Cheap Quartz Clock movement stops at 1.48V

[Buriedcode]

Hi,

So My folks have a carriage clock that they like, which stopped working.  It has one of those cheap quartz clock movement modules that are everywhere, but rather old, I'm guessing late 80's. 
Initially, I took it apart, carefully cleaned any crud off the ends of the gears, added tiny amount of lithium grease, replaced the battery - and it worked.

Few weeks later, it stops again, so I'm assuming I've unintentionally added dust/contaminants to the mechanics, but just out of curiosty I measured the battery voltage and current draw. 0.3uA, and 1.48V.
A new battery got it working again, with an open cell voltage of about 1.55V.  So I hooked it up to a small PSU I built ages ago that simulates an AA, variable output from 0.9 to 1.54V, and slowly reduced the voltage to the point at which it sopped - about 1.48V. 

I then removed the small magnet that the coil drives, and probed the coil wires - assuming that it was a machnical issue and that it needed the higher voltage to drive dirty gears -  whilst varying the voltage - again, pulses nicely at 1Hz with >1.48V, but stops completely lower than that.

Lastly, probed the crystal lines with my scope and noticed a nice healthly ~350mV sine on one leg, with ~320mV on the other when running at >1.5V.  But once its dropped to 1.48V the sine decreases amplitude until it stops. Only returning when > 1.52V, slowly increasing in amplitude over about 5 seconds.

I realise this is a lot of effort for what is essentially a $2 module, but:

a) I cannot find one with the same specs of shaft diameters for the hands, or the thread and
b) I can't resist taking things apart.

I am unaware of crystals actually failing except in cases of mechanical damage so I'm assuming its the very old DIP8 driver chip that simply cannot drive it at lower voltages anymore.  Is this an age thing? or have I killed it with ESD when I took it apart?

In order to keep the same mechanical parts I think I'll just try a board from newer unit - just to drive the coil, but I am curious if anyone has seen anything like this before - googling just throws up the usual "buy a new one" and "clean the gears".

[Whales]

> added tiny amount of lithium grease

This is probably too thick.

What's the thinnest oil you have?  Can you get some sewing machine oil?

[RoGeorge]

The 1Hz coil moves a little magnet.  Maybe the metallic core of the 1Hz coil somehow become magnetized, and now the same electric pulse doesn't have enough magnetic swing to rock the little magnet wheel.  Or maybe the little magnetic gear-wheel lost part of it's magnetic field with time.

I would try removing the U-shaped magnetic core and de-magnetize it.  Only the metal core, not the entire coil, or else it might fry the electronics.

Maybe try a 10uF+100nF in parallel with the battery.  Maybe add a second battery in series with a diode, so to make it run at 2V+ instead of 1.5V.  Or, as a last resort try to magnetize harder the little magnet inside the rocker geared wheel. 

[floobydust]

Similar situation, I was given an old clock with sentimental value to repair.
What I found is the mech simply wears out and there is too much slop. Calculate how many pulses, gear turns over the decades. Plastic wears and no amount of oil can overcome that.
I had the second hand intermittently get stuck at ~9'oclock because that is when the most torque is needed to rotate up.
Perhaps yours has too much friction and needs the high drive voltage to overcome that. You can check for a VBATT capacitor see if it has dried out as well.

These clock mechanisms come in standard configurations and it was just a matter of finding the arm size and height. I did find it after a long search like Ali.
Newer mechs are not doing the noisy clunk clunk clunk, they actually have a motor that smoothly moves the second hand and I went with "silent".

[Koray]

I once cleaned and lubricated one of these mechanisms. It worked, but similar to your case, it depleted batteries very quickly and stopped after a while. This time I degreased the whole mechanism, reassembled completely dry. It has been working flawlessly, going with one alkaline battery for more than a year. I suggest degreasing with pure alcohol or naphta and check again.

K.

