How is my Robotron is working with only 3.2 volts?

TheYeti

Well-known member
Joined
Sep 3, 2015
Messages
5,699
Reaction score
5,326
Location
Centennial, Colorado
I have a newly acquired Robotron and Joust. The Robotron is working 100% but the Joust didn't have sound.

Since the 2 games are so similar in hardware, I was able to swap the sound boards between the 2 games and quickly isolated the problem to the sound board. (Joust sounds started working but then Robotron sounds stopped) I'll work on the sound board later.

However, after I confirmed that sound started working on Joust...about an hour later the game screen turned yellow and displayed error 1-3-1 on the board's LED.

So I immediately checked the voltage reading on the Joust CPU chip and it was reading 3.1 volts. Yea...that's not good.

I suspected the issue was related to power so I removed and re-connected all the molex connectors on the power supply and it fired up just fine. (I eventually need to convert to Trifurcon connectors to avoid this issue in the future.

But now that the game turns on, I checked the voltage at the CPU again and now I'm getting 4.7 volts. Still low, but the game works so I guess that's OK?


So just for giggles I decided to check the CPU voltage on my working Robotron and to my surprise, the reading was only 3.2 volts!?

As far as I can tell my Robotron is working fine. So how in the heck is working with only 3.2v?
 
i'd turn that 4.7 up a little to an even +5volts and be done with it

all kinds of screwy stuff can happen if you use earth ground instead of logic ground when testing voltages
 
Dumb question: How do I do that? I didn't think there was a voltage adjustment pot on a Williams Robotron or Joust?

There's a mod you can do to add a POT to adjust the 5V. cambridgearcade.dyndns.org/?q=node/76

But you didn't answer the most important question -> when measuring the 5V voltage with your DVM, where did you put the black reference lead? :001_scool:
 
There's a mod you can do to add a POT to adjust the 5V. cambridgearcade.dyndns.org/?q=node/76

Awesome! Thanks, I will look into that in the future.

But you didn't answer the most important question -> when measuring the 5V voltage with your DVM, where did you put the black reference lead? :001_scool:

I put the black reference lead on one of the negative solder pads where the battery pack was. (battery pack not currently present)
 
It likely isn't working at 3.2 volts, so verify your reading... Check on power supply as well as game board, check a known good 5v to make sure your meter is working right, make sure you arent probing corrosion, etc
 
Ideally, you should measure DIRECTLY across the IC you're testing with no remote connection. Even a few inches of PCB may have considerable voltage drop associated with it.

Also, it may also depend on your meter. What meter (make/model) are you using?
 
your ram would not work at all at 3.2V. so you're testing wrong I would think.

I fixed the links in my signature, nobody told me they were wrong, but in the PSU guide it shows how to measure voltage at chips. I do not recommend measuring voltage at the rams themselves.

I always replace the headers and caps on the MPU and rom boards. there's a multitude of factors that go into losing voltage between the power supply and boards; the molex plugs being equally as important.

the Mark Spaeth voltage mod is kind of dangerous because that adjustment will go north of 6 volts if done wrong.
 
Ideally, you should measure DIRECTLY across the IC you're testing with no remote connection. Even a few inches of PCB may have considerable voltage drop associated with it.

Oh wow...I had no idea that you needed to measure directly across the IC. But that makes total sense now. Considering the voltage drop that could occur over distance. Thanks for correcting me on that. I'm going to take new measurements when I get home. Rep for you!

Also, it may also depend on your meter. What meter (make/model) are you using?

This is the exact multimeter I'm using:

https://www.amazon.com/gp/product/B00KXX2OYY/ref=oh_aui_search_detailpage?ie=UTF8&psc=1

(I know!...I should be using a Fluke!)
 
Oh wow...I had no idea that you needed to measure directly across the IC. But that makes total sense now. Considering the voltage drop that could occur over distance. Thanks for correcting me on that. I'm going to take new measurements when I get home. Rep for you!

This is the exact multimeter I'm using:

https://www.amazon.com/gp/product/B00KXX2OYY/ref=oh_aui_search_detailpage?ie=UTF8&psc=1

(I know!...I should be using a Fluke!)

It's funny, sometimes I read sarcasm when there might not be any ... LOL :)

$18 meter. Humm. Might be OK.
Might just be your measuring technique.

