Asteroids Cocktail restoration

[csa3d]

So today I got the last of the caps in the HV board, plugged everybody in and flipped it on to see what would happen.  I smelled some smoke and hear some arcing in the HV cage, I haven't opened it back up to see what went wrong yet.  I was hoping everything would magically light up, but that didn't happen.

So turning on self test, I unplugged the monitor and figured mind as well see if she plays bind.  With the self test, you're supposed to hear a series of beeps to tell you what's going on.  High means good, low beep means bad.  There should be multiple high beeps.  Luck not being on my side, I hear only 1 beep, and that beep is LOW.  This means that the first RAM chip fails self test. 

If I have to buy 1 ram chip, I'm assuming I mind as well replace them all.  If I only buy one, what's to say any of the others are good.  I'll have to order more anyways.  So this being said, can anyone steer me towards the proper place to get these ram chips?  The manual says I need "Item #152, part # 90-7033, Random Access Memory  (D2, E2, M4, N4, P4, R4)"

I also don't really where to go from here.  I'm assuming I should discharge the monitor, investigate what's smelling/burning/arcing on the HV board.  Any suggestions would be great.

Thanks
-csa

[csa3d]

Today I decided to to take some readings and photos, and hopefully generate some discussion on what to do next.

Here are my multimeter readings off my power block and AR-I, based on this KLOV thread about Asteroids Power readings:

=====================================
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)

reading: 14.15 (high)


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

reading: 37.5 (high)

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

reading: 7.1 (high)

Meter set to 200vac
Red lead to pin 11
Black lead to Pin 13
Reading should be 65vac
 
reading: 62.9 (low)


=====================================
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*
*********
=====================================

Meter set 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)


reading: 6.3

This has me thinking something on the power block is still not working right, as I'm not getting any readings anywhere near 60vac.  I also have no idea where to look on a schematic to find out what the readings should be, and why I have no pin in slot #12, but I have one in slot #11.

Anyone?
-csa

EDIT:  Started a discussion over here at KLOV as well
EDIT2:  Figured out proper pin-outs, attached image and updated chart above.  Readouts look good, though a bit high or low.  I've attached the pin-out diagram for reference.

[csa3d]

I also did a Self Test with the cab, since my sound appears to be working fine.  According to page 7 of the Asteroids Cocktail Manual you are to hear a series of beeps.  High beeps mean something passed, low beeps mean that item failed.  When tested, I hear only one low beep, meaning it's failing on the first RAM I'm guessing?

I am confused though, because the manual states there are to be 6 RAM locations (D2, E2, M4, R4, N4, P4).  Given my board...

D2    slot appears factory blank, no chips there, what do I do here? (see attached photo)  <-- Low beep, guess I am failing here??
E2    is shared with F2 and has a big chip in there (guess that means the RAM is part of the ROM?)
*see edit below *

M4    no questions here
R4    no questions here
N4    no questions here
P4    no questions here

I'm attaching all RAM/ROM pictures along with the manual descriptions.  Any suggestions??

-csa

EDIT:  Upon further investigation of the PDF manual I have, the PARTS list shows RAM being in locations (L2, M2, M4, N4, P4, R4). This differs from the chart on the SELF TEST page, which appears to incorrectly identify D2 and E2 as Ram locations.  Based on the PARTS list, all 6 chips are identical in that they share a common "UPD2114LC" code on the chip.

[csa3d]

more images..

[SirPeale]

If the RAM is socketed, you could pull it and swap it with another of the same kind on-board.  If it then passes that chip, it's either a bad chip or an intermittent connection.

[csa3d]

How does one identify a chip by popular name to order new ones with?  After researching a bit.. my best guess is that it has something to do with the printing on the chips.. and I'd also gather that it should begin with "UPD"?  I'm guessing my ram is type "2114", but I'd like to learn for sure so I can ID other chips/Roms on the board properly.

