Please help - Stargate (Williams) power supply

[vinito]

Hi Folks.
First let me thank in advance for any help here.
I read a couple posts regarding the repair of a Williams Stargate game and they are somewhat helpful, but the OPs had different problems than I do so I was hoping somebody could help lead me through my specific problem.
The game I've inherited has a power supply problem. First, here's a picture of the same thing I have:


On powerup, LEDs #2 and #3 light up, but #1 does not. This indicates a problem with the +5V circuit, and checking voltages with multimeter verifies this. Other voltages are within range (didn't check ripple yet) but +5V measures down around 1.3V. BTW, I checked the voltages with output headers unplugged, so it was under no load.
Also, it may be obvious with so low a voltage, but of course absolutely nothing is happening on the MPU, i.e. I get no reading at all from the 7-seg. display.

I hobby with electronics enough to be dangerous I suppose - I have a pretty nice Hakko soldering/desoldering station and do pretty well with it. But with my limited knowledge, I'm not sure what I should replace "shotgun style" to see if I can revive the 5V circuit. I plan to replace the header pins and sockets (I already have the stuff to do that from my days doing pinball repairs). I have verified that the connections are not causing my 5V problem on this board though, so there is more to it.

I figger resistors and ceramic caps are probably OK just because they don't usually fail, and there is no darkened (burned) looking areas on the board. What should I replace for the 5V circuit that could be a likely problem? Electrolytic caps are suspect of course, but which ones? I guess it wouldn't hurt to replace them all, but I'd have to order and wait for a kit if it's available somewhere. If it's nothing too strange, it's very possible that I have what I need to get the 5V section working again so I can fire the machine up and see if that is the only problem I need to address.

Alternatively, I suppose I could separate the header and supply the 5V from an ATX power supply or something. But if I do that I'm not sure which ground lines I should segregate from the original board and pair up with the +5V lines since I wouldn't have a common ground plane anymore if I did that.

So if anyone could offer advice for either option I would appreciate the counsel.

Thanks.
Vinito

p.s. I'm not sure what the separate heat sink board does yet, but when I did my checking I had it disconnected from my board too. Not sure if that would affect the 5V or not - I would suspect that board to apply to a higher voltage circuit and not the +5V. Was I wrong there?

[smartbomb2084]

 The heatsink assembly has to be connected to the board in order to have +5V.

[vinito]

Thanks.
That connector had a couple browned slots so I replaced the header and crimped new trifurcon pins to rebuild the socket. I made double-sure that the pinout was the same as original (haven't done this in a while and was paranoid I'd screw things up so I was extra careful). That all turned out pretty nice and solid.

Then I plugged it in and powered it up. Well the connector needed to be replaced anyway so it wasn't wasted effort, but still getting no love from LED #1 and still measures a tad more than 1V.

I have the book with schematics, but it only includes the single-board power supply types (8373 and 8359 right?). Any links for schematics and assembly drawing for my actual board out there? That would have to help.

While I was in there I measured BR1 since according to my iffy schematic, that's the one used for the 5V supply. Getting around 30V dc from the output pins and virtually no AC ripple. That seems to be working fine anyway IF that voltage is within the correct range.

What do I check or replace next?

[SavannahLion]

http://www.coinop.org/repair/PS_Williams.aspx

Mentions the two board power supplies and links to a bunch of schematics. Unfortunately, you have to log in to see most of the schematics. You might be able to find it by seeing if you can find the schematic as some place like Crazy Kong Arcade.

Remember, it's likely that Williams used the same power supply in more than one cab. So even if you can't find what you seek by looking at the Stargate schematics, you might find it for another game. Just be cautious of any changes between cabs.

[vinito]

Yea I found that site the other day but registration is either buggy or dead and I have no word back from them so I can't view the documents.
However, I did find that the Robotron power supply seems to be the same thing and I found the drawings for that. I'll be looking it over to see if I can make any sense of it. So far what I've seen looks to be identical.