WMS boards won't work with VCC under 4.7V (more or less). That's just how it works ... or doesn't. (the thresholds are lower; but not VCC)

log5.gif
 
It's funny, sometimes I read sarcasm when there might not be any ... LOL :)

$18 meter. Humm. Might be OK.
Might just be your measuring technique.

WMS boards won't work with VCC under 4.7V (more or less). That's just how it works ... or doesn't. (the thresholds are lower; but not VCC)

log5.gif

LOL. :) Definitely no sarcasm here! In fact, quite the opposite! I really appreciate all the help you and others have already given me. THANK YOU!

I agree, I'm sure that I have been measuring wrong. Matter of fact, I think I might have even been measuring the wrong leg on the IC!
 
There's a mod you can do to add a POT to adjust the 5V. cambridgearcade.dyndns.org/?q=node/76

But you didn't answer the most important question -> when measuring the 5V voltage with your DVM, where did you put the black reference lead? :001_scool:

https://www.youtube.com/watch?v=d1ZZ-VOP5GQ&t=3s

here is a quick video i made for installing the adjustment pot if you so choose to do that, but as Mecha said you can turn it up too high so make sure to adjust and seal the pot down with a dab of silicone.

BTW: i have the rebuild kits and cap kits for all these williams power supplies if you need.
 
Dumb question: How do I do that? I didn't think there was a voltage adjustment pot on a Williams Robotron or Joust?

also worth mentioning if you don't want to recap your original power supply you can get a switcher kit from arcadeshop it will take a few min to install ;)

http://www.arcadeshop.com/i/892/williams-power-supply-conversion-kit.htm

arcadeshop said:
Compatible games:

Blaster
Bubbles
Defender
Inferno
Joust
Joust 2
Make Trax
Moon Patrol
MotoRace-USA
Mystic Marathon
Robotron
Sinistar
Splat!
Stargate
Turkey Shoot
 
https://www.youtube.com/watch?v=d1ZZ-VOP5GQ&t=3s

here is a quick video i made for installing the adjustment pot if you so choose to do that, but as Mecha said you can turn it up too high so make sure to adjust and seal the pot down with a dab of silicone.

BTW: i have the rebuild kits and cap kits for all these williams power supplies if you need.

Awesome video! I'm going to see what voltage readings I get tonight and go from there. I already know I need to get Trifurcon molex connectors on both my Joust and Robotron PSUs. The Joust (as noted above is flaky to turn on) and the Robotron has burned areas of the existing molex. Yikes.

Everyone has been very helpful...such a ton of great knowledge here. Thanks guys.

So here is my "To Do" list:

1. Properly check the voltages directly across the CPU
2. Get trifurcon molex for the PSUs
3. Install the 5v POT mod (if need be)

I also should probably install a lithium battery on the PCB for game saves. These AA batteries are no good.

phew! lots to do!
 
Personally I think your voltage meter isn't reading correctly. OR, there's slight corrosion or a little bit of oxidization on that lug you're connecting your ground to so it's putting a little resistance between the ground and your probe.

I used to use the cheaper meters, but as the battery dies on those things the voltages get off. So the danger is you measure it as 3.2, then turn it up to read 5... when it's actually 5.1 and you just turned it up to 6.9 or something and fry chips.

I screwed up a couple boards doing that, eventually bought a Fluke and the thing with them is, if the battery goes weak or starts dying, the meter will just turn off, the voltages stay steady and don't drift off.

An $18 meter, though, isn't really a 'cheapo' per se... so you may be alright. Also you got something more likely on the molex connector so it may just be oxidation on the ground pad you picked...
 
Personally I think your voltage meter isn't reading correctly. OR, there's slight corrosion or a little bit of oxidization on that lug you're connecting your ground to so it's putting a little resistance between the ground and your probe.

I used to use the cheaper meters, but as the battery dies on those things the voltages get off. So the danger is you measure it as 3.2, then turn it up to read 5... when it's actually 5.1 and you just turned it up to 6.9 or something and fry chips.

I screwed up a couple boards doing that, eventually bought a Fluke and the thing with them is, if the battery goes weak or starts dying, the meter will just turn off, the voltages stay steady and don't drift off.

An $18 meter, though, isn't really a 'cheapo' per se... so you may be alright. Also you got something more likely on the molex connector so it may just be oxidation on the ground pad you picked...

