Williams High Score problem

1500points

New member
Joined
Sep 2, 2011
Messages
5,706
Reaction score
107
Location
Iowa
I have a Joust upright and a Robotron upright.

On Joust, the high score save works just fine. it'll save new high scores and all existing scores.

On Robotron, if you get the highest score it will do odd things with existings scores. sometimes one or two will disappear, sometimes all of them disappear.

I have tried the Joust CPU in the Robotron, with the same effect of flakey high scores. So it appears the problem is independent of cpu's or battery issues.

Which leads me to presume the problem is on the Robotron ROM board?

Anyone seen this before or have some technical insight about what is wrong?

The next step would be to pluck all the joust Roms and insert the Robotron roms. I'd really like to avoid that due to all of the potential new problems it introduces with sockets, chips, etc.

Thank you.
 
The ROMs are read only, they don't save the HS table. You should be able to Google which chips hold the high scores. Worst case, if you look at the schematics you should be able to see which chips are being powered by the batteries and start changing those out.

.
 
switching power supply.
1 CPU has original battery holder, not corroded.
Other has a battery pack all black taped up (not my job, but it works).

I'll get out the schematics and start tracing the battery power, as it appears you guys are assuming the HS is saved on the CPU ram, right?
 
I've read about this exact issue and wondered how practical it would be to just wire up a simple little linear voltage board that could be run to only the chip(s) that handle the high score save. Just so you could still keep the switcher in there and handle the single HSS issue separately.
 
I've read about this exact issue and wondered how practical it would be to just wire up a simple little linear voltage board that could be run to only the chip(s) that handle the high score save. Just so you could still keep the switcher in there and handle the single HSS issue separately.

This will not work, or fix the problem. Switchers dump power immediately which does not allow the game to write to the CMOS RAM, which is why you keep losing the high scores.
 
Gotcha Dokert - makes sense. So it sounds like his ultimate solution is to just go back to a linear PS.
 
Interesting stuff.

Guess I'll just let it be.

What this doesn't fully explain to me yet.
Is that the Joust has a switching powersupply and it handles high scores properly, using a same CPU.

On Robotron, it will save normal scores. The only problem is when you get the highest score, it throws everything off. if you never beat the high score it'll save and retain the normal score list just fine.

These Williams games are a constant lesson in electronics.
Thank you again.

Oh and I should add- the scores will go awry while the game is still on. for example, beat the highest score, enter initials, look at the high score table and some of the lower scores have disappeared.
 
Last edited:
I'll check that out.
I'm going to put in a CPU with all 24 fresh 4164 rams tonight and batteries that are supplying the full voltage.

i'll check on that cmos ram, too.

I'm so glad I found a full set of joust schematics in the bottom of the locked coin box when i got that machine......
 
Based on the symptoms, it may also be the connector to the ROM board. The address lines all go across the ribbon cable to the ROM board. You can recrimp the ROM cables and see if that helps.

ken
 
My Joust has a switcher installed and I was constantly losing high scores on it. I ended up building an NVRAM adapter which replaces the original CMOS SRAM with an M48Z02 battery-backed SRAM, similar to this: http://www.robotron-2084.co.uk/techwilliamscpunvram.htm It's been working really well; I haven't lost any high scores or settings since. If the CMOS ends up to be the cause of your problem and you want an adapter, PM me. I made some PCBs/kits for this last year and could do another run of them if needed.
 
My Joust has a switcher installed and I was constantly losing high scores on it. I ended up building an NVRAM adapter which replaces the original CMOS SRAM with an M48Z02 battery-backed SRAM, similar to this: http://www.robotron-2084.co.uk/techwilliamscpunvram.htm It's been working really well; I haven't lost any high scores or settings since. If the CMOS ends up to be the cause of your problem and you want an adapter, PM me. I made some PCBs/kits for this last year and could do another run of them if needed.

I have this NVRAM adaptor in my Robotron, too...got it from the guy who runs the website you mentioned here several years back when he was still making them as I had the same problem in mine after the switching PS went in there. It worked great and I never had a problem after I installed it with the high scores.

But I picked up JROK's multiWilliams board when he first made them which of course now bypasses my original boardset and saves the scores on its own. So this NVRAM chip now sits on the original CPU board unused until such time as I might need to switch back to my original boardset...don't know why that might happen, but you never know ;) . Or it may run out of juice before that time...

AFWIW, I have only had this problem with Robotron even though all of my Williams games that use this CPU type (Stargate, Bubbles, Robotron, Joust, and Sinistar) have switchers in them.
 
It looks like Robotron high scores are working.
i inspected the ribbon cable from ROM board, looked solid.
put in a CPU with fresh batteries and confirmed voltage output on connectors.
put in all 24 brand new 4164 ram with the conversion connector for J1.

When it fired up the high score table reset properly to factory default, willy elektrix

and after half a dozen games the high score reset properly and saved other scores.

fingers crossed that it is fixed. i'm going to assume there was a fruity old 4116 ram even though they passed the onboard diagnostic tests.

Thank you everyone!
 
Back
Top Bottom