Here's the document I'm looking at:

http://www.robotron-2084.co.uk/manuals/robotron/robotron_upright_drawing_set_apr_82.pdf

[vinito]

I have a shopping cart going at Mouser for a bunch of parts. So far I've got all the electrolytic caps (except the 18000mfd), IC1 & 2 voltage regulator chips, Q1 & 2 scr, Q3 & 4 transistors, and Q5 7905 regulator.

I have to look around as I may have an 1800mfd cap hidden somewhere - seems to "ring a bell" for some reason.
I would like to go ahead and replace the bridge rectifiers while I'm at all this but having trouble picking out the correct ones at the Mouser site.
Obviously some of the parts don't have anything to do with the +5V section, but I'm thinking (maybe wrongly) that it shouldn't hurt to replace anyway.

So any advice here is welcome. Maybe there's a rebuild kit which has all this available somewhere from an arcade machine supplier?

Thanks for the help so far.

edit to add:
I found Bob Roberts's site and they have a rebuld kit. (link - about 2/3 down the page) Not sure if it has everything I've put on the Mouser list but the little picture looks like it has most of it I guess. I'll try to contact and see if I can get a lits of what is actually included in the kit.

[smartbomb2084]

Did you check th 2N6057 transistor on the heatsink yet?

[vinito]

No - how do I go about doing that?

Edit:
OK I looked it up. Wouldn't hurt to verify this is correct.
In case it matters, my multimeter isn't a Fluke, but it's a decent old Radio Shack one. Sometimes it acts a little buggy, but usually it does pretty well.

For testing purposes, this is supposed to be the correct conceptual schematic (NPN on left):



My method according to the interwebs:

Switch multimeter to "diode" mode.
+ on E and - on C should not pass current. (fail : current passes)
- on E and + on C should not pass current.
+ on B and - on C should pass current.
+ on B and - on E should pass current.
- to B and + to C should not pass current.
- to B and + to E should not pass current. (fail : current passes)

Looks like I have a bad one, yes?

Also on the webbernet, I discovered that NTE247 crosses as a replacement for the now-pretty-much-obsolete 2N6057. Does that sound correct?
(NTE247 datasheet)
(2N6057 datasheet)

[smartbomb2084]

Was the 2N6057 transisitor tested with the heatsink connected to the main power PCB?

If it was, test it again disconnected and see if you get the same results.

An NTE 247 should be an ok substitute but sometimes you just have to try the NTE part in the circuit to see if it will hold up.

[vinito]

No. It wasn't connected. It was still mounted to the heat sink of course, but the connector was just dangling.

[smartbomb2084]

Then I would say the transistor is junk.

Whenever you are testing a Bi-Polar transistor and 'current passes' between the Collector and Emitter, regardless of the polarity, the transistor is junk.

Anyone else around here disagree with that statement?

[vinito]

I just stuffed an envelope for Bob Roberts to order a rebuild kit from them. Hopefully everything I need is in there. I know the picture they have shows a 2N6057 in the bag so we'll see.
I'll replace everything they send me and see how it works. I'm guessing it will be a week or two, so unless I stumble across a 2N6057 while cleaning my house (HA! Not sure what's less likely - cleaning my house or stumbling across that piece) then I'll probably be on hold until then.

Thanks for the help. Looks like I found one bad suspect part at least, and I seriously doubt I would have checked it out if you hadn't mentioned it.

edit to add:
Got an email from Bob Roberts today saying that to look for the parts around Thursday! Well that will work out great even if it takes a day or two more.
In the meantime, if there are any typical problems with the Stargate game I should check out, feel free to let me know and I'll look into it as much as I can without powering it up.
I figure I'll look all the headers over and for sure replace any which look like they got warm.

I was able to push the test button on the sound board early-on and it seems to be cycling through just fine. Might be a bit crackly if I remember correctly, so maybe I can clean that up.
I can hear that familiar high pitched whine from the CRT circuitry and when powering it off it does a kind of squishing flash, so at least it's not a black screen. Maybe on boot-up it will just work.

