Asteroids PCB RAM Fault but ???

dezbaz

Well-known member

Donor 4 years: 2011, 2013-2014, 2016
Joined
Feb 22, 2008
Messages
8,889
Reaction score
529
Location
Ballajura, Western Australia, Australia
Hi Guys

I have a faulty Asteroids PCB (Rev 06)
  • Random vectors when I play asteroids.
  • 1 x beep when I go to test mode

I look at the Asteroids repair manual and see this
Self_Test_Ast.jpg


The problem is the manual says D2, E4, M4, R4, N4, P4 are RAM

I am looking for the equiv to D2 as highlighted

I have RAM in these locations . . L2, M2, M4, N4, P4, R4

How do I know which on corresponds?

Could someone please help me?
Is there a manual for this revision? Or a cross reference

Cheers

:)
 
Last edited:
That's odd. Every version of the schems and manual I've got show the same RAM locations. If you have RAMs in L2 & M2, I don't find any documentation about that arrangement.

L2 & M2 were used on older ("-01 version") boards for PROMs. I assume there are some jumper wires on/around those ICs?? Someone would presumably have to changed the enable/select lines to function as the RAM...

To answer your question: The one the game (and manual) considers "D2" is which ever one is connected to the low nibble (bits 0-3) of the data bus. Specifically, pins 11-14 of the RAM should connect to either pins 2-5 on the chip at E3 (if E3 is a 74LS245), or pins 1-4 on the chip at E3 (if E3 is a AM8304B). If, instead, they (11-14 of the RAM) connect to pins 6-9 at E3 (or 5-8 if AM8304B), then that RAM is the upper nibble, i.e. equivalent to the "E2" RAM in the manual.

What concerns me, is that according to the schematic, the original wiring had both locations L2 & M2 connected to the upper nibble...
 
Last edited:
[*]1 x beep when I go to test mode

Wait a sec...
Is the 1 beep kinda high pitched, or low-pitched (more of a "boop" than a "beep")?
Is it quite short, or a full second long or so?
And, does it display the crosshatch pattern after the beep?

If it's a high, short beep, followed by the crosshatch pattern, that's not an "error beep."
IIRC from when I had a bad RAM, the "error beep" is a low, longer beep, and the crosshatch pattern does NOT appear after it.

(This is all regarding original Asteroids behavior; I have no idea about Scott's new code)
 
Last edited:
Wait a sec...
Is the 1 beep kinda high pitched, or low-pitched (more of a "boop" than a "beep")?
Is is quite short, or a full second long or so?
And, does it display the crosshatch pattern after the beep?

If it's a high, short beep, followed by the crosshatch pattern, that's not an "error beep."
IIRC from when I had a bad RAM, if it's an error the "error beep" is a low longer beep, and the crosshatch pattern does NOT appear.

(This is all regarding original Asteroids behavior; I have no idea about Scott's new code)


It's not the extreme high pitch I once heard on an atari board, it is the medium long pitch.

* * * *NOTE * * * The pitch happens in the original asteroids config, and with Scotts kit fitted

Yes I do get a crosshatch screen, but the screen's lines aren't right. I will get a picture tomorrow.
 
OK, there were apparently more versions of the Asteroids PCB than I ever knew about...
I found Manual and Schematic printings up to the 7th on arcarc. It's looking like your PCB may be a -05 or -06 version. I can't be sure, becasue the sticker on your pic covers it (it should be just after the "A034986"). I found a manual that shows your PCB layout, but haven't yet found any printing of the schematics which reference the updated RAM locations.

Perhaps just replace both of them for good measure :)

However, your problem (random vectors) sounds more typical of bad vector RAM (the other 4 RAMs) :/
 
Last edited:
That's odd. Every version of the schems and manual I've got show the same RAM locations. If you have RAMs in L2 & M2, I don't find any documentation about that arrangement.

L2 & M2 were used on older ("-01 version") boards for PROMs. I assume there are some jumper wires on/around those ICs?? Someone would presumably have to changed the enable/select lines to function as the RAM...

I'd like to retract pretty much all of the above. RAM at L2 & M2 would seem typical of -05 and -06 PCBs. It is clearly different from the -01 through -04 arrangment I was referring to.

To answer your question: The one the game (and manual) considers "D2" is which ever one is connected to the low nibble (bits 0-3) of the data bus. Specifically, pins 11-14 of the RAM should connect to either pins 2-5 on the chip at E3 (if E3 is a 74LS245), or pins 1-4 on the chip at E3 (if E3 is a AM8304B). If, instead, they (11-14 of the RAM) connect to pins 6-9 at E3 (or 5-8 if AM8304B), then that RAM is the upper nibble, i.e. equivalent to the "E2" RAM in the manual.

This part I still stand by. You should be able to determine which RAM is connected to which half of the data bus by this method (and thus determine which is "eqivalent" to D2 & which is E2).

PS- your E3 appears installed in the holes for the 74LS245 (not the ones for an AM8304B). Can't read the lettering on the IC in your pic...

What concerns me, is that according to the schematic, the original wiring had both locations L2 & M2 connected to the upper nibble...

