Another Williams battery thread

JDGJr

Member

Donor 4 years: 2012, 2016, 2019, 2024
Joined
Oct 27, 2009
Messages
659
Reaction score
8
Location
Longmont, Colorado
About 2 months ago, i set up my Robotron with 3 Duracell batteries that tested at 1.5v each. A couple of weeks ago, the machine started booting with the high scores cleared, and requiring me to open the coin door.

So i figured i'd swap in the lithium battery kit i had on hand. I pulled the board and checked the combined voltage across the 3 batteries in the holder (replaced in 2008, looks very clean) - the total was .44 volts. I took out the batteries and tested each, giving .78, .79, and 1.12.

How can those three batteries give a combined voltage lower than the average?
Is there something odd on the board that i need to look into?

I saw other threads about older defender boards and battery drain, so I'll ask if anyone knows if these board markings are defender, or robotron: 5770-09656-00 Rev B.

Should i choose to install a battery pack instead of the lithium?

thanks for any suggestions,
John
 
Last edited:
About 2 months ago, i set up my Robotron with 3 Duracell batteries that tested at 1.5v each. A couple of weeks ago, the machine started booting with the high scores cleared, and requiring me to open the coin door.

So i figured i'd swap in the lithium battery kit i had on hand. I pulled the board and checked the combined voltage across the 3 batteries in the holder (replaced in 2008, looks very clean) - the total was .44 volts. I took out the batteries and tested each, giving .78, .79, and 1.12.

How can those three batteries give a combined voltage lower than the average?
Is there something odd on the board that i need to look into?

I saw other threads about older defender boards and battery drain, so I'll ask if anyone knows if these board markings are defender, or robotron: 5770-09656-00 Rev B.

Should i choose to install a battery pack instead of the lithium?
John

John,

5770-09656-00 Rev B = Robotron - http://www.robotron-2084.co.uk/techwilliamshardwareidcpu.html this website is a good resource for you!

batteries, when not underload will recover, but when under load, will drop so how much the loaded voltage drops depends on what capacity remains and the load (drain) itself.

if battery pack goes below 2.7 volts, the memory contents are lost. refer to the defender "Theory of Operation (Early Series Boards)" , CMOS RAM section manual page 26 here
http://www.robotron-2084.co.uk/manualsdefender.html

i did some calculations recently to work out what the Amp/Hour rating should be for 10 year memory retention, and 12 month retention - use the highest rating 1.5v AA lithium cells you can find and they'll last a long time.

also make sure the batt holder contacts are clean AND tight as sometimes even bumping the game may jolt the connection, so if the battery contact breaks even momentarily then contents lost!

if you fit the lithium button cell kit, use a nice high Amp hour rating cell, the highest you can get. you'll need to check the battery specs sheets to find a good capacity cell of the same dimensions as the holder you have.

sometimes some issues with the cmos or associated circuitry may cause a higher than usual current drain, (the threads here have related many such incidents) so review the defender schematics and compare to the logic ics connected to your battery/cmos ram/cmos mem protect circuit anc check the parts used are low power parts, or possibly a cmos ram fault may cause extra drain. fit your new batteries first and time how long they last in practice before you go changing chips, and check the AH rating (either in AH or mAH) as some cheap cells have a very low rating and wont last long..
 
It is also possible that you got a bad battery that drained the other two. If one of the internal cells shorts out this can happen.

Also possible is a "whisker" short where a small whisker of material grew shorting two barreties together. The possibiity of this is much higher when there has been corosion from a previous battery leakage.

When you change the batteries it is also a good idea to clean the area around the diode just to the left with alcohol to minimize corrosion shorts.

ken
 
I've never gotten Williams to work right.. My Stargate starts in bookkeeping mode even with the AS lithium battery kit, a Defender cocktail I just sold started in test mode every second or so start with new batteries.. had 2 Sinistars, corrupt highscores all the time, one new batteries, the other an AS lithium kit and neither solved the issue even with volatages fine.. Never looked beyond the batteries but man, what a pain in the arse they have been to me.
 
