CMOS settings failure

I have the Archimedes 410/1 running Morley SCSCI on RISCOS and lately CMOS settings were sometimes lost. Replacing the two batteries did the trick but one week later the Configuration settings are lost again, not being collected on Boot Up. New batteries again – system ok, but only for a couple of days. Now not even new batteries get the full correct Boot Up. *Configure is all good. Manual says maybe ‘CMOS internal battery’ needs attention.
Anyone any idea what this is about ?
Thanks in advance,
Gary

I have the Archimedes 410/1 running Morley SCSCI on RISCOS and lately CMOS settings were sometimes lost.

Losing CMOS settings like this has been covered many times on the comp.sys.acorn news groups. If you do a search for RISC OS CMOS settings in an archive (half a chance with google groups) you may find the answer you want.
You may find Chris E popping up here to suggest a conversation off-line.

‘Chris E’ here. Unless I’m losing the plot big time Gary’s problem is not related to the RTC problems that we developed our: RTC podule for, details here
As the Archimedes RTC Batteries were not soldered to the motherboard any leakage is unlikely to damage anything on the motherboard.
Gary: Is the clock maintaining the date and time one day to the next?
Does a meter confirm the batteries have lost significant voltage?
The RTC is DIP chip soldered in, if it is the fault, then it’s not difficult to replace with standard soldering equipment. If something other than the RTC chip is draining the battery then it could be difficult to trace.

As Chris has mentioned, it sounds like a hardware issue in the CMOS charge circuit and unrelated to battery leakage damage. More than likely one of the caps has failed and needs replacing.

If you ask over on the 32bit Hardware forum of stardot, someone will almost certainly be able to assist in tracking the faulty component down.

The batteries used on an Archimedes A310, A440 etc are not rechargeable, so no charging circuitry!
The A3000 was the first Acorn machine using rechargeable batteries.

n.b. Andrew tells me he has seen battery leaks cause motherboard damage on an Arc.

The A3000 was the first Acorn machine using rechargeable batteries.

The first? I’m puzzled, because I seem to remember replacing BBC Master Acorn battery packs with DuraCell batteries with a series diode to defeat the charging system and prevent subsequent damage. Or am I going even fuzzier than I have hitherto thought?

Your memory is 100% correct John. My statement was wrong for a few months:-/ The BBC Master was designed to take a AA sized rechargeable and they shipped it with one for the first six? months or so. Then one burst into flames in a School and they quickly came up with the 3 x AA battery nest and offered them free.

Your memory is 100% correct John.

Well, that’s reassuring! I can even remember the school! Isobel! Where are you now? Long time!

Thanks Everyone,
The 2 x AA batteries are nested in a pack/cradle behind the cooling fan and cannot come into contact with anything. Their condition also looks good. I have replaced them (again) and the new batteries caused perfect Boot Up. Clock and date are maintained perfectly. ‘Old’ batteries discharged. So I suspect new batteries are necessary every few days to maintain otherwise great performance and therefore am guessing something is draining the 2 x AA battery pack.

I thought perhaps someone else would have a ready answer, but the PDF’s here will probably help.

page 34 of the A300 manual has the following

4.4.6 CONFIGURATION, NON-VOLATILE MEMORY & REAL TIME CLOCK
If the NVM suffers data retention problems and the RTC fails, then, with the computer power off, check for about 2.8 V on IC16
pin 8. If this voltage is not present, inspect PL11, D3 and the charge state of the batteries (>1.4 V per cell).
If the NVM IC16 consistently fails on the same data bits, change the device.
If the clock fails to run or runs inaccurately, check and if necessary replace XL3. TP1 allows access to the clock signal.

The component numbering matches the diagrams for the A400 (same link page)

BTW. D2 and D3 are likely to be about 0.7V drop across them (anode to cathode)

Thanks Steve, unfortunately I am no ‘Techie’ merely an Acorn RISCOS devotee. Sounds like I need to get this into a dealer – CJE Micro’s perhaps ?

Thanks Steve, unfortunately I am no ‘Techie’ merely an Acorn RISCOS devotee. Sounds like I need to get this into a dealer – CJE Micro’s perhaps ?

Perhaps, but then the typical cheapo meter from CPC or wherever is a simple high impedance Digital meter which will show you the voltages.

Reading the circuit diagrams is as simple as following a street map if you’re not trying to figure out what they do and redesign it.
Open the zip file of A400 diagrams (2 off) and look at the bottom middle of the second PDF – nice big letters say PL11.
When you find PL11 on the board (it’s the battery plug in) the diodes are marked and the symbol on the diagram has a nice cross bar that matches the end of the diode marked with a bar.