I have