And this I retract similar to the first, as I was referring to the -01 thru -04 layout (which you don't have).

It looks like on the 7th Printing of the schematics, Atari had updated the locations of the ROMs to reflect being in column 2, but it doesn't look like they ever redlined the locations of the RAMs to reflect being further up column 2, as they clearly are.

Keep in mind, there are numerous other things (other than an actual bad RAM IC) that could cause it to report a bad RAM... bad address decoding, bad logic chips in the RAMSEL line, bad buffers, bad bus...
 
It is normal for asteroids testmode to produce a very short 3khz beep when it
enters testmode. If any RAM failure, it will beep out a code using the lower
thrust frequencies and it will not enter testmode. That is, if you have a ram
failure you will not see a cross hatch because it does not run that code when
the ram fails. So if you are seeing a cross hatch, the RAM tests passed and
what you heard is probably just the normal quick 3kz beep. Same beep as is
played when you press any button in normal asteroids testmode. Note the
multigame testmode is different and has more features. One of them being
the buttons are mapped to play different sounds allowing you to test out
more things. Anyways, the point here is that you cant have a ram failure
and a cross hatch. -Scott
 
It is normal for asteroids testmode to produce a very short 3khz beep when it
enters testmode. If any RAM failure, it will beep out a code using the lower
thrust frequencies and it will not enter testmode. That is, if you have a ram
failure you will not see a cross hatch because it does not run that code when
the ram fails. So if you are seeing a cross hatch, the RAM tests passed and
what you heard is probably just the normal quick 3kz beep. Same beep as is
played when you press any button in normal asteroids testmode. Note the
multigame testmode is different and has more features. One of them being
the buttons are mapped to play different sounds allowing you to test out
more things. Anyways, the point here is that you cant have a ram failure
and a cross hatch. -Scott

Makes sense
I will get more info, clearer pics / video of the screen and the beep etc

It looks like ROMS and RAM are ruled out for now

Thanks guys
 
I uploaded a few videos

The Vector problem is shown about 18 seconds into the video
http://www.youtube.com/watch?v=386mwXO1rZY

http://www.youtube.com/watch?v=Sxz_VOo3zSQ

Not sure how to get the video in the site instead of a link, but it can be watched at least

Any ideas?

Somewhere in the vector generator. LOL. Like that helps. The text is drawn
with short vector draw (svec) and those seem to be working ok. But some of
your normal vector draws are not drawing correctly. I'm not sure how the experts
debug the vector generator. Thats a good chunk of the board circuitry and much
harder to debug than the CPU sections which appear to be fine.
 
Somewhere in the vector generator. LOL. Like that helps. The text is drawn
with short vector draw (svec) and those seem to be working ok. But some of
your normal vector draws are not drawing correctly. I'm not sure how the experts
debug the vector generator. Thats a good chunk of the board circuitry and much
harder to debug than the CPU sections which appear to be fine.

I am going to check for bad traces first and sockets, as this board was working 1 year ago, but went into storage. My storage situation was not the best :eek:

Thanks Scott
 
UPDATE

I have a few other Asteroids boards so I tried on in the cab. One of them is a REV 04 and was fixed by a repairer in USA a few years ago. I forgot I had it. (Or rather, I thought the one in the cab was that one)

Here's what happened - exactly the same thing

Random horizontal vectors exactly like the prev PCB
http://www.youtube.com/watch?v=41BtJnraI3o


I then thought what's the chance of that??

So I fitted the Lunar Lander PCB I have into the cabinet and this is what I got
http://www.youtube.com/watch?v=uT0oYyJXWqo

It looks like all 3 of my PCBs are OK, right?
 
Further developments

I swapped the entire chassis, with the Lunar Lander chassis - Complete

Same problem

I swapped the last part of the harness, as some of the braiding was a bit sus - same problem

I am now thinking the Yoke or the Power Brick are the problem
 
I think it's big blue

From here
http://arcadecontrols.com/BBBB/bigblue.html

* 1 Intermittent lightning streaks in pic/hum bars/weaving.
* 2 Blank raster... steady or intermittent.
* 3 Jumpy pic sometimes described as jiggles.
* 4 Failure of H hold to lock in.
* 5 Intermittent vertical roll or failure to lock in.


Mine is point 1

I have a spare BB, will fit tomorrow and end this
 
dez

could you post a picture of the rev 04 board?
like the one you posted for the rev 06 board.

Thanks
 
Could you put the Asteroids board in the Lunar Lander cabinet?
Maybe put the two games back to back and swap monitors, I mean just the wires between the cabinets if their the same. You should be able to swap things around to isolate the problem. If it's not Big Blue then it must be a monitor problem if all the boards do the same thing.
 
Could you put the Asteroids board in the Lunar Lander cabinet?
Maybe put the two games back to back and swap monitors, I mean just the wires between the cabinets if their the same. You should be able to swap things around to isolate the problem. If it's not Big Blue then it must be a monitor problem if all the boards do the same thing.

Yes, I can, but I have borrowed the chassis and harness so LL is not a straight swap at the moment

I have the Big blue to fit, so I will do that anyway tonight (in about 12 hrs)
 
Back
Top Bottom