[vinito]

OK. Got some progress.
Received the parts from Bob Roberts today and replaced all the parts that came with it. Hooked it up and switched on power.
All three LEDs are on. Excellent <wringing hands>

Hooked up the other headers and powered back up.

It came up with a 1-3-4 RAM error code. Screen shows the carpet thing, then a RAM error message, then reboot, plays a snippet of sound, etc. Switching around ram chips nets not much, but I did get it to change to a 1-3-7 error.
So I decided to look closer. A cursory inspection the other day didn't look like there was any board corrosion, which surprised me since the batteries were still in and corroding some. Looking closer today I can see that various legs of the RAM chips have flecks of the greenish corrosion on them, some are kind of coated in a black film and others look like they are flaking just a bit. Also, there is a kind of dusting of corrosion-colored film on the board, but only on the lower area near the battery and extending through the RAM socket array. The rest of the board looks pretty clean actually.

So I was thinking I could probably strip all the RAM sockets out since they are probably iffy at this point and replace them with new ones, then as per a bit of interwig research I could replace the 4116 ram chips with 4164 and alter the power header socket a bit to suit (replacing the 12V input with a 5V via jumper wire).

So does that sound reasonable or should I maybe step back and try some simpler things first? I'm into simple if it will do the trick.
I guess the question that keeps entering my mind is whether something else that's messed up would cause a RAM error code to come up. Don't want to waste $30 if I don't have to (for $1 per I figured I'd just get 30 in case I bugger something up and need spares).

Whew! I knew this wouldn't be as simple as just a power supply issue, but at least it's doing something now.

BTW - I checked the voltage at the MPU board and the 5V supply indicates just shy of 5V at 4.98, so I think the power is getting to the board fine. Other voltages seem OK too.

[smartbomb2084]

Why not just try cleaning the RAM chips first?

You do know that a RAM ERROR of 1-3-7 means.....

1--- RAM ERROR

3--- RAM BANK 3

7--- RAM CHIP 7

Just checking...

[Mysterioii]

You can clean general filth off of there pretty easily with just regular old rubbing alcohol and a plastic bristled brush like an old toothbrush...  I guess maybe there are special ESD-safe brushes out there but I wouldn't really worry that much, personally.  Just let it dry really well before powering it back up.

[vinito]

I considered cleaning things, but it's nearly impossible to clean chip socket holes and that's proably the most likely weak point wouldn't you think?
Plus i don't know if the ram chips are good or bad to begin with so thought another shotgun approach might save time and possibly improve reliability using the updated ram.
I have switched entire banks around and still get the same error, for what that's worth.
Well I'll stare at it a bit more tonight and see if I have any revelations.

[Mysterioii]

I dunno man, I don't usually like the shotgun approach, I was all set to suggest that you not replace all the parts you got from that one guy at the same time but before I could suggest it I saw that you'd already done it.     I tend to like to replace things one at a time, I like the satisfaction of knowing what's wrong (and not wasting good spare parts).

If you swapped the banks around and got the EXACT same error, then based on what smartbomb said I would guess that it's not the ram (I would have expected the error to persist but to show up as a different bank).  Maybe there's a bad solder joint or lifted trace or something somewhere.  Me personally, I'd clean things off first.  You might see some chip damage or a lifted trace or something better without a layer of gunk there.

I hate trying to debug things without a schematic though.   

[vinito]

If I was better at electronics, I'd agree with you. I normally like understanding what's wrong and just fixing it. Also, I've had the shotgun backfire before too by making a little error in the process of replacing dozens of parts then taking quite a while to find where I went wrong.

In this case the reasons I'm not hesitating too much is that 1) I've read that the 4164 RAM is more stable and uses less power so supposedly will cause less grief in the future, and 2) the bigger (potential) problem is that corrosion in the chip sockets seems impossible to try to get reliably clean other then replacing them.

I wish I knew more about troubleshooting so I could just dig in and find the problem. I have nice soldering/desoldering equipment and even have an o-scope and function generator and a couple decent bench power supplies. now if i just knew more about how to use the stuff.