[Buriedcode]

I initially thought it was the mechanics, but that doesn't explain why the oscillator stops when the power is <1.48V, I meant I had removed the board and coil and just put a scope on the coil terminals whilst adjusting the input voltage.

> added tiny amount of lithium grease

This is probably too thick.

What's the thinnest oil you have?  Can you get some sewing machine oil?

I avoided machine oil as most of the gears were nylon, with only two being metal.  I applied a small amount to the axles of the gears with a fine brush, then wiped off any I could see - should have left a tiny film over them.  But as I said, I can worry about the mechanics later.

[bdunham7]

Just use an Energizer Ultimate Lithium which will stay above 1.7V until it is totally dead.

[tom66]

Just use an Energizer Ultimate Lithium which will stay above 1.7V until it is totally dead.

Second this, Ultimate Lithium's are actually considered to be at around 0% SoC when they are at 1.55V.  They have almost no useful charge left at that point.

Your clock will probably run for over a decade with a lithium cell.  Yes it is expensive to buy them but you will waste more time trying to fix it so it can work on a normal alkaline.

[mikerj]

I initially thought it was the mechanics, but that doesn't explain why the oscillator stops when the power is <1.48V, I meant I had removed the board and coil and just put a scope on the coil terminals whilst adjusting the input voltage.

Odd that everyone seems keen to avoid this and blame everything on the mechanism...

It's possible that by probing the crystal you are loading it sufficiently to make this happen, they are very low power oscillators (where you using a 10x probe?).  Another possibility is contamination on the PCB causing leakage currents, careful and thorough cleaning with IPA or similar may be worth a go.  It seems unlikely for the crystal to be damaged but still mostly working without a significant frequency shift.

If it's a standard integrated quartz movement then replacements are very inexpensive, probably not worth spending a lot of time on it.

[magic]

Given that the motor only activates for a brief moment once per second it would be weird if it somehow managed to disrupt operation of the oscillator. These clocks usually run from AA cells which easily deliver plenty of current.

Probably something wrong with the chip, or maybe crystal(?) or other electronics, if there's anything more there... Capacitors maybe? These love to go bad.

[Buriedcode]

Odd that everyone seems keen to avoid this and blame everything on the mechanism...

Yeah I'm not sure I made that clear in my first post

It's possible that by probing the crystal you are loading it sufficiently to make this happen, they are very low power oscillators (where you using a 10x probe?).

Yup, across the crystal stops it completely so I had the ground clip to the battery negative and tested both sides of the crystal - thats the nice sinewave sitting at roughly half the supply.  It does decrease in amplitude for a few seconds when the probe is attached so its clearly loading it in some way.

Before I used my scope I hooked up both the coil lines to a dual schottky diode driving an NPN via 47k - that triggered a 555 monostable so I could see the pulse each second. one of those one-minute-circuits.  It still stops at roughly 1.48V and won't restart until the supply goes above 1.53.

Another possibility is contamination on the PCB causing leakage currents, careful and thorough cleaning with IPA or similar may be worth a go.  It seems unlikely for the crystal to be damaged but still mostly working without a significant frequency shift.

I'll clean the board, looks fine, but its also pushing 40 years old so who knows.  It has one fixed ceramic cap of 39p, and a green trimmer cap which measures about 22p where its set.

If it's a standard integrated quartz movement then replacements are very inexpensive, probably not worth spending a lot of time on it.

Yeah, as I said in the OP it does have a standard thread, but a non-standard shaft for the second and minute hands, otherwise I would have just used the replacement one I ordered (same thread, but slightly different shaft diameters

My option now is a board swap - modern ones have a PCB of similar size, but needs modifying to fit, one of those slow weekened afternoon jobs 


Anyways cheers guts guys, I'm assuming its just the driver chip is failing.

[Whales]

Apologies, yes I didn't notice you mentioning the lack of electrical oscillation in your OP.  My fault.