-csa

[csa3d]

I should also add, that I can hear some ticking/static somewhere in the monitor.. either in the EHT or the around the monitor neck. With lights turned off, the monitor neck has glow.. and I'm unable to see any sparking in the EHT.. so I'm not sure what's going in there. I cannot see any parts on the EHT board which appear burnt or blown since I've replaced them.

I can say that every time I've discharaged the monitor, I have never once heard a zap noise.

Also, when I was cleaning out parts on the monitor, the diode that connects the D904 EHT Rectifier (connects to anode tube) was SUPER sticky. I disconnected it, cleaned it as best I could, then put it back together.

There is a slight "hot" smell that I get after keeping the unit on a while poking my in there for inspection which makes me nervous. Not sure if this is something I should expect from 40 year old electronics or if something bad is going on.

-csa

[csa3d]

I placed an order for a logic probe, because I was told I need to test to see if my watchdog is barking:

If you have a logic probe, check pin 40 of the 6502 and see if that is pulsing. If it is, then either your watch dog is barking or your reset circuit has an issue.

If the watch dog isn't barking (pulsing) and the reset is, then you probably do have a bad RAM or something to do with the RAM (like address decoding, etc..)

In the meantime, does anyone have any suggestions what I can do to test the monitor?  Do I need working RAM/board first?  I'm holding off ordering 6 new rams until I test the watchdog, so hopefully if I have to order parts, I can order them all at once.

-csa

[SirPeale]

Wish I had an answer for you on that one.  Rasters are easy, there a million ways to get a raster signal.  Vector?  You'll need a vector generator.

[csa3d]

I have tested pin 40 of the 6502, and it tests as constantly being high. Does this test satisfy the watchdog test, or is that a separate test. If it's a separate test, can anyone tell me how to test that now that I am the proud owner of a logic probe?

Also on the good news front, today I realized there is a single point of light in the middle of the monitor, and no LEDs lit up on any boards in the monitor. I can hear what I"m guessing is "chatter", but my board is still not passing self tests.

Is the next step to purchase 6 things of new 2114 ram? Do I need to buy a HV diode or should I run with the one I have? I've read these are sometimes suspect as well.

-csa

[Spyridon]

Glad to see you are taking on this project, but you might be in over your head a bit right now (I know I would be).  Have you considered buying a working PCB and working monitor to get this going and using that to help trouble shoot the non working parts?  Once you get them working, you could sell them to make back the money you spent on the working parts?

[csa3d]

Glad to see you are taking on this project, but you might be in over your head a bit right now (I know I would be).  Have you considered buying a working PCB and working monitor to get this going and using that to help trouble shoot the non working parts?  Once you get them working, you could sell them to make back the money you spent on the working parts?

I have not considered that, no.  I am in a little over my head, but I don't feel overwhelmed just yet.  I feel pretty good that I can pick up what's going on with a little guidance and some hours of reading.  My biggest thing is you experienced guys call chips by names I'm unfamiliar with, and don't see listed as such on schematics.  It takes me some thread searching and cross referencing different information to figure out what is what.  From there, the tests seem rather simple and I'm pretty comfortable with program logic since I write code at work.

It's also kinda fun to me, so it's something I feel I would like to learn if possible.  If it continues to along the path of problem after problem, I might eventually get frustrated/impatient and just give up.  That being said, it's not really in my nature to do so.

I've placed an order from Bob for some new ram chips.  After all, the board is failing self test and blaming ram according to the test results.  Given that I've ensured my machine is not constantly resetting, I'm willing to buy that they are bad.  Especially given the number of blown fuses/pots/and burned traces on the machine when I got it.

-csa

[Spyridon]

Good to hear.  Keep at it and good luck.

[SirPeale]

The RAM chips aren't socketed (at least they don't look socketed in those pictures).  Being able to remove the old parts CLEANLY is very important.

I recently invested in a desoldering unit - the Aoyue 474A+.  It was ~$130 shipped plus a few extra parts (get more spring filters and sponge filters!)  A 100% necessary purchase IMO.  I was using a solder sucker and braid for years, and it was taking me forever to remove components.  Now I can remove a 40 pin chip in a matter of a few minutes. 