Good tips indeed! I'd never considered what would happen when a battery starts to die in a meter. Holy cow, I'b better be careful. I know my meter has a fresh 9v in it so I think that I'm good. Eventually i just need to get a Fluke.

I know my meter might be cheap but I have tested it while rebuilding my Qbert power supply and it seemed to read accurately...for the most part.
-------------------------------------------------------------------
UPDATE:

So last night I took voltage readings DIRECTLY across the CPU IC on both Robotron and Joust. Here is what I found...


ROBOTRON: I'm getting 4.7v pretty consistently. It's hard as heck to keep a steady hand while trying to hold the probe tip on the legs and not bridge them. I noticed that if I didn't make REALLY solid contact with the legs with both probes...then my readings were very low.

Also, in the process of turning the Robotron off and back on...the monitor didn't come one! I was like huh? Well then I looked at the PSU and noticed that only 2 of the 3 led lights were one. So I removed the Molex plugs and the main input molex is really torched.

EHTu9x5xFDmmdf3WsYSLZ2GY2s7hWXUi2CKCq8QUYpkpX92IB



So I re-seated all the molexes and the game and screen fully turned on again. So now I know that I REALLY need to repair those headers with Trifurcons.


JOUST: I'm getting 4.2v on the CPU. Again, really hard to keep a steady hand and make good contact at the same time. Need to probably bust out my fiberglass pen and some alligator clips to get good readings. Some tarnish could be to blame here. Also, the Joust does NOT like to have the voltage read on the CPU. Almost every time I do it, the game reboots. That started to scare me so I stopped taking readings. And the PSU is really flaky on the Joust so I might as well rebuild this PSU to while I'm at it.

So ultimately I'm not sure if my "low" readings are because of my crappy meter, poor probe contact, or that my PSUs are just old and need rebuilds.

Either way...I definitely need to rebuild these PSUs for sure.
 
Last edited:
You can also bump up the +5v by removing R10 on the power supply if it's on the low side. See the robotron manual, page #8.
 
Good tips indeed! I'd never considered what would happen when a battery starts to die in a meter. Holy cow, I'b better be careful. I know my meter has a fresh 9v in it so I think that I'm good. Eventually i just need to get a Fluke.

I know my meter might be cheap but I have tested it while rebuilding my Qbert power supply and it seemed to read accurately...for the most part.
-------------------------------------------------------------------
UPDATE:

So last night I took voltage readings DIRECTLY across the CPU IC on both Robotron and Joust. Here is what I found...


ROBOTRON: I'm getting 4.7v pretty consistently. It's hard as heck to keep a steady hand while trying to hold the probe tip on the legs and not bridge them. I noticed that if I didn't make REALLY solid contact with the legs with both probes...then my readings were very low.

Also, in the process of turning the Robotron off and back on...the monitor didn't come one! I was like huh? Well then I looked at the PSU and noticed that only 2 of the 3 led lights were one. So I removed the Molex plugs and the main input molex is really torched.

EHTu9x5xFDmmdf3WsYSLZ2GY2s7hWXUi2CKCq8QUYpkpX92IB



So I re-seated all the molexes and the game and screen fully turned on again. So now I know that I REALLY need to repair those headers with Trifurcons.


JOUST: I'm getting 4.2v on the CPU. Again, really hard to keep a steady hand and make good contact at the same time. Need to probably bust out my fiberglass pen and some alligator clips to get good readings. Some tarnish could be to blame here. Also, the Joust does NOT like to have the voltage read on the CPU. Almost every time I do it, the game reboots. That started to scare me so I stopped taking readings. And the PSU is really flaky on the Joust so I might as well rebuild this PSU to while I'm at it.

So ultimately I'm not sure if my "low" readings are because of my crappy meter, poor probe contact, or that my PSUs are just old and need rebuilds.

Either way...I definitely need to rebuild these PSUs for sure.

those burned up wires are the AC source for BR1 which produces regulated +5 and unregulated +/-12. you will lose power there. and probably trip out your voltage regulator.

Williams eventually figured out there was too much current being pushed there and doubled the wires. the empty pins are actually duplicates of where the existing wires go. if you got 18 gauge wire on hand you can do this yourself.
 
Some of these can make life a lot easier. Especially if you're using probes with clips on them...

That's an interesting beastie. You clip that on a chip and it carries the signals up those pins so it's more convenient to probe it?
 
Back
Top Bottom