Clock Battery Shenanigans

[Mikey405]

Good morning all.

I wonder if anyone can identify the type (and / or voltage) of the cells in the picture please (there are no markings on them - even underneath the plastic but they are quite thick - a bit thicker than a 2 pound coin). They are part of the ultra-simple power supply from a "Bodet" clock rescued from the computer room at work. I'm not sure if they're 1.2V NiCads or something else. The clock is apparently from the 1970s so I'm assuming NiCad. I'm currently feeding in 3V from a supply and the clock is working fine.

The current-limiting resistors have both burnt up too so if anyone might hazard a guess at what value they may once have been...

The circuit is astonishingly simple - mains-in on the two terminals - each leg passing through one of the resistors, then to the bridge rectifier and then the two cells. The cells then supply a voltage to the clock through the small switch. And that's it... Great as long as the cells don't change their internal resistance.

Unless I'm missing something, if one cell goes high resistance then presumably the voltage will shoot up and burn out the clock circuit, and if the resistance of one or both goes down then the resistors will burn out.

I can hardly believe that there isn't more to it than just using the resistors / cells to limit the current and voltage, but I can't see anything else to it. Am I just being daft?

Thanks all.

Kind regards.

From Mike.

[Vintage_RC]

They look like DEAC NiCad disk cells which were common in the 1960's and 1970's

[paulsherwin]

I suppose using NiCds like that will keep the clock running during a power cut, but in all other respects it seems a poor design. I would fit a different power supply using a linear regulator or a zener. You could even power it from an external wall wart.

[TonyDuell]

Using NiCd cells as a shunt regulator in that way is surprisingly common. Some older HP calculators (The 20-series Woodstock, the 30 series Spice models) did it. It's also not uncommon for a 3.6V NiCd used to battery-back a RAM chip to be charged from a 12V power rail via a resistor.

All is fine if the cells are good (the voltage across them when charged is pretty constant). But if the cells fail open-circuit, as they often do, the results can be expensive.

[Nickthedentist]

Hi Mike,

What an unusual thing.

Presumably, it's essentially a mains-powered quartz clock with battery back-up.

Like you, I don't like the idea of the circuit design at all. The fact that the SRBP PCB with mains on it is so toasted is also worrying.

Personally, I would disconnect that PCB entirely and power the clock directly from a battery or wall-wart power supply, depending on current consumption. But leave the PCB in situ in case anybody really wants to reinstate it at a later date. You might want to remove the NiCd cells in case they leak though.

N.

Edit Sorry, crossed with Paul.

[Nickthedentist]

Or use something like this: https://www.rapidonline.com/vigortro...output-84-2723

More powerful readily versions available.

[Graham G3ZVT]

Originally Posted by Nickthedentist

Hi Mike,

What an unusual thing.

Presumably, it's essentially a mains-powered quartz clock with battery back-up.

Looks like a flip clock movement. All the ones I've seen have used a synchronous motor. Were there quartz ones, with stepper motors too?

[Refugee]

It looks like one of those flip clocks that ran on a single C cell.
They require a pulse every 90 seconds.
They worked better with dry cells.

[factory]

Are those resistors actually open circuit? I'm not implying you should reuse it though.

David

[Mikey405]

Hi all.

Crikey - Thank you for all those replies. I wasn't expecting that many.

Yes, I'm going to put a new safer looking power supply inside - although that may just be a couple of D-cells in a suitable box. The case is well big enough to hold them.

The comment by "Refugee" is interesting because there is a battery holder the size of a C-cell (she sells C-cells on the sea shore) built into the motor assembly but a single 1.5V supply isn't quite enough to power the large complicated mechanism of this clock and the 1Hz (stepper?) (pulse?) motor eventually stalls. It has all sorts of cogs and wheels, levers and springs - and gives the time, the day, the date - and even compensates for leap-years with a weird cam / switch mechanism where you set the year within the 4-year cycle.

Re the resistors - I haven't actually checked to see if they're open circuit. I'll grab the clock down a bit later and measure them. As Nick suggests, I'll leave everything "in-situ" in case I (or anyone else) wants to return the device to standard again in the future, but I'll velcro a small battery-holder inside somewhere.

Thanks again everyone for the interesting posts.

Kind regards.

From Mike.

PS. The "minute" number is all-but silent as it flips over but I almost jump out of my skin every time the hour changes with a loud "clack".

