Good one ! S+ MPU kills batterys Fast

[coorslight115]

Heres a good one guys.

I have a 10Mhz S+ MPU board that when you install a new battery it will play fine. No errors or problems.
As long as you leave the machine on it is fine.
If you turn off the machine and return in an hour or so it has no display at all. Reels are stiff. Reset does nothing. If you pull the MPU and test the battery it has less than 1 volt remaining. This was a brand new battery. It you install another battery the board works fine.
We can rule out everything else in the machine because it works perfectly with other MPU boards.
I am using the same SS and SP chips from the other boards.
I have swapped out the Sound chip, Dallas clock/cmos chip, CPU chip. No change.
The best part is it will kill the new battery sitting on the bench out of the tray.
No repairs have been made to the board, no burnt traces, no corrosion from old batteries, not shorted to the tray.


Any ideas

Also...never get a error 12 for the dead battery.   

[stayouttadabunker]

I would take a look at the S+ MPU schematics and locate something very nearby
that may be shorted before the signal for the low voltage is sent out.
I honestly do not know what actually sends out the low voltage [21] error code
but I would suspect that whatever IC component performs that voltage check routine -
is the source of the problem? 
One thing for sure - the problem lies on the MPU board somewhere!

[coorslight115]

The fact that the problem occurs with only the MPU board laying on the bench, You managed to narrow it down to the MPU board.....thank you Bunker!! I would have never guessed that

Only kiddin!

Strange problem for sure

[cowboygames]

If it's killing the battery that fast then something on the board should be getting warm. Granted, coin batteries don't have a lot of power behind them, but I'd use a finger to see what's to warm to be right. Maybe something stuck under the coin try or a bug shorting a connection. Hard to say

[coorslight115]

If it's killing the battery that fast then something on the board should be getting warm. Granted, coin batteries don't have a lot of power behind them, but I'd use a finger to see what's to warm to be right. Maybe something stuck under the coin try or a bug shorting a connection. Hard to say

It is a 1/2 AA battery. Goes dead out of the tray and very clean top and bottom. No bugs or debris. Nothing warm at all. Removed one lead of the battery and put a volt meter in series to measure draw. Nothing removable changed the reading

[cowboygames]

You might try putting a meter lead on the positive battery tab and check it to ground in diode scale, see if it tones. Or throw the board out, they're cheap

[Tilt]

It sounds like the battery is providing power to more than just the CMOS RAM/Clock like it normally would.  More like it's powering (or trying to) the entire board.  What's the history with it?  Worked fine and just started doing it, or used board but new to you?

[poppo]

On a S+ the battery does 2 things. It provided backup for the CMOS via a set of diodes. It also goes into a 3302 comparator. This chip compares the battery voltage to the +5v and when there is a big enough difference it sets the low bat signal which then triggers a 12 error.

See the attached schematic. It's possible CR14 is bad, and as mentioned above the battery then tries to power the whole MPU. To test this, use a good battery and with the MPU out, see if you have battery voltage between ground and both sides of CR14.

[poppo]

After re-reading the original post, I would look hard at the diodes. You won't get a error 12 unless the MPU actualy boots first with a low battery and if the diode is bad, the now dead battery may be pulling down main voltage. In any case, I would look close at CR14, CR15 and U65.

[coorslight115]

After re-reading the original post, I would look hard at the diodes. You won't get a error 12 unless the MPU actualy boots first with a low battery and if the diode is bad, the now dead battery may be pulling down main voltage. In any case, I would look close at CR14, CR15 and U65.

Thanks Poppo for the suggestions, will try them later today and post back

Also the board history is un-known. took it in as a core when I sold someone another MPU.

[coorslight115]

Good catch Poppo.  CR14 conducting both directions. Now I have to dig around and find a IN4148 in my collection of stuff.

[Tilt]

A shorted CR14 diode will allow the battery voltage to power all the +5v components, and kill it really fast as you've seen.  Good call Poppo.  If you don't have a 1N4148 a 1N914 should work fine too.  They're both very common switching diodes.

[cowboygames]

An NTE519 will do it to. That's the ECG cross

[stayouttadabunker]

I knew that if I steered someone like him in
the right direction - he'd find the problem....

Okay, okay...maybe saying that the problem layed on the MPU was a little vague but
I did mention that it could be an IC component on the board schematics...!!

I DO have a question though...what IC component triggers the low battery voltage
and how does it tell the MPU to send out a [21] low battery error code?

[poppo]

I DO have a question though...what IC component triggers the low battery voltage
and how does it tell the MPU to send out a [21] low battery error code?

U65 It's a comparitor. It gets a reference voltatage on pin 6 and the battery voltage on pin 7. If the difference is too great, pin 1 goes low and sends the low battery signal. The processor then gives the error 12.

[stayouttadabunker]

I DO have a question though...what IC component triggers the low battery voltage
and how does it tell the MPU to send out a [21] low battery error code?

U65 It's a comparitor. It gets a reference voltatage on pin 6 and the battery voltage on pin 7. If the difference is too great, pin 1 goes low and sends the low battery signal. The processor then gives the error 12.

Oh! That's fantastic info Poppo!
I wish I knew about how IC components work like you guys!