I was previously cutting out the old components, which sucked if I was unsure if they were bad or not.  With the desoldering unit it preserves their integrity until I can determine their worth.

And: don't forget SOCKETS.  Any time you remove a component, put a socket in its place.  They're cheap, and if you need to remove the component again for any reason it's just a matter of popping out the old one and replacing instead of unsoldering and possibly damaging the board.

[csa3d]

The RAM chips aren't socketed (at least they don't look socketed in those pictures).  Being able to remove the old parts CLEANLY is very important.

I recently invested in a desoldering unit - the Aoyue 474A+.  It was ~$130 shipped plus a few extra parts (get more spring filters and sponge filters!)  A 100% necessary purchase IMO.  I was using a solder sucker and braid for years, and it was taking me forever to remove components.  Now I can remove a 40 pin chip in a matter of a few minutes. 

I was previously cutting out the old components, which sucked if I was unsure if they were bad or not.  With the desoldering unit it preserves their integrity until I can determine their worth.

And: don't forget SOCKETS.  Any time you remove a component, put a socket in its place.  They're cheap, and if you need to remove the component again for any reason it's just a matter of popping out the old one and replacing instead of unsoldering and possibly damaging the board.

Yeah, I've placed an order with Bob for more Ram and sockets to match if he has them.  I agree, if I've gotta spend the time taking them out, sockets will make it 100 times faster in the future.

On the front of a desoldering station, what about that device makes it cleaner?  I've never seen one work, and assume the old solder is heated off then trapped somehow.  I've been using a braid.  The sucker I purchased first never seems to do anything other then irritate me by not removing anything I've melted.  Guess I should look into one of those.  I was hoping the desolder braid would allow me to clean them up enough to reuse them for testing purposes.

-csa

[SirPeale]

Oh...my...god...the desoldering unit is so awesome.  It surrounds the entire joint with heat, then uses a pump to suck away the solder, leaving a nice, clean, solder free hole.    It literally takes just a few moments to clear away the hole.

I'd been using a solder sucker and braid, and more often than not the solder on the TOP of the board wouldn't even melt.  Then you'd have to solder the joint again and resuck...might have to do that several times.  And of course the constant application of heat can damage the traces, the plated thru-holes, the PCB itself.

I always thought "what does a desoldering unit do that an iron and a solder sucker can't?"  And since they're usually pretty pricey I thought I'd never get one.  At ~$130 SHIPPED, it was worth every penny.  The time saved removing components...the money saved by being able to salvage possibly good components...but especially the time. 

http://www.sra-solder.com/Soldering_irons.htm is where I got mine from.  Practically within driving distance from me, but I had it shipped anyway.  I can't say enough good things about it and the company.  One of the pieces had a small defect on it, and they shipped a replacement piece out that day, and didn't want the old one back (of course it was just a gasket, but they certainly could have asked for the old one).

If you're going to be in this hobby for any length of time - heck, even if you aren't and just want to use this to fix this one board - the time saved alone will make this machine pay for itself its first use.

[csa3d]

No real news here on this update.  I started a thread in the monitor section over at KLOV to get some advice.  They gave me a bunch of good ideas and some more parts to test/check.  I'll likely look into that after I get the board passing self tests and playing blind.  So today, I'm really just posting parts order/price information here to keep a running tally.  Parts should arrive Monday:

Quantity       Price      Part
----------------------------------------------------------------------------
  6              $15.00     2114-2L ram
  6              $  3.00     IC sockets for above ram
  1              $  2.50     PCB mounting feet
  1              $  1.00     PCB mounting feet to cab screws
  1              $  4.00     Bob's Plastic Alignment tool kit
----------------------------------------------------------------------------
USPS Priority - $7.00
Total - $32.50

-csa

[csa3d]

I just placed and order for a new tube since the one I has not one, but two different burn in patterns offset slightly. My plan is to get everything working with this older monitor, then do a swap out once I know it's all up and running. I picked up a 15ST4730R from Video Display Corp for just under $60 shipped. This is the same tube the B&W Vector Monitor FAQ references on page 21.

