*1***2***3*
*4***5***6*
*7***8***9*
*10**11**12*
*13**14**15*
***********

Meter set to 20vdc
Red lead to pins 1, 2, 3
Black lead to pins 4, 5
Reading should be 10.3 vdc (could read high 13vdc as it is unregulated)

I got 14.31 vdc without load. Might be a bit high...I tested the diodes on the Rectifier Board, (since I have an older version it doesn't have the bridge rectifier like many atari blocks have). They checked out fine. Other than big-blue and the fuse block and fuses (which were replaced with new ones), there really isn't much to this block other than the isolation transformer which apparently never dies!

So I'm not quite sure why the voltage out of these pins are so high. I'll do a bit more reading and researching, it might be a bad connection somewhere?

Dokert said:
Meter set to 200vac
Red lead to pin 6
Black lead to pin 7
Reading should be 36 vac

36 vac for me, perfect.

Dokert said:
Meter set to 20vac
Red lead to pin 8
Black lead to pin 9
Reading should be 6.3vac

6.5 vac, slightly high, but might be ok.

Dokert said:
Meter set to 200vac
Red lead to pin 10
Black lead to Pin 14
Reading should be 80vac

I have no pins for 10 and 14. I think his guide is for the stand-up, which may be the reason I don't see these pins on my brick. Those pins may be for powering the Marquee light and extra coin lamp?

Dokert said:
Meter set to 200vac
Red lead to pin 11
Black lead to pin 13
Reading should be 65vac

I got 61.1 vac, this might be too low.

Dokert said:
A/R I
Remove the 9 pin connector P7 (P5 on the power brick must be plugged back in)

__^__^__
*1**2**3*
*4**5**6*
*7**8**9*
*********

Set meter to 20vdc
Red lead to pins 5, 6
Black lead to pins 1, 4
Reading should be +5vdc (could be as high as 7 or 8 vdc because there is no load)

With my voltage selector turned all the way down, without load the reading was 5.46 vdc.

I plugged in the game PCB and got it to fire up. P1 and P2 were blinking, and the PCB LED was lit:
index.php

At least we have juice running through!
I checked voltages at the 5v and 10v PCB test points. Was able to adjust the AR board to get a good 5v running through, but the 10v line was sitting around 12.4. That's pretty high.

That pretty much brings us up to date, and figured I could get some of you guys to chime in.
I'm going to see if I can clean up the edge connect more thoroughly, and check all header pins. I read that it could help lower the voltages I'm getting.

Checking through the Asteroids Troubleshooting Encyclopedia, blinking P1 and P2 start buttons could mean one of three possibilities listed that I can look into:
a. Fix: Check the power supply, the large capacitor sometimes comes loose, you'll have to remove it from the cabinet to check it from the bottom.

b. Fix: When my board had that problem, it turned out to be loose caps on the board. Specifically, the largish ones on the right upper side. They mostly have to do with the sound FX circuits. But, if they come loose and make/break contact, they send surges that reset the board. The other suggestions I've seen are valid too. Those caps however, since they stick up off the board, may tend to get loose first, but you should probably check all solder joints.

c. Fix: Resocket all EPROMs.

Anyone have any other tips to try to lower that voltage?

In the meantime, I'll be taking apart the monitor and applying the repair-kit.
 
Good luck with your troubleshooting. That cabinet looks awesome.
 
All of the voltages that you have posted are within spec. The +5vdc could be as high as +7vdc with no load, and the +10.3vdc is unregulated and can be as high as +15vdc. You are correct, you do not have a marquee light thus no pins at that location. The 6.3vac is fine also.

Coin it up and see if you can hear the game playing when you push the start button.
 
Thanks for the replies. Good to know that the voltages are on point.

I should have been more specific, the P1 & P2 were flashing rapid fast. I attached my speaker and got nothing but buzzing.
At random, powering it up would cause the start buttons to flash slower, in demo mode like it should. I can press P1 to start a game (the dips are set to freeplay), and it'll stay constant. I still get only buzzing from the speaker. Though it seems when I physically move or bump the pcb it goes back to rapid flashing.
There must be a bad connection somewhere ? I'm assuming the rapid flashing means it's constantly resetting.
 
You have done well so far, but you do not have a known good monitor or game board. One or both could be bad. What make and model is the monitor?
 
Thanks Dokert, I really appreciate your help.

The monitor is an Electrohome G05-805:
index.php


I haven't tried connecting it just yet since the cabinet was gutted and the boards re-capped. I wanted to concentrate on getting the game playing blind before I tackled the monitor repairs.