Anyway, I'll look at it again over the weekend and maybe something will come to mind. I guess I could just replace the RAM sockets and clean the chips and see if that helps any. At least that would allow me to clean the board well and inspect traces and such to make sure there's not any hidden issues. Maybe I'll get a different error code just for some variety.

and BTW, I have dug up some appropriate documents and found how to interpret the error codes. That's why I tried swapping a whole bank of chips with another - to see if I could get a 1-2-? error code or something which would point to a bad chip.

At this point I wish I could just get the board repaired by someone who knows what they're doing. I like learning things and all, but I'm quickly reaching my limits in ability.

the saga continues for now.

edit to add:
I do have a set of schematics I downloaded. They aren't as clear as a printed one, but so far I've been able to work out what the little print says with some deduction. It has been helpful looking up a couple things so far. Too bad I can't make better sense out of how it all works together.

[smartbomb2084]

Anyone brave enough to replace 24 RAM sockets on an 80's Williams Video CPU board is a better man than I.

The solder is super hard and loaded with impurities that will gum up your solder sucker fast.

The biggest problem I see with these boardsets are not in the chip sockets as much as the ribbon cable connecting the ROM and CPU boards.

The 4164 RAM to me is a luxury type fix unless you plan to operate this game on the route where it will be subjected 12+ hours of run time per day and people beating it to a pulp.

When you have a 1-3-7 RAM ERROR swap out only that ONE chip (37) with any other ONE chip and see if the problem follows that suspected bad chip.

[TOK]

Anyone brave enough to replace 24 RAM sockets on an 80's Williams Video CPU board is a better man than I.

The solder is super hard and loaded with impurities that will gum up your solder sucker fast.

The biggest problem I see with these boardsets are not in the chip sockets as much as the ribbon cable connecting the ROM and CPU boards.

The 4164 RAM to me is a luxury type fix unless you plan to operate this game on the route where it will be subjected 12+ hours of run time per day and people beating it to a pulp.

When you have a 1-3-7 RAM ERROR swap out only that ONE chip (37) with any other ONE chip and see if the problem follows that suspected bad chip.

I just did the 4164 swap on my Defender, and have to disagree with you. All the new RAM cost about $25, and it's likely I'll never have problems with it again. If you put any value on your time, this is a cheap and effective fix for a common problem.

You don't even have to mod the board or RAM itself if you get/make an adapter to eliminate the 12 and double up the 5v on the correct legs.

[CheffoJeffo]

I just did the 4164 swap on my Defender, and have to disagree with you. All the new RAM cost about $25, and it's likely I'll never have problems with it again. If you put any value on your time, this is a cheap and effective fix for a common problem.

 

There is a reason why even the cheapest hobbyists make this change, and it ain't because they are putting their games out on route.

[vinito]

I figured I'd try a bit of "due diligence" before I start ordering parts.

I went ahead and replaced the .156 header pins and built new plugs for the pigtails with good trifurcon crimp connectors. I didn't like the old insulation displacement plugs anyway.
I'd kind of like to replace the ribbon cables too, but I don't have those so maybe I'll go ahead and throw those in with my order just to make sure they are good too.

So plugging it all back in and powering up resulted in the same 1-3-7 error. So rather than swapping whole banks I tried just swapping the 1-3-7 RAM chip with another one which looked clean. This resulted in a different error number, though not the one I swapped out. So I swapped the new error numbered chip out with another different clean-looking one and got yet another error number, though again, not the one i swapped out. Did this a couple more times with similar results. Currently I'm getting a 1-1-3 error number.

So in my ignorant mind, this tells me that I probably have some problem RAM chips at least and if I replaced them them I might get some better result maybe.
So I'm planning to go ahead and get some. Since they are about the same price and the "hack" is so simple, and since folks are recommending the update, I'll go ahead and get some 4164 RAM and do the power hack.
If replacement ribbon cables are available, I'll go ahead and get those too. Question - the other end of the ribbon cables are directly connected to the board right? It looks like that is via insulation displacement connector of some kind, but it's not a plug I guess. If I'm wrong let me know.

