Amiga 500 Clock

[Sim085]

Hello, I'll be honest. I do not know exactly how the Amiga A500 clock is expected to work as I had never used it before. Anyways I replaced the clock battery and set the time to 24/04/2015 13:30 and switched the Amiga off.

I came back after a couple of hours and the time was 24/04/2015 13:40!

I do not remember using the A500 for that long but could be. That said, should the time be the current time? or it keeps the time when the Amiga was switched off?

[zipper]

Of course it shows the current time - if it's working/battery isn't dead.

[Sim085]

Originally Posted by zipper

Of course it shows the current time - if it's working/battery isn't dead.

The thing is this; without battery it always went back to around 1995 when set to 2015. Now after I placed the battery it is staying in 2015 but it seems when switched off the time is not going forward.

[kipper2k]

it sounds like the oscillator circuit is bad, check for corrosion etc, you may need to replace the oscillator or get a new mem card. If it is an a500 + with onboard battery, there could be corrosion on the motherboard.


i think that while the computer is on there is voltage reaching the chip which will update the clock, when the computer is off, the clock is not either getting the voltage or a problem like i mentioned earlier

[Sim085]

I did have a battery leak and I removed the battery and cleaned with vodka and surgical spirit. However years after I removed the original battery I re-opened I could see some chips legs in the battery area had turned blue! I re-cleaned with surgical spirit!

Where is this oscillator?

Originally Posted by kipper2k

it sounds like the oscillator circuit is bad, check for corrosion etc, you may need to replace the oscillator or get a new mem card. If it is an a500 + with onboard battery, there could be corrosion on the motherboard.


i think that while the computer is on there is voltage reaching the chip which will update the clock, when the computer is off, the clock is not either getting the voltage or a problem like i mentioned earlier

[Sim085]

How can I test if crystal is working? With multi meter?

[kipper2k]

I'm thinking that your battery is not supplying voltage to the OKI RTC chip (U9), there is a test you can do checking voltage at the RTC chip while the computer is turned off...

no voltage = no updating RTC when turned off so the time on the chip will be the time when the A500+ last updated it while turned on...

Check voltage on pin 18 (while computer turned off), should be >2.2v (one lead on pin 18 and other lead on -ve terminal of the battery

[wolfchild]

Hi,

In addition to kipper2k's tests, you may want to check that the RTC crystal (Y9) is intact and that the trimmer (TC9) is not covered in corrosion.

My opinion is that the while the RTC chip may be getting power (confirm as per kipper2k's post) the crystal is not oscillating. The crystal's job is to make time 'pass', but in your case it stays the same even after a few hours. The fact that the time was retained seems to indicate that the RTC chip was supplied with power all along.

To test the oscillator you really need a frequency meter or oscilloscope, although touching one of the oscillator pins will probably cause it to stop running.

Ed

[hooverphonique]

Originally Posted by wolfchild

Hi,
My opinion is that the while the RTC chip may be getting power (confirm as per kipper2k's post) the crystal is not oscillating. The crystal's job is to make time 'pass', but in your case it stays the same even after a few hours. The fact that the time was retained seems to indicate that the RTC chip was supplied with power all along.

This.. If there wasn't power, it would lose the time..

Also, the battery backed clock is only 'used' when the amiga is powered off, meaning that each time you power it on, it loads the clock into the RTC of the CIA chip, and that's why time progresses while the amiga is on, but goes back to the last time stored by setclock after a power cycle.

[Sim085]

Thanks for all the replies.

So in short, if the RTC chip is getting power I will need a frequency meter or oscilloscope to confirm / check of the crystal is working or not...

Originally Posted by wolfchild

To test the oscillator you really need a frequency meter or oscilloscope, although touching one of the oscillator pins will probably cause it to stop running.

Ed

[wolfchild]

Well, that's the normal equipment needed to check oscillator circuit functionality.

On reflection, you may also confirm the fault indirectly by executing SETCLOCK LOAD multiple times from CLI. If the time is always the same, then have confirmed 100% that the RTC oscillator is busted. Suggested course of repair would be replacement of Y9 first, followed by TC9 and finally the RTC chip.

Ed

PS. Your location is Lija, Malta? I live in Attard and have a spare 32kHz crystal (and an old frequency counter too)

[Zippy Zapp]

Sorry if this sounds dumb, but in IT administration I have always learned to ask the simple stuff first.

How did you enter the time? Did you do it from CLI with Date? Did you set it in the Prefs? If at a command prompt did you SAVE the clock by SETCLOCK SAVE?

If I remember correctly even without a battery backed clock present the OS can pull time from the date stamp of a pref or file that was updated last, so it might be close to the actual date, but not the current date/time. So if you did nothing else and turned it back on a few days later it would go right back to that same time. I have seen this on a couple of my Amiga's but I don't remember where it is pulling this date from.

That is why I asked about how you saved the time/date. Sorry, if this comes off like I am implying that you are dumb or something, as that is not the case, but it is easy to miss a step.

[Sim085]

I had some time to dedicate to this issue again and still can't find the problem. This is what I did:

Booted into Workbench 3.1.
Opened CLI
Entered > DATE 10-Jul-2015 01:04:00
Entered > SETCLOCK SAVE
Entered > DATE (I can see date changed)
Entered > SETCLOCK LOAD (multiple times)
Entered > DATE (I can see date changed in line with mobile)

I switched of Amiga. Let it for around 5 minutes off. Turned it back on. Booted into Workbench 3.1. Opened CLI.

Entered > SETCLOCK LOAD
Entered > DATE (the date / time was ok ... I think it was 01:15:00).

I went to sleep and this morning i turned the Amiga back on. When I entered the command DATE I get the time 01:40:00! So it seems the clock did tick forward for some time but then stopped ticking forward...

Something I noticed is this...
I pulled the battery out and I checked its voltage and this was 2.9v while yesterday this was 3.2v!

[Sim085]

In just one night another 3v battery was dried to just 2v.

I have pulled the motherboard out again. I got a new battery. Checked it is 3.2V. Put it in. Checked with a multi-meter, put pins on + and - of battery (from back of motherboard) and this shows ... 1.32V.

I put the multipeter + on battery + and the - on U9 - and this also shows 1.16V.

However what I find strange (maybe someone can confirm if this is right or not) is if I put multimeter + on battery + and I put the - of one of the legs of FAT AGNUS I get a 1.10v reading!

Should the battery be giving voltage to the fat agnus. I checked some other points round the board (ex: EVEN CIA (B)) and i also get a 1.10v reading!

Is this normal?

Note: motherboard works perfectly normal when switched on.

[hooverphonique]

are you measuring that voltage directly on the battery terminals? if, there is a (near) short of the batery circuit somewhere..

the battery probably shares ground with the motherboard, that's why you can get a voltage by probing in other places on the motherboard..

[Sim085]

Thanks to all for the help. I think the problem is now solved (time kept for nearly a day, but still checking). I think I will categorize this under "my usual luck"!

So what was the problem?

The battery holder I had bought was made of a battery holder soldered to a pcb that in turn had the three pins to be soldered to the A500+ motherboard. However the + leg of the battery hold soldered on this pcb was a little long and was touching the resistor on D192 (the one going diagonal). I pushed this pin a little and I immediately noticed the voltage went up (i.e. before with a 3V battery was getting a reading of 1V, now I got a reading of 2.8V). So I tried it out and so far it is keeping the time

Thanks again to all for help and suggestions; the SETCLOCK LOAD / DATE / SETCLOCK SAVE was a real helper because that showed me that crystal was ok.