The thing that bugs me is that when I turn the speaker pot up high, I only get a loud buzzing, not even any faint game-sounds in the background.
After a couple searches here, I found a few posts indicated that the loud buzzing (w/ no audio) could be due to a failing on-board audio amp. Richie Knucklez has a post here pin-pointing his fix.

But of course, the audio sits secondary to getting the game to run consistently without going into constant rapid-flashing-reset mode.

I'm going to take a look at the wiring harness tonight, the pins on the connector might not be making solid contact. Also going to take a look at all solder points, and test all diodes and resistors with my DMM.

Any other suggestions are highly welcomed!
 
Another thing I wanted to get you guys to chime in on, which might be the cause of the buzzing:

On the AR board originally had 1000uf 25v caps installed for C9 and C10, so I replaced them with new ones of those (1000uf 50v). Whereas Bob Robert's site has a note about the AR boards were actually shipped with 3300uf 35v caps installed.

The manual and schematics states it as being 1000uf 25v, and it directly connects to the speaker output.
*Actually the Cocktail manual's part list states 100uf 25v, the upright manuals state 1000uf 25v, and Bob 3300uf 35v!

Who should I believe?
To the cocktail owners out there, what do you guys have on your AR boards?
 
Last edited:
Asteroids uses an A/R I, which has 1000uF 25v caps. The A/R II uses 3300uF 35v caps at the same location.

The 1000uF 50v ones that BR sent are correct, the manual has a typo in it. A lot of manuals obviously did not get proof read BITD.
 
Rapid flashing P1/P2 lights indicate the board is resetting. If on the other hand it was set for free play, you'd get a slower flash rate in attract mode, and a solid P1 light after pressing it.

Atari boards are notorious for flaky sockets, so depending on your board repair skills, those could be replaced. First, I'd try reseating each ROM and the 40 pin 6502 processor.

Also, put it in test mode and note the tone sequence you hopefully hear, it could be a failing RAM too.
 
Hi Mike, thanks for the information. I believe the sockets are what the problem is. Here's why I think it is:

Last night I reseated the Roms including the big processor, flipped it on to no change in the blinking.

So, I figured I'd go around the board and re-flow some key solder points. Particularly the sockets, and resistors. I also double and triple checked my cap soldering.
Flipped it on again, to see that it went into steady free-play blinking! Yay! I was so happy! :) I was able to hit start and have the constant lit P1 button on.
But, it didn't last. :(
Eventually it started resetting again. I wiggled a few things around, the edge connector, some caps and resistors....but when I wiggled the roms I got it to change. I lightly pressed down on one corner of the Roms furthest from the processor, I was able to get it to go back into steady free-play blinking!

This leads me to believe that the socket isn't making good contact with the Rom legs when seated. Perhaps when I re-flowed the socket points, it shifted the connection for a moment then returned to the previous state.

So I've been pondering whether or not I should just replace the old sockets with new ones, or just get the Braze Hi-score kit and render these sockets obsolete. I've read on the Braze site that the old eproms aren't needed when installing the kit.
 
get the Braze Hi-score kit and render these sockets obsolete. I've read on the Braze site that the old eproms aren't needed when installing the kit.

That is correct. I have the Braze HS kit in my AST, and my OEM EPROMs are pulled out and stored away in a box. Your CPU socket needs to be good, though.

If you're patient, I predict there will be a good supply of used HS save kits for sale when the Scott releases the Multigame kit for Asteroids...
 
Yeah good idea, I just might wait till then. For now I wanted to test my theory of the sockets, as well as eliminate another audio possibility, so I picked up a few things from a local electronics store:
- New IC sockets for the EPROMs (24 pin) and CPU (40 pin), both dual-wipe.
- New edge connector for the game PCB
- 2 new TDA 2002 audio amps for the A/R board (figured I'd get these since they didn't come with-in Bob Robert's cap-kit, and they were available right away)

I replaced the old sockets with the new ones, they came out pretty smoothly since I use a solder sucker and braid to remove the old solder keeping them in place. They just lifted right out.

Here's a picture comparison of the old (top) to the new:
index.php

As you can see, the new ones look much more effective in making contact. The chips now slide in really snug, and connections feel more solid than the old.
So, as a first try, I flipped on the power, and voila! The game boots into free-play attract mode perfectly! No more rapid flashing, just steady flashing, and after I hit the P1 start button, it lights up solid as if in game mode.
I'm pretty happy that I was able to narrow down on that problem, now onto the audio.