So since I'm replacing the RAM, I am still planning on replacing all the RAM sockets too since they are in the "corrosion zone" and their internal condition is unknown. Replacing them would make their condition a known good thing again (yea I know, but the odds are pretty high). I have a good Hakko desoldering gun, so though it will be a little time-consuming, it should go OK. After I remove the sockets and probably a handful of other components with a bit of corrosion showing, I'll clean the board well and clean the legs/wires of the odd components and tidy it all up and file it under the Corrosion Repair category. I guess I ought to kludge together some kind of remote battery pack so it will retain settings too.

So that's the plan. If there are suggestions, corrections or advice, don't hesitate to let me know. I'll thoughtfully consider anybody's input whether I ultimately use it or not.

Thanks.

[smartbomb2084]

I guess I should consider myself lucky to have never had RAM ERRORs so bad that I need to modify the system.

With DEFENDER it seems like ROM ERRORs are way more prevalent-- at least in the boards I have.

I am personally against system changes because at what point does that affect the originality of the game?

Will the game still skip and sputter at the start of Wave 20 with extra RAM or will it now transistion smoothly?

If it does transition smoothly, you could say that the game is 'better' now but it is no longer original.

Just like if you run a switching power supply in these games your High Score Table gets periodically erased but if you keep the original there's no trouble found----- anyone got a fix for that?

[TOK]

A legit concern... I don't think the Williams hardware is able to address the extra capacity (4x the capacity, so it is substantial!).
With the 4164, it does all the same "tricks" of warping enemies away when there are too many on screen and slowing down when things get busy. One of the things I'll pay attention to is Landers warping through the top of the screen without mutating when there is a ton of stuff onscreen.

I am running a rebuilt linear supply for the reason you mentioned.

PS: The bad RAM also tripped a ROM error, and I couldn't some of the changes in book-keeping. That went away after the upgrade.

[vinito]

PS: The bad RAM also tripped a ROM error, and I couldn't some of the changes in book-keeping. That went away after the upgrade.

That reminds me. When I was swapping individual chips and getting a variety of errors, I occasionally got a ROM error a couple times. That concerned me, but after reading your post I'll hold on to hope that the bad RAM was mistakenly tripping that ROM error.

I ordered a bunch of stuff a couple days ago. Got the RAM and replacement ribbon cables from Bob Roberts coming (supposed to be here tomorrow). I had only 20 sockets so I ordered more of those, plus some various capacitors and diodes and such to spruce up the iffy-looking ones in my corrosion zone. I figure it will take an hour or two to rebuild the board, but time flies when I'm doing that so it should be pretty painless as long as I don't run into unforeseen issues. At least the board will be all cleaned up and the RAM all upgraded in the end.

IF this reanimates the game, I'll have a total of about $70 in parts. That won't be too bad. I just hope this takes care of it. It appears that the video is coming through OK and the sound test verifies that system is working too, so I might just be OK.

[vinito]

I received the RAM and cables from Bob Roberts and another package from Mouser for various other components today.
All I need is the third package with the new sockets and I'm ready to put this thing back together.
Man I sure hope this works. Kind of just want to get past it.

Once it's working, I'm donating it to my Mom's basement so it can be played by company and at family gatherings. It will be cool sitting next to this second donation, which I am also in the process of rebuilding. But it's all relay logic so much simpler (for my concrete brain) to work with:

[vinito]

I have dug into the socket removal. I have about half of them off and my suspicions about the internal condition proved to be conservative if anything.
On the first socket I used the desoldering gun, but it almost pulled one pad off and while there was a dusty film of corrosion on the board where things were in the open, under the socket it was built up pretty thick and green, and pulling the clips out of the case revealed that they were in pretty rough shape. They were at least 50% covered in fairly thick, green corrosion.