If it's a 1-tick-per-second design then the coil won't be part of any resonant circuit, so mechanical load shouldn't affect the oscillation.  If it ticks faster (eg many-ticks-per-second-movement, aka a silent mechanism) then perhaps it's possible, but I think it much more likely they drive the coil with some sort of buffering transistor. 

Maybe, possibly, physical load might cause some sort of protection in the circuit to kick in.  But that sounds like complexity they wouldn't bother with.  You could test for this by running it with no physical load/gears attached and see if it oscillates at lower battery voltages.

40 years old is impressive.  I wonder if the silicon could have changed over that time?  Moisture ingress could perhaps change its resistive properties? 

Could you please post a photo of the IC?  Is it a black epoxy blob, are there any external resistors, etc?

[magic]

BTW, is it possible to disconnect the coil leaving only your NE555 indicator in place to completely remove coil and mechanical issues from consideration?


It's not clear what could cause the driver chip to go bad, but I think that damaged quartz (excessive resistive loss) is also at least a plausible theoretical possibility. If all else fails, I would try replacing it to see what happens.

[Gyro]

As it is a 'sentimental item' carriage clock (presumably meaning limited space to piggyback a board from another movement), I'd be tempted to just follow tom66's advice - stick an energizer lithium in it and put it back on the mantelpiece.

[Buriedcode]

As it is a 'sentimental item' carriage clock (presumably meaning limited space to piggyback a board from another movement), I'd be tempted to just follow tom66's advice - stick an energizer lithium in it and put it back on the mantelpiece.

Yeah, thats probably the simplest option, and any worries about "cost" is kind of pointless given the battery life on these clocks. Would be intersting to see how long the cell voltage holds up >1.5v for.

BTW, is it possible to disconnect the coil leaving only your NE555 indicator in place to completely remove coil and mechanical issues from consideration?


It's not clear what could cause the driver chip to go bad, but I think that damaged quartz (excessive resistive loss) is also at least a plausible theoretical possibility. If all else fails, I would try replacing it to see what happens.

I could easily desolder the the coil as well (when I say I removed the mechanics, I mean I removed the PCB + coil from the housing, nowhere near the rotor or gears) so I could try that.

Whilst I added the wires for the test circuit, I did desolder the crystal and tried some I found in my "used parts" bin (no idea of their quality).  I tried 3, with 2 behaving exactly the same and the third requiring a higher voltage to start - which I'm guessing is because its a lower capacitance one, most branded ones I have are 12pf laod anyway.  So xtal looks fine.

Again, I posted partly in case anyone had any knowledge of known issues, or if I was just missing something, and partly because I couldn't really think of a mode of failure that could be the cause, so.. curiosity really.  Its an odd one eh? 

[Gyro]

Yeah, thats probably the simplest option, and any worries about "cost" is kind of pointless given the battery life on these clocks. Would be intersting to see how long the cell voltage holds up >1.5v for.

Well they will run Blink outdoor cameras, PIR, RF link, wifi, IR filter shutter solenoid IR LEDs etc. for a good couple of years. When the ones in my cameras failed (suddenly and close together), the ones I removed read 1.5V open circuit. Coincidentally,  I put them in a couple of quartz clocks, hoping to give them a second life but they stopped working after about a month. Given that brand new standard alkaline cells are not capable of powering the Blink cameras at all, I doubt that the Lithium cells ever drop below 1.5V in their useful lifetime.

I doubt you will ever need to replace one in the clock [Edit: unless the problem is making it pull significant current].

[IanB]

It's not completely unusual for badly designed electronics to require a relatively high voltage. Lazy design, and also that I think zinc chloride cells when fresh start out at about 1.6 V. With only a tiny current draw, you would still have obtained a long run time, and zinc chloride batteries are cheap.

I add to all the previous suggestions: use an Energizer lithium cell and be done with it.

I've done the same thing with another clock that also stopped working at a relatively high battery voltage.