Capcom 1942 - Advice sought

DarrenF

Well-known member
Joined
Jan 7, 2011
Messages
5,855
Reaction score
526
Location
Florida
Working on a dead 1942.

Lots of nice tile/character/sprite garbage on screen when powered on. This tells me the video section is more-or-less working. However, there's no movement or change, it's not "booting".

Inspected for physical damage; none found.
Voltage is good (5.0-5.1) all over both boards.
Pulled & verified all of the non-gfx EPROMs; they're all good.
Checked continuity of all lines in the interconnect cables; good.

Probing both Z80s: the reset lines are staying high; the clocks are pulsing; there is activity on all data lines and most address lines. Only thing curious to me is that the INT (not NMI) line is staying low.

I also noted that the RAMs (TMM2016s) are hot. Measured them at ~150°F. It's hot enough that if I touch them wile probing around, I pull away by reflex. It's uncomfortable to touch for more than a couple seconds, but it's not going to give you burns or anything.

Anyhow, it's a real pain for me to diagnose, as I can't find any schematics for 1942. The best I can do is look at schematics for similar era Capcom games and hope they did things similarly. So I've been looking at the GnG and Commando schems for clues. I suppose it's better then nothing, but it still sucks.

So, anyone have any sage advice? Does that RAM sound likely bad? Or would you try chasing down whatever's pulling the INT line low? Anyone ever run across the schems for 1942? Anyone drawn their own by tracing the circuitry? (Cause I'm about to try...)
 
My repair logs are also at www.retrocomputermuseum.co.uk in the arcade section which you can view without registration. I don't bother putting them on KLOV as there is a very low word limit per post and I think you are only allowed 3 photos or something equally daft.

Anyway - your RAM chips are bad, they should be slightly warm to the touch and TMM2015/2016/2018s are known to turn into heaters when they go bad, expecially if they are the black shiny chips, the matt finished TMMs were later chips and seem to be fairly reliable. The symptoms you describe are exactly what you would expect to see from a board with bad system RAM too, so if you replace the hot ones it is likely your board will work perfectly again. If you dont have a desolder station then use side cutters to snip the legs off the chip and remove them one by one. You may need to apply a lot of heat to get the 5V and grounded pins out, if this is the case and you are still struggling to get them out then use a hair-drier to get the surrounding board good and hot then re-try.

In terms of what to replace them with, they are more commonly known as 6116 SRAMs, as long as you make sure you dont replace them with slower chips you will be fine. Also bear in mind they came in two physical sizes, 0.6" and 0.3" wide if you are going to order some sight unseen, they are very common chips so should be cheap.
 
Last edited:
Thanks guys. You were dead-on, Womble.

I stopped being so lazy, and pulled the scope out. The signals on the data lines were very ugly...

Both TMM2016s were the shiny type. I replaced one (the one that had an active enable line) with the only 6116 I had. It then booted up to show the title screen before hanging at the start of attract--success. I dug around a little more, and found I did have a tube of 2016s... 200ns, but 2016s. Anyhow, put it in place of the other RAM, and now it boots up and runs attract normally.

Now I need to add a speaker to my test connector and see about the audio...

Thanks again, Womble!
 
A final follow up:

No sound when I hooked up a speaker.

The TMM2014 hooked up to the audio CPU was also really hot, so I replaced it. No change.
I then looked at the data lines there and they looked pretty wacky, so I swapped in another Z80. No change. I traced the data signals up to an LS245 that didn't seem to be acting right; replaced it, no change. Then I stopped being stupid, and went to the other end of the circuit...

There was output from the PSGs, and even from the amp! Replaced the output-coupling 1000uF cap and now I have sound.

Lesson learned: When you don't hear anything, start at the amp and work backwards.
 
Last edited:
Lesson learned: When you don't hear anything, start at the amp and work backwards.

