Memory backup batteries

[Theo Markettos]

I have various bits of 20-year-old equipment with memory backup batteries.
These are all NiCd, and are quite capable of keeping the memory contents for
years. But they've all started crystal growth, and need to be replaced.

For example, one takes the obsolete Varta V280H, 280mAh 1.2V. The suggested
replacement for this is a NiMH, the CP300H. But the self-discharge is poor:
90% of capacity retained after a month. That means it takes 7 months for
charge to decay to 50%. Given the old batteries could retain memory for 10
years (memory supply current typ 2uA), that's a bit poor. Apparently also
the memory chip (PCF8583) has a habit of not reviving if it's been unpowered
for a long time (not sure if that's charge buildup on unterminated pins or
something else).

So:

Is that the best I'm going to get with a NiMH? Are there any better types?

If I use an alkaline button cell plus Schottky diode instead (to disable the
charging circuit), are there any pitfalls? Electrolyte leakage? Will the
Schottky behave sensibly at such low series current? The RAM chip is
specced up to 6V so a higher voltage won't harm it, but when I look up the
datasheet of a possible Schottky (ST's BAT41) all my currents are off the
scale. Reverse leakage doesn't look like it's a problem, but I'm a bit
worried about the Schottky not conducting at all in the forward direction
with such low currents.

For battery I'm thinking something like an good quality SR44 (one of the
locations is a very confined space, and originally used an 11mAh NiCd - I
might need something even smaller than an LR44/SR44). Renata's SR44 has a
self-discharge of 5% per year: http://www.renata.com/pdf/watch/DB303.pdf

Any other ideas?

Thanks!
Theo

 

[Lostgallifreyan]

Theo Markettos <theom+news@chiark.greenend.org.uk> wrote in
news:SSr*ZcPjt@news.chiark.greenend.org.uk:

Is that the best I'm going to get with a NiMH? Are there any better types?

Uniross now sell a 'hybrio' variant of NiMH in AA and AAA cells. Basically it
has a better separator for its electrolyte, or some such. I read that much,
the blurb wasn't very forthcoming. Anyway, the result is a much lower self
discharge rate (retains 90% after a year), and for memory backup purposes
this is obviously useful, so I imagine that industrial batteries (OEM
purposes) are now made this way too. Try asking Uniross directly. They might
take several days to respond, but you'll get good info when they do. And if
they make them, others will too.

You might also use 3.7V Li-ion types but you'd have to regulate strictly for
4.0V at a very low charge current unless you want to take serious risks with
them.

 

[Lostgallifreyan]

Theo Markettos <theom+news@chiark.greenend.org.uk> wrote in
news:SSr*ZcPjt@news.chiark.greenend.org.uk:

If I use an alkaline button cell plus Schottky diode instead (to disable
the charging circuit), are there any pitfalls? Electrolyte leakage?
Will the Schottky behave sensibly at such low series current?

I mentioned a Li-ion battery in the other post, though I now realise you need
a lower voltage. You can still do it this way. ust use a couple of extra
series diodes to drop the excess. You can use bigger batteries if there's
room to improvise a bit. In 20-year-old gear there often is. Find the best
battery you can fit in there, and use whatever wire length you need, and a
small string of silicon diodes in insulating sleeving, etc.. The lithium
thionyl chloride batteries (if you can justify the cost) are an excellent
primary cell for memory backup. They should guarantee you ten years easily.
The currents are so low that there's no great losses in the voltage drop on
those series diodes. Standard practise was to use at least one, just for
safety purposes.

 

[Lostgallifreyan]

Lostgallifreyan <no-one@nowhere.net> wrote in
news:Xns9DFC639F83560zoodlewurdle@216.196.109.145:

The currents are so low that there's no great losses in the voltage drop
on those series diodes. Standard practise was to use at least one, just
for safety purposes.

Another safety feature I saw for lithium primary cells was a series 1K
resistor. If your circuit works with that added, use it.

 

[Theo Markettos]

Lostgallifreyan <no-one@nowhere.net> wrote:

Uniross now sell a 'hybrio' variant of NiMH in AA and AAA cells. Basically
it has a better separator for its electrolyte, or some such. I read that
much, the blurb wasn't very forthcoming. Anyway, the result is a much
lower self discharge rate (retains 90% after a year), and for memory
backup purposes this is obviously useful, so I imagine that industrial
batteries (OEM purposes) are now made this way too.

Yes, those are available from a number of manufacturers. I have some AAs
(from Vapextech), and they still do self-discharge, but over maybe a year
not a month.

And that's what I'd assumed would be used in memory backup batteries. But,
looking at Varta's datasheets, the self discharge isn't much better than a
generic AA bought at the corner shop (I assume, most battery manufacturers
aren't good at quoting self-discharge on their datasheets).

Try asking Uniross directly. They might take several days to respond, but
you'll get good info when they do. And if they make them, others will too.

Uniross' website only deals with 'consumer' (phones, cameras, AAs) or
'industrial', which appears to be about bespoke batteries. Saft and
Panasonic do a few, but they're quite expensive and have too high voltage
(2.4/3V). Sanyo have some NiCds for memory retention, but far too large
(1200mAh). So it appears Varta is the only manufacturer I can find of
PCB-mount memory retention cells, and those have poor self-discharge.

You might also use 3.7V Li-ion types but you'd have to regulate strictly
for 4.0V at a very low charge current unless you want to take serious
risks with them.

Hmm, I think I'll pass on those. Simple is better. Silver oxide seems like
the way to go, but I can't seem to find any suitable button cell holders (in
the UK).

Theo