Does anyone have any links to any info outlining how one does the actual swap?

-csa

[csa3d]

Ok, removing these ram chips and installing sockets is no joke.  I'm on night #2 of it, and I've got all 6 ram chips off, and sockets mounted in 4 of them.  Should finish this tomorrow evening but man.. what a pain in the 

-csa

[SirPeale]

Should have bought the desoldering iron. 

[csa3d]

Should have bought the desoldering iron. 

LOL.  The irony here is that by the time I got the first chip off, I was mumbling those same thoughts.  I'm 90% certain that after this task I'll own one.

-csa

[csa3d]

Last night I got all 6 sockets mounted into the RAM banks and inserted new 2114 ram into the sockets.  Upon running self test, the first thing I noticed was that the testing tones were much different sounding.  The bad news is that I only get two test beeps.  This means self test did not complete, and there's something wrong with ram, specifically at position #2.  The odd thing is that the beeps were both low tones, and the manual states it's to be a high beep followed by a low beep.. high meaning item passes test, low meaning item has failed.

So I removed all the ram (thank god for new sockets) and put only 1 ram chip on the board in the first ram position and re-ran the test.  I get a high beep low beep!  I put a second chip in, I get low beep low beep.  Hmm... so i tried all the other 5 ram chips in position 2 to see if it had something to do with a faulty chip.  Same results.

This got me thinking that there's probably some circuit item around chip number 2 that is faulty, which is preventing the self test from working properly.  I checked all the traces on the sockets I just soldered for like the 17th time, and they all seemed to check out fine the best I can tell.  Looking at the traces, I started testing different components around ram #2.  There aren't that many components near by, so that didn't seem very complicated to do.

Call it luck, but I discovered a peculiarity in the bank of resistors at the top left corner of the board that all appear to have the same resistor installed.  Checking each with the multimeter, I found one which was not giving me a reading at R1, while all the others gave me readings around 22ohm.  I check the manual, and sure enough all those resistors are to read 22ohm.  So I think I found the offender all by myself  

Hopefully a trip to the store, a pack of 22 Ohm 1/4watt resistors, and the replacement of the resistor at R1 will lead me closer to completing self test.

-csa

[SirPeale]

Pull it out of circuit and read it to be sure. 

[csa3d]

Pull it out of circuit and read it to be sure. 

You're talking about the "bad" resistor, yes?

-csa

[SirPeale]

Yes.

[MaximRecoil]

I recently invested in a desoldering unit - the Aoyue 474A+.  It was ~$130 shipped plus a few extra parts (get more spring filters and sponge filters!)  A 100% necessary purchase IMO.  I was using a solder sucker and braid for years, and it was taking me forever to remove components.  Now I can remove a 40 pin chip in a matter of a few minutes.

I disagree. A good desoldering unit may be gentler than a good solder sucker, but it is not faster; nor does it work better.

However, there is only one solder sucker worth owning, and it is the classic fullsize Edsyn Soldapullt (models DS017, DS017LS, and AS196; they are all the same aside from their level of ESD-safeness). This is what we used at the PCB factory I worked at in the late '90s, and they work perfectly for what they are designed to do; with a single trigger pull; assuming they are used correctly. When used correctly, most components will simply fall out of the board when you are done (sometimes they may need a slight nudge on some of the legs with the iron).

We had high-end Metcal desolder units at the factory also, but they mostly collected dust. Most everyone used the Soldapullts for everything; including removing the terminal blocks off an entire board that some new-hire decided to solder on backwards or with the wrong blocks altogether (usually about 100 of them, 4 legs each); which took about 10 minutes.