I had a Defender board that had a bad (leaky) Diode that let the DC flow to the rest of the board, when It was turned off. Once that was replaced the high score worked properly for months. Recently, I also put in a battery Cap on mine. As long as it gets turned on every two months or so I don't loose my high scores and I don't have to worry about batteries.
 
I had a Defender board that had a bad (leaky) Diode that let the DC flow to the rest of the board, when It was turned off. Once that was replaced the high score worked properly for months. Recently, I also put in a battery Cap on mine. As long as it gets turned on every two months or so I don't loose my high scores and I don't have to worry about batteries.

I know the information is on here, but... can someone point me to instructions on testing the diode referenced above? thanks
 
Pull the board out of the machine. Use the continuity tester on your meter. Mine ended up being one of the ones right next to where the battery holder was. The diodes are intended to act like one way gates to hold DC power into CMOS section of the board. To test place red lead on one side of diode and black lead on the other side. Then reverse the leads. One way will act like a closed circuit and the other side will act like a open circuit or very close to it, a reading of .02 is ok. If the reading is the same either way it is shorted out.
More background:
There is two Diode by the Positive side of the battery holder those would be the ones I'd check. One keeps the positive voltage from going to the batteries when the game is on. The other one keeps the voltage contained in the cmos section when the game is turned off.
I hope this helps.
 
I hope this helps.

Very much so.

the smallest diode (closest to the batteries) reads: 587 / 40+ and climbs higher if i keep touching it
larger diode reads: 165 / 114

might both be bad? what kills these things?
 
The capacitors may be interfering with the test on the top one closest to the battery. You are going to have to desolder one leg and lift it away from the board to test properly.
I just pulled out a meter what it is showing when I used on a 1n4004 as an example. 1.0 with nothing connected
Red lead on positive side indicated by line at the top of the diode No change in continuity meter reads 1.0 no voltage going through.
Black lead on positive side / Red lead on negative side .800. (Voltage is going through)

On one of my old defender pcb's this is what I got
.840/1.0 (One closest to the Battery)
.180/1.0 (One next to it)
I would suspect the large one on your board is leaking
 
Last edited:
I would suspect the large one on your board is leaking

I think you are right. Looking at that one, it is marked 8108, not 5817 like the schematic i have. Plus, while the black part looks ok, the leads on both ends have a slightly different patina than the one beside it. Of course, i have a 4001, but not a 5817, so this will have to wait until the weekend.

Oh, taking the batteries out allowed me to see a much clearer .607/1 from the small diode - is that acceptable?

thanks for the info!
 
Yep, sounds ok on that one. That diode is preventing power supply voltage from getting to the batteries, which is a good thing
since they are not rechargeable. You can try using the 1n4001, make sure it tests good first and does not leak. Once installed, make sure your cmos is still getting the proper voltage around +5 volts on that line. It should not hurt anything.
 
Last edited:
I've never gotten Williams to work right.. My Stargate starts in bookkeeping mode even with the AS lithium battery kit, a Defender cocktail I just sold started in test mode every second or so start with new batteries.. had 2 Sinistars, corrupt highscores all the time, one new batteries, the other an AS lithium kit and neither solved the issue even with volatages fine.. Never looked beyond the batteries but man, what a pain in the arse they have been to me.

I don't know what was in the Sinistar I got from you but it's been fine until about a month ago...
 
You can try using the 1n4001, make sure it tests good first and does not leak. Once installed, make sure your cmos is still getting the proper voltage around +5 volts on that line. It should not hurt anything.

Won't work...

1n5817 is a schottky diode and has much lower forward voltage drop than a 1n400x (about .7V). Battery voltage minus the drop will be too low for the CMOS to hold data.

You could always just drop in a NOVRAM adapter :)
 
I got the 1N5817 from a local shop and installed yesterday. The original diode was definately leaking, and the new install seems to be working as desired.

Thanks everyone for the support!
 
Back
Top Bottom