Juan smash Williams linear power supply!!!

JuanUp

Active member
Joined
Jan 25, 2007
Messages
904
Reaction score
36
Location
Orange County, California
OK, I've been tackling this for too long now. I'm not making much progress but I've narrowed it down to why my Stargate won't boot up. I ordered the deluxe Bob Roberts Williams Power Supply, replaced practically everything and still nothing. Also, I've checked and cleaned all of the connectors, pins and fuses. The only components I haven't replaced are the TO92's and the TO220's, (not including the resistors and diodes which checked OK). I realized as I was driving home that I should of checked the voltages off the transformer as well. EDIT: I believe these are 2 MPSA55 and 2 C122A respectively. This is from the D-8784 power supply that I reading from Sinistar and Moon Patrol schematics. It looks like there are tons of test points I can read off of to isolate this problem. Something to check tonight. Gee, woulda' helped to have this available when I was working on it. ;)

Anyway, off the power supply (4J2) to the decoder / RAM board I'm getting:

1: -5.1 (-5)
4: -13.7 (-12)
6: 10.57 (+12) <-- problem voltage
12: 4.97 (+5)
13: 4.97 (+5)
14: 4.97 (+5)

These numbers are pretty much the same after I replaced most of the components.

Also, there was a 2N6059 (top) and a 2N3055 (bottom) that came in the kit which I installed on the heat sink. EDIT: Verified.

Another note is that the 2N3055 was bad when I tested it. Could this of fried the RAM? If so, is there a way to test them? EDIT: No, they weren't.

One other thing, is there a schematic of this newer power supply? I believe the Stargate drawing set I finding only has the older power supply. EDIT: OK, it looks like the Sinistar and Moon Patrol has the newer power supply schematics.

Again, I'll check voltages off the transformer tomorrow morning. Any help would be great and please correct me if I'm wrong on anything. I'm new to all this but I've been reading other posts related to this before I posted.

Thanks,
Juan
 

Attachments

  • WPS_before.jpg
    WPS_before.jpg
    61.3 KB · Views: 32
  • WPS_after.jpg
    WPS_after.jpg
    62.6 KB · Views: 30
  • WPS_heatsink.jpg
    WPS_heatsink.jpg
    81.9 KB · Views: 27
Last edited:
Progress!!!

OK. After a few hours of poking, prodding, cleaning, sanding, measuring, hair-pulling and frustration; I finally got the Stargate at decent voltage levels and finally showing the rug pattern.

CPU 1J3
9: -5.04
8: 11.48 - 11.64
7: 4.97
5: 11.89

The game now boots, shows rug pattern, the LED displays error code '131' then reboots doing the same over and over.

I'm guessing at this point that pin5 is still too low or a failed discrete component. Please chime in if you have any insight. ;)

Juan
 
Check the voltages at the 4 comers of the RAM chips. If they are 4116s the voltages should be +5, +12 and -5 volts. Also swap RAM 31 with a different RAM once in a while it really is that one that is bad.

Ken
 
Thanks, Ken. Yep, one of the first things I did was swap them. :) I just tested the voltages before heading to work. Here's the poop:

1: -5.04
8: 11.85 <-- ugh, yeah... bad.
9: 4.93 <-- had to cut R10 from the power supply to get it up to here.
16: 0.0

I'm thinking I'll double-check everything coming out of IC1 and R2 on the power supply and go from there. Has either the MPSA56, C122F or the Zener known to go bad?

Juan
 
Thanks, Ken. Yep, one of the first things I did was swap them. :) I just tested the voltages before heading to work. Here's the poop:

1: -5.04
8: 11.85 <-- ugh, yeah... bad.
9: 4.93 <-- had to cut R10 from the power supply to get it up to here.
16: 0.0

I'm thinking I'll double-check everything coming out of IC1 and R2 on the power supply and go from there. Has either the MPSA56, C122F or the Zener known to go bad?

Juan

The overvoltage circuit has been known to go bad (the three chips you referenced) but usually the output voltage on the corresponding power supply (+5 or +12) is 0.

It sounds like one of three things right now. First would be the LM723 voltage regulators may be bad. Second would be that one or more of the reference resistors on the input lattice for the voltage regulators have drifted. It's a pain, but you may have to desolder one leg at a time and measure the resistance and match against the specs. Third, if you replaced the large bridge rectifier, you may need to solder the wires to the top pads. The plated through holes can be damaged and the solder won't flow to the top pads.

ken
 
w00t!

I ran back and followed your advise. Check out the voltages off of one of the RAM chips:

Pin1: -4.96
Pin8: 11.95
Pin9: 5.08
Pin16: 0.0

Success! However, I'm still getting the '131' error code. Yesterday, I re-seated ROM3 when I noticed that it was getting super hot. Would the ROM board even show error codes if they were bad? I'm reading from elsewhere that this can be due to the interconnect cable, discrete components or those little filter caps next to the RAM chips. I'll keep probing. Anymore wise words would be helpful. :)

Thanks, Ken!

Juan
 
Good news and bad news...

I got the power solid as a rock and know a little something about those 723 voltage regulator ICs. I got the 12V regulated spitting 12.6V and hits the RAM at around 12.3V now so there's no doubt voltages are to spec.

Unfortunately, I still have the 131 error code. :( I busted out the oscilloscope and it looks like one of the pins, (3 or 5, can't remember at the moment), of the RAM has intermittent 2V pulses in the 5V signal. Crap... I can't remember which pin! Anyway, looking much like bad connection somewhere or TTL IC failure. I'll update which pins are doing what but 2V pulses sneaking in there doesn't look good.

EDIT: It's pin 2. Looking at the schematics, it looks like I might have a few bad 74LS374s since I don't see that pulsing on the lower RAMs. Fingers crossed.

If I'm talking out of my ass, please correct me. I want this damn thing fixed.

Juan
 
Last edited:
If you can determine which pins have the extraneous signals, you can use the schematic to determine which of the 74LS374s are going bad. The processor data bus is also shared with the 74166 chips which have also been known to go bad.

ken
 
Frustration...

It's been a few days since I got back to this. I wanted to give a heads up.

I decided that before I start pulling chips that I check the 74LS374 and the 74166 against my Robotron. Sure enough, the Robotron has the same 2V pulses on the same pins and on the RAM and the game works perfectly fine.

I'm left with to test all ICs on the main board; a possible bad interconnect cable, (might of damaged it from pulling it off a few times); and / or a slim chance that the decoder ROMs might need to be replaced, (which Bob Roberts recommends when all else fails).

I don't know at this point. I'll start with the ICs.

Juan
 
Last edited:
If you are going to be testing chips, watch on ebay for an HP 10529a logic comparator. It has proven to be a serious time saver for me, once I figured out how to use it. It works by taking a known working chip and running it in parallel with the chip in question. If the test chip screws up, an LED lights up on the 10529 and you can replace the bad chip. It is not infailable, but it will catch a lot of things the logic probe won't.

There is one on ebay now, but the seller is way high. I have seen them as low as $20 each. I think I paid $50 or $60 each for mine and they are worth it.

I just wish that they had made an 18 or 20 pin version of it.

ken
 
Excellent news!

The Stargate is up and running... intermittently. ;) Looks like the 74166 shift register for bank 1 is konking out causing the game to reset every once in a while.

Either way, the power supply is solid as a rock now. :) Thank you so much for the advise and help, Ken!

Juan
 
Back
Top Bottom