Having the right type of solder sucker is important, but knowing how to use it is important as well. A mistake I've seen often is that people are afraid to let the plastic tip touch the hot iron; afraid that it will melt; so they are trying move the iron out of there and quickly get the sucker in there; and that never works right; because the solder will be partially cooled and you will only get some of it out. Here's a diagram I drew for another thread to show the proper relationship between the iron and the sucker when you pull the trigger:



And here's an old post from Ken Layton which sums it up perfectly:

I use an Edsyn "Soldapullt" model DS017 solder sucker. It workers great and it's a professional tool. I've worn out 3 of 'em over the past 25 years since I use it so much. I have no problem removing flybacks with this tool. In fact, it takes about 30 seconds to remove a flyback with this.

There are now appearing some similar looking (and much cheaper) solder suckers that look like a Soldapullt. These knock offs are cheap junk and you're throwing your money away on those knockoffs.

[csa3d]

When I unsoldered the bad resistor it cracked in half, so it was well on it's way out; no need to test it after removal  Put the new one in, did some MM tests, and put the PCB in the game and fired it up.  I have the same results I had before replacing it.  Two low beeps then nothing.  At this point, I have no clue what's wrong or where to test.  Unless someone here has some ideas, I'm off to KLOV to beg for help  

-csa

[MaximRecoil]

When I unsoldered the bad resistor it cracked in half, so it was well on it's way out; no need to test it after removal   Put the new one in, did some MM tests, and put the PCB in the game and fired it up.  I have the same results I had before replacing it.  Two low beeps then nothing.  At this point, I have no clue what's wrong or where to test.  Unless someone here has some ideas, I'm off to KLOV to beg for help   

-csa

Sometimes things can go wrong with boards that are well beyond being able to be fixed by someone who is learning as they go; asking questions on forums. For example, I got a Missile Command last summer, and when I first turned it on it worked. A few minutes later the picture went screwy. I figured that it probably wasn't anything major, because afterall, it was just working a second ago, and it is not as if anything was smoking. I tried the obvious things, like reseating all the socketed chips, replacing all the RAM chips (adding sockets in the process), cleaning contacts, reflowing dubious solder joints, etc.; and I asked plenty of questions on the forums. However, I got nowhere.

I finally sent it out to a Missile Command board repair specialist, and it turned out to be a challenge for even him to fix:

Yeah this one was a doozie. I'll give you some details if you're interested. When I got the board, did the inspection, replaced the crystal, etc, I plugged it in and the screen rolled. Hmmm. Usually a sign that either HSYNC or VSYNC is missing. I did some poking around and found the VBLANK was missing. I tracked that down to a 5-input NOR gate. It seemed that all of the inputs were firing, but there was no output. That's a classic sign of either a bad chip or a shorted load on its output. I replaced the NOR gate - still same problem. Uh oh. It must be a shorted load somewhere sinking a bunch of current from the output. I shotgunned all the loads, took them off the board, and replaced the NOR gate with a socket and new chip. STILL same problem! WTF???  As it turns out, one of the inputs to the NOR gate was *just* strong enough to trigger the voltage threshold on the logic analyzer, but still weak enough to NOT trigger the NOR gate! Son of a gun! I'm going to replace the driver tonight and see what happens.

And then:

Your MC board is now up and running! It was defiant until the end.

Yup ... if I lived to be 100 I never would have fixed it myself (unless I shotgunned the whole board).

BTW, his rates were extremely reasonable (don't know if they have changed since last summer or not):

Flat fee of $35 + parts + return shipping.

His name is bit_slicer on KLOV. I don't know if he works on Asteroids boards or not, but if you hit a deadend, you might want to send him a PM. He was a Godsend for me (and my Missile Command still works perfectly).

[csa3d]

Sometimes things can go wrong with boards that are well beyond being able to be fixed by someone who is learning as they go; asking questions on forums.

   Yeah, perhaps I'm at that spot.  I guess I'm hesitant because I've changed out so many parts on the board myself all ready, that part of me feels it's going to be one or two more tiny things that are broke that bring this back to life.  I'll give it a few more days then I'm likely gonna start contacting people about board repair.  Bummer..