On the first socket I discovered that simply pulling on the plastic case would pull the clips out pretty easily, so after that I tried simply prying the case off first. That works pretty well !! So I just started prying off the cases and leaving the clips/legs soldered to the board and sticking up, which allowed for some simple loosening of the solder with the soldering iron and gently slipping each leg out individually. I'm making pretty good progress with this method, and it allows me to use a gentle touch to insure that I don't lift any traces off. From the second socket on it has been going smoothly and the board has been completely undamaged. I just have to make sure to remember to solder the first socket in place and verify continuity in case my pad flub messed anything up. The rest of them ought to just be standard soldering practice.

Fortunately only about the bottom three rows or so are corroded and it's getting "cleaner" as I go up. I'll clean it up with some vinegar, then some alcohol and water and let it dry well before I get to the reassembly. I think it should clean up OK.

Wish me luck. And again, any suggestions are welcome.

[vinito]

OK Kids, I'm gettin' close

All the new sockets are in. I ran two little bridge wires through and connected them to their respective traces to make sure my pulled pad flub didn't cause a continuity problem. Under magnification it looked worse than I originally thought and the bridges turned out good, so I feel pretty good about it. The SN74LS04N chip near the crustal had corroded legs for whatever reason and the capacitor that goes to it (the one that kind of looks like a hack) was disconnected, so I replaced both of those components too. A couple diodes (D9 and D10) looked iffy and some of the .1µF caps next to the RAM sockets were rough-looking too, so I bought replacements and just need to solder them in.

Just a few more components, the power plug adapter and a remote battery pack and I'm ready to give this thing a try.

Any advice here is definitely welcome, especially cautionary stuff. I don't want to fry something at this stage for sure.
I'll be pretty busy most of today so I probably won't be working on the board until later this evening.

[vinito]

Well here's the rebuild.

First a before pic of the RAM area:


The diodes before. Just a bad looking solder joint mostly:


And the chip & capacitor I was talking about. I noticed the capacitor had come loose before so soldered it back on, thus the resin showing.
There's some green on there if you look close enough:


And here's the big finish. RAM isn't inserted yet, but showing the sockets is more informative at this point anyway. I made the remote battery pack so it can be unplugged via connector if the board needs to come out later.


The ribbon cables for the interconnect and ROM boards have been replaced. Just have to make a power plug adapter to suit the new RAM and I'll be able to check this thing out.
Lesseee... Turn around three times, throw some salt over my right shoulder, hold your mouth just right and power it up?
We'll see I guess.

[vinito]

Excelsior !!!


I inserted the RAM, checked everything over a bit and hooked it up.
[drum roll.........]

Oh man... Two RAM errors again... but wait.. it went to the main screen on the monitor... Then reset. Huh.
Then on second boot got a different RAM error. OK hit reset then. New error.
Reset again. "All systems GO". Well OK then. Put a couple credits in and start....

Hey ! It works! "zero" on the LED too. Boot up a couple more times. Power off & on a couple more times. It keeps working. Woohoo!

I guess it needed a couple warm-up boots or something. It seems to be working just fine. Screen looks good too.
Alrighty then!

Just a couple glitches to work out. Can't seem to get into setup mode. Probably a connection somewhere I would think. I'll trace it down later and see if I can find the problem.
For now it seems to play unlimited number of lives and smart bombs. Is that right for a first boot-up?

Well I'm pretty happy with the success I've got so far. I forgot how difficult the game is to play. I haven't played it since it was first released  

Thanks for the help and suggestions. Any further suggestions regarding my current wrinkles would be appreciated.

 

[TOK]

I think booting it with the batteries removed will force it into setup/book-keeping. I only have a Defender, but think its similar enough to share some of the same issues.

Sounds like you're on the right track! 

[vinito]

(I started a new thread on this different problem: http://forum.arcadecontrols.com/index.php?topic=119913.0)
Well I guess as I get closer to getting it working, things are getting more elusive.