[G6Tanuki]

Originally Posted by Mikey405

The comment by "Refugee" is interesting because there is a battery holder the size of a C-cell (she sells C-cells on the sea shore) built into the motor assembly but a single 1.5V supply isn't quite enough to power the large complicated mechanism of this clock and the 1Hz (stepper?) (pulse?) motor eventually stalls.

I wonder if it was originally intended to take a mercury-battery? I remember there was a C-cell-sized one of those that gave nearly 3V.

[paulsherwin]

You may be able to modify the holder to take a pair of alkaline AAs in series.

[Brigham]

Originally Posted by paulsherwin

You may be able to modify the holder to take a pair of alkaline AAs in series.

..or an Ever Ready No.8!

[Robsradio]

Originally Posted by Mikey405

Hi all.

Crikey - Thank you for all those replies. I wasn't expecting that many.

Yes, I'm going to put a new safer looking power supply inside - although that may just be a couple of D-cells in a suitable box. The case is well big enough to hold them.

The comment by "Refugee" is interesting because there is a battery holder the size of a C-cell (she sells C-cells on the sea shore) built into the motor assembly but a single 1.5V supply isn't quite enough to power the large complicated mechanism of this clock and the 1Hz (stepper?) (pulse?) motor eventually stalls. It has all sorts of cogs and wheels, levers and springs - and gives the time, the day, the date - and even compensates for leap-years with a weird cam / switch mechanism where you set the year within the 4-year cycle.

Re the resistors - I haven't actually checked to see if they're open circuit. I'll grab the clock down a bit later and measure them. As Nick suggests, I'll leave everything "in-situ" in case I (or anyone else) wants to return the device to standard again in the future, but I'll velcro a small battery-holder inside somewhere.

Thanks again everyone for the interesting posts.

Kind regards.

From Mike.

PS. The "minute" number is all-but silent as it flips over but I almost jump out of my skin every time the hour changes with a loud "clack".

Great clocks, I have a few!
You can run this from 2 C cells and remove/isolate the mains backup part, never looked safe to me.
If you look above and behind the hour and minute flaps you will see a U channel, this will accommodate 2 C cells, all you need to do is fabricate 2 battery contacts from some thin springy metal, the slots to fit them in are provided. Then just connect to the movement.
Hope that makes sense
Rob

If you ever decide to part with it please let me know ;-)

[Herald1360]

Originally Posted by Mikey405

Unless I'm missing something, if one cell goes high resistance then presumably the voltage will shoot up and burn out the clock circuit, and if the resistance of one or both goes down then the resistors will burn out.

I can hardly believe that there isn't more to it than just using the resistors / cells to limit the current and voltage, but I can't see anything else to it. Am I just being daft?

Thanks all.

Kind regards.

From Mike.

Maybe you are..... the DEAC cells will limit the voltage to about 3V when good, the rest of the rectified 240V (call it 237V ish) is dropped across the resistors. So.... resistor current will be 237/(R1+R2). Shorted DEACs will only increase the current to 240/(R1+R2). Only a bit over 1%! Open circuit DEACs will cause all the charging current to go through the "works" but the chances are this won't be that much higher than normal since the two high value resistors make a pretty good constant current source.

Apart from the dissipation issue with the resistors, the design isn't so different from the series capacitor PSUs found in much modern kit and similarly the only really unfortunate failure mode would be short circuit R instead of short circuit C. Neither is likely and for good measure there are two resistors effectively in series providing a degree of redundancy if the design is well toleranced.

I think it's an "engineers' 10p" sort of design and a good bit of cost conscious work. How long did it run quite happily in service anyway?

[emeritus]

CPC now do packs of helical springs for C and D cells for about £2 for a pack of 10. I bought a pack with my last order (which despite long Royal Mail delays in some areas, arrived in 3 days by second class mail: unlike the Christmas card received last week).

[Mooly]

Originally Posted by TonyDuell

Using NiCd cells as a shunt regulator in that way is surprisingly common......

I'd not seen this done until very recently when I had an Argos plug in timer fail. I say fail, the LCD display was over contrasted and all segments were black.

When removed from the wall the display returned to normal... and the 1.2v NiCad cell was showing around 2.5v when plugged in.

Haven't fixed it yet... I objected to paying nearly £5 (CPC) for a tiny (they are very small) replacement cell when a new timer costs no more.

[Herald1360]

Replace the NiCd with a stack of silicon dodes, two per cell if you're not bothered about battery backup?