On a brighter note, my tube replacement has arrived from Video Display Corp.  To my surprise, it arrived without being broke!  They packed the tube extremely well, and the box says in big black letters all over it "CATHODE RAY TUBE", "GLASS", "FRAGILE".  hahah  I suppose that gave the shipping guys enough warning to not beat the crap out of the box.  Too bad it's now gonna sit in that box until my board is repaired.

-csa

[csa3d]

Here's some images from the Ram --> IC Socket replacement I made this week.  I'm currently talking with another KLOV member about possibly sending out my board.  Quite a shame.. I feel like I've all ready dumped a bunch of money into trying to fix this myself, but I'm stuck so what do you do 

-csa

[csa3d]

more images..

-csa

[MaximRecoil]

Double-check all of your soldering. Some of those fillets look iffy. For example, on the M2 RAM, in the left-hand column of joints, the fillet on the 4th joint down appears to have a void in it. All of your fillets should be shiny and smooth, and shaped like an Atari Fuji logo.

[csa3d]

Double-check all of your soldering. Some of those fillets look iffy. For example, on the M2 RAM, in the left-hand column of joints, the fillet on the 4th joint down appears to have a void in it. All of your fillets should be shiny and smooth, and shaped like an Atari Fuji logo.

Yeah, that was a hard area to solder.  What looks like a void is actually a small tail which means too much solder.  Each of 'em passes continuity test, and I also checked that the traces which appeared bare wouldn't accidentally connected to any of the pins nearby which shouldn't have been.  Do you think resoldering them in this case would make a difference?  I was thinking about testing voltages on a chip level next and see what that says.

-csa

[MaximRecoil]

Do you think resoldering them in this case would make a difference?

Probably not.

BTW, it sounds like you need a smaller/thinner iron tip and/or smaller diameter solder wire if those seemed to be in a hard area to solder.

[Spyridon]

His name is bit_slicer on KLOV. I don't know if he works on Asteroids boards or not, but if you hit a deadend, you might want to send him a PM. He was a Godsend for me (and my Missile Command still works perfectly).

I've used bit_slicer for some repair work and highly recommend him.  He's an Atari expert, but I don't think he does any vector games.  You'll have to check with him to see.

[csa3d]

I've contacted road.runner.  I'm gonna ship it out to him to give it a shot.  He comes highly recommended for asteroids boards.

-csa

[Bangarang]

Any word on this yet?  Summer is closing in fast and you'll want something to do indoors with the AC on!

[csa3d]

Ok, here's the update:

Got the boards back from Road.Runner from KLOV.  Excellent work, highly recommend him if you have a board issue.  My asteroids board now plays blindly, as does my newly acquired asteroids deluxe with ROM mod to play in my cocktail.  This is the good news.

The bad news is that my deflection board in my monitor keeps blowing F700, R718, and R713.  This has happened twice, causing a small component fire each time, and melted the two big 6800uf caps thus far, and other various resistors in that area.  I've replaced parts twice now, and don't want to plug the board in again until someone helps me figure out what is going wrong properly.  If I plug in the deflection board again, it's just going to fry these parts, make me angry, and waist 2 hours of my day removing, cleaning, and repairing those parts for a third time now.

Here's my KLOV thread on this monitor issue, which I'd love if someone would chime in with advice, so it's not me just rambling on about things.







 
-csa

[csa3d]

Here's the schematic section that's causing me troubles on the G05-805.. parts in question highlighted in yellow.



-csa

[opt2not]

Sorry for resurrecting this post, but have you got anywhere with this lately?

I may have an asteroids cocktail "fall into my lap" in the near future that I will be restoring as well.
This one has been left out in the rain for about 4 months, so I'm not sure how much salvage there will be... 
Hopefully the glass allowed the rainfall to trickle off the sides, and not into the cab. But I've been told it hasn't been powered up since it's exposure so finger's crossed!