If I leave it off for a while, when I turn it on I get a ram error. It seems to be a different number every time. I've got 1-1-1, 1-1-7, 1-2-6, 1-2-4, 1-3-4, 1-3-6 just since I've been keeping track.
If I don't leave it off for long, then turn it off and back on (or push the reset button) then every once in a while it will come up with a different RAM error, but most of the time it will just go to zero and boot up with no RAM errors.
There's a door plunger switch near the cash box. It has been "hacked" to stay closed. If I leave it that way, it just boots up (when it passes the RAM test) and skips the "Adjustment Failure" message and goes to the game over screen, but then I can't access the setup mode. With the switch open, it boots up to the "Adjustment Failure" "Door Open" etc. message. But leaving the coin door open and turning off & back on does nothing different, i.e. "factory resets" are not restored, but rather the same "Adjustment Failure" message comes up every time. However, I can get into the setup/audit mode kind of. It gives the basic audits, then pushing the advance button again goes to a screen which has the text (and only the text) "EXTRA SHIP EVERY" along the top left hand, but a kind of blippy block of short lines of varying colors kind of flashes from left to right and slowly scrolls up the screen. So strange video there - a kind of partial game adjustments screen but then that strange flashing/scrolling line thing.

The odd thing to me is, if you let it boot up, the game will play and the video looks pretty good. All the controls work and you can play through the levels, but it starts with a ton of lives and smart bombs - way more than 10.

So part of it works and part of it doesn't. The RAM errors vary greatly and go away if you hit reset once or twice, and the Adjustment Failure message is consistent and won't correct to factory settings as long as the cash box switch is open. Then there is the strange video once you enter into setup mode.  If the cash box switch is closed, it will boot up and let you play a game with superman settings.

Anybody have any ideas?

Help is welcome. Please!

And by the way, booting with batteries connected or not seems to make zero difference.

[pinwizz69]

Try one of these for trouble shooting and you'll understand why I rave about them.
One of the best possible tools you can buy for the $ is one called  " The Lil' bitty Tester " .
They cost around $25. I,m a 30 year Pro Arcade game tech. and this tester is the best tester I,v ever bought.
It does active tests on Diodes, Transitors, Mosfets ( Makes the gate latch ", Bridge Rectifiers and more.
Mine is an easy 20 years old and has never failed.
I still keep it in the original ( Lil' bit of tape these days ) box and have the manual.
Waaaaay better than even my Pro series Fluke meter and easier than a scope.
I wish they could test caps but the best way to do that is with an Analog scope.
If you can't buy one locally look at MCM Electronics.
I,m sure there,s other sellers, I just like there AAA+ service they have provided me over 25 years as a Service Manager with 2 distributors and several Routes.

[boardjunkie]

From the look of the original ram sockets,  you better check the rest of them. Inspecting/cleaning/replacing sokets  is *always* the first thing I do when t-shooting problems on boards this old.

Always clean the chip pins too. I use a Dremel with a wire wheel on it and it is fast and effective. You need to be careful doing it this way so it doesn't catch and launch the IC across the room. Bottom line....if the chip pins are not clean and shiny, chances are they ain't makin' good contact with the socket.

Double check your solder on the ram sockets. You didn't damage any thru plated "vias" when removing the old ones did you? Its easy to do if yer new to this.

And...with the 4164 retrofit, its a good idea to short the unused pins (where the 12v and -5v would appear for the old ram) to either gnd or 5v. Ram usually doesn't like unused pins just left floating. Pin 1 is unused on the 4164, so you'd just need to  jump A7 to the "new" 5v pin (short pins 8,9). Just do it on 1 chip and the connection will be common to the rest of them.

[RobMcRaf]

I too am having similar problems.  I have gone back to a linear power supply (the cab was using a switching one that supposedly was bad for the CMOS) and have switched to 4164 ram and made the appropriate mods.  With the linear power supply on and all the boards connected, I am having getting less than 4 volts at the pins.  I cant find the schematics for the 2-board type of power supply to do proper trouble shooting.  If anyone can help me out it would be appreciated.

I have 5 volts when the boards are not connected.  I havent done the plug-1-at-a-time test yet,  will do that asap, but really need schematics.  Thanks!