I replaced the old audio amps on the A/R board figuring it could help in fixing the bad speaker hum/buzzing I'm getting. But to no avail, it didn't. I still get the same audio results as before. So this leaves me to my last solution of changing out the PCB edge connector, after that I'll have to dig a lot deeper.
I'll do this step next and report back.
:cheers:
 
I'm happy to say I got the sound issue resolved. I now have a working PCB!



I did a lot of reading, researching and pouring over the schematics in the last week or so, on how the sound works and what to look for. I ended up swapping out the LM324N located at P11 with a new one, and it turned out to be exactly what the problem was. I previously read this in Richie Knucklez's KLOV thread, but at least now I have a better understanding on how the audio is mixed. I removed the old LM324N, mounted a socket and installed the new one:
index.php


In addition to this fix, I also went ahead and outfitted a new PCB edge connector, with better shielded connections.
Here's the comparison from the old to the new:
index.php


You'll notice the new one is slightly thinner, but has better contacts (looks gold plated), and fits much more snug now.
index.php

I wrapped the connections with heat-shrink-tubing to limit any interference I might get, I've read this is a good thing to do for reducing the amount of general speaker hissing. I'll clean up the wiring a bit more, so it's not so messy.

I didn't get any RAM or ROM failure tones when putting into test, so it sounds like my PCB is good to go!
Next up, the monitor repair.

Keep on rocking in the free world! :D
 
I cleaned up the monitor this weekend, installed new transistors, and re-capped both the chassis and HV board.
index.php

Got a lot of dust out of the back of this thing.

Here's a before and after of the Chassis board. After looking around the board, I didn't see any damage to the components at all:
index.php


Here's a shot of the cleaned up neck:
index.php


I didn't' take a picture of the cleaned up HV board, but here's a before picture:
index.php


After all the clean-up and part replacements, I connected it all up to give it it's first go.
Unfortunately, I got no image at all. :( Not even a dot. I turned up the brightness, to no avail. BUT, the neck was glowing strong, there was good monitor chatter, and the Chassis had it's LED lit. I've read that when the LED is lit, the spot-killer is enabled, this happens to prevent any burning onto the screen when the monitor isn't getting any signal input. So I disconnected the monitor, and started checking voltages on connection pins from the PCB.

According to the B&W monitor faq, the following voltages (AC) should be present:
1 - X Out 6.68
2 - X Ground 6.68
3 - Y Out 6.69
4 - Y Ground 6.69
5 - Z Out 6.70
6 - Z Ground 6.70
7 - 30VAC Center Tap 30.3
8 - Ground Center Tap 0
9 - Ground Center Tap 0
10 - 30VAC Center Tap 30.3
11 - 6.3VAC Center Tap 9.7
12 - Heater Ground 3.0

From my readings, I was only getting 29.8 on Pins 7 and 10... Zero on everything else. The PCB is not giving out any image signals, so I'm going to look at the components that relate to the vector generation and hopefully am able to pinpoint the problem. Any tips on what to look out for would be much appreciated! I'm off to study the schematics...
 
Long time for a bump, but here it is with an update.

I ended up getting my hands on another monitor that isn't 100% but it's closer than what the other one is has going on with it.
This monitor has good neck-glow, chatter, and the spot killer is on when powered up. If I turn the brightness level up I get some lines on the screen that look like this:
index.php


index.php


I applied a cap kit, replaced all the fuses on the deflection board, and installed new chassis-mounted transistors. Still get the same image above and can't get the spot-killer to turn off.
I haven't capped the HV section, but at this point I'm really suspecting that the signals aren't coming through properly from the board. The board I have is a known working one, I've already verified this. But what I'm thinking is that the PCB edge connector I replaced earlier may not be wired-up correctly or making good contact. I'm going to go back and verify my connections. I've read and re-read the Monitor FAQ, with not much of an explanation with what's next to look at. But at this point I'm willing to try anything.
 
I know you stated: "The board I have is a known working one, I've already verified this." but I feel you do not have a functioning PCB yet.

Can you still start a game?
Is the PCB still in free play and you can start a game?

Looks like the spot killer is on because you don't have deflection signals coming from the PCB.
 
The PCB runs fine, plays fine, all sounds accounted for. These images were taken when not in test mode. The lines move, and i can see they're trying to draw the asteroids.
I don't have a o-scope so I can't test it on that, but I got this board off a friend who told me it was 100% working.
 
Back
Top Bottom