Which is why I love logic probes with audio sounds! Put it on the outputs of the audio chips (Yamaha FM chips, etc...) and listen. :)
 
Which is why I love logic probes with audio sounds! Put it on the outputs of the audio chips (Yamaha FM chips, etc...) and listen. :)

I use a crystal earpiece to trace through certain types of sound fault, ie to figure out if there is any recognisable sound in the sound circuit (much better than looking at a 'scope as at least you can then hear the sounds). Like this:

http://www.ebay.co.uk/itm/HIGH-IMPE...ecialistRadioEquipment_SM&hash=item43a89b8ca3

Chop off the jack plug, attach one wire to ground, attach the other wire to a small cap (0.1uf ceramic disc or similar) and use the other end of that to probe the relevant component legs. Works a treat. :)

A logic probe with audio would be more useful though - what make/model do you have? And can the audio be turned off?
 
I'm pretty sure Capcom forbade the release of the 1942 schematics to make it harder for people to bootleg the boards. Fortunately for us someone must have, against the rules, photocopied a copy (probably from capcom usa) to help an operator fix a bad board, and the operator or capcom after adding notes and made a second copy. I snagged this copy on ebay on sundayish, and it arrived yesterday and I finished scanning it a few hours ago. The papers were 8.5x11 double sided copies of the larger probably C or D sized original sheets. The papers I have show signs of being run through an HP or similar auto document feeder scanner with bad rollers (you can see the two 'burn' marks by the bad rubber scraping on some pages) and the marks are on THIS generation copy's paper only, so someone else likely scanned this same set of sheets at some point but never released them. The mixed up sheet order as I received them and restaple holes add to that suspicion.

http://www.sendspace.com/file/hpstiu

PDF version pending.

LN
 
I'm pretty sure Capcom forbade the release of the 1942 schematics to make it harder for people to bootleg the boards. Fortunately for us someone must have, against the rules, photocopied a copy (probably from capcom usa) to help an operator fix a bad board, and the operator or capcom after adding notes and made a second copy. I snagged this copy on ebay on sundayish, and it arrived yesterday and I finished scanning it a few hours ago. The papers were 8.5x11 double sided copies of the larger probably C or D sized original sheets. The papers I have show signs of being run through an HP or similar auto document feeder scanner with bad rollers (you can see the two 'burn' marks by the bad rubber scraping on some pages) and the marks are on THIS generation copy's paper only, so someone else likely scanned this same set of sheets at some point but never released them. The mixed up sheet order as I received them and restaple holes add to that suspicion.

http://www.sendspace.com/file/hpstiu

PDF version pending.

LN

Wow. Thanks for snagging them, scanning, and sharing! And welcome to KLOV.
 
Grabbed a copy. Huge! But I'm very appreciative. While my board is working fine, having the schematics is a security blanket so I'm very happy to have it for whenever the day comes that I need it. I'm presuming the pdf version will be smaller so I'll be watching for that to show up. And I would suggest submitting it to arcarc.xmission.com (email address is right there on the home page), or I'd be happy to if you don't want to deal with that and you don't mind me doing it. I submitted a couple of manuals recently and they were really quick to post them to the site.

Thank you for this nice contribution!!!
 
PDF version pending.

LN

I managed to use ImageMagick to convert them to pdf ("convert *.png -compress jpeg outfile.pdf") Then I rotated them all clockwise using pdftk ("pdftk outfile.pdf cat 1-endR output rotated.pdf") The resulting PDF has all 8 sheets in the proper order, rotated landscape, and is ~19MB.

I'm sure someone skilled in the art of image manipulation and pdf creation could do much better. Perhaps I should convert to jpeg and trade a little quality for size before PDFing...

EDIT: I'll be damned. Looks like someone already did a better job... there's a 5.7MB version on arcarc already: http://arcarc.xmission.com/PDF_Arcade_Manuals_and_Schematics/1942_Schematics_(Capcom).pdf
Nice work.
 
Last edited:
Back
Top Bottom