Amiga 4000 battery problem

[doctorq]

Hi guys, I hope you can help me out on this issue.

I got 3 A4k motherboards which all were broken at some time (none of them wanted to boot). I decided to get them all repaired at amiga.fr and he also did a splendid job on getting them to boot.

I also wanted him to fix the problem with the batteries so I was able to save my time changes for good on all 3 boards once again, so I didn't have to set the time every time I have made a reset. Unfortuanately he failed on this job doesn't want to fix it under the warranty.

As I don't want to spent more money on sending the boards back and forth, I now cry out for help on this forum.

This is how it is. I have bought a new NimH battery, and soldered it in (I know these are working as I have one axactly the same in my 4th A4k, and tested both batteries on that one). After I power up and run the Time prefs the save button is enable, and the time is set to xx May, 2002. This is the same for all 3 boards, even for those without a battery. If I save the time on the one with a bettery in it, the clock is reset when I reset the computer.

How can I fix this problem? I have tried to turn the small screw above the battery, but that didn't help. What to do next?

I hope someone can help me out on this one.

[Castellen]

For a start, try erasing the non volotile memory in the real time clock using this program:
http://amiga.serveftp.net/MiscFiles/RestartClock.lha

If that doesn't clear things up, you may have a hardware fault, like open circuit data or address lines going to the RTC.  It would seem unlikely all of the boards have the same hardware problem though.

Refer here for repairing the RTC, it's basically just a lot of continuity checks:
http://amiga.serveftp.net/RTCrepair.html
http://amiga.serveftp.net/Battery.html

Failing that, I'd suggest replacing the RTC data latch, U177 (U195 in A3000).
Instructions are aimed for the A4000D, but the A3000D uses an almost identical circuit.  Component references will be different.

Don't touch the "small screw" (trimming capacitor) by the RTC, this is used to set the crystal frequency to exactly 32.768kHz.
You'll need a frequency counter if you want to set it, or lots of trial and error adjustment.

[TjLaZer]

I was going to mail my 2 board to him to do just that!  Both worked but would not save time.  One had save ghosted and one did not.  Both had slightly leaking batts but nothing that you would think would cause damage:  But they did do damage to the RTC area.

So you asked him to fix that and he didn't?  On my 2 4000s with this problem, I ended up fixing one of them on my own with the tutorial that was posted by the previous poster.  The 2nd one I have not had time to try to fix, but am hopeful I will be able to.

[doctorq]

@TjLaZer

Yeah, I wanted him to fix the battery problem on all boards, as well as getting them to run, but the bettery problem wasn't fixed That's the reason why I have made this topic.

@Castellen

Hopefully all I need to do is to run the proggy you told me about, as I'm not that god at using a soldering iron, so surface mounted chips isn't going to be replaced by me.

I will try to look through the two links, maybe something comes up that will help me.

[TjLaZer]

It was fairly easy to fix my 4000.  I followed the very well written guide.  You simply check traces to see if they are connected and examine the board carefully.  I noticed the "darker" traces near where the batt leaked were indeed the broken ones and I added a hookup wire.  I had 2 that I had to do and after that my 4000 was 100% working!

[doctorq]

Hopefully I will be as lucky as you :-) But congratulations on fixing your motherboards.

[doctorq]

Hmm, didn't have much luck here. I downloaded the RestartClock utility and tried to run it. My clock was reset to sometime in 1978, but then I powered down the computer to put an other harddrive in it, and the clock is back to may 2002, so I still can't save the clock.

I can't see any broken traces, so I don't think that's the problem. I can see that the 3 + poles for the battery ends at R179, but I can't see a trace for the - pole. Don't know if there should be one, so please inform me.

I would be able to solder in a new chip if that is needed. Can't replace surface mounted chips though, so I don't hope that I have to do that. Can anyone tell me what to look for next?

[zipper]

> I can't see any broken traces, so I don't think that's the problem.

Didn't you measure them? My nearest trace to the battery didn't look too bad, but it was broken yet.

[doctorq]

[Castellen]

Without seeing it and testing the board myself it's impossible to say exactly where the fault is.

Refer to the RTCrepair link I posted above.  I'm assuming you don't have a logic analyser handy, so in that case it's just a matter of testing each of the PCB tracks for continuity.  All you need is an ohmeter/multimeter.

Failing that then replace the 74HC174 data latch, and failing that, replace the real time clock IC.
Most of the time it's just a PCB track open circuit and/or a dead latch, both of which are easily fixable.
The RTC ICs are becoming hard to get, so with a bit of luck it's not that.
As circuit is not very complicated, it's no big deal to get it going.


I'm planning to update the repair pages sometime over the next few weeks to make them a bit easier to follow, add some more info, and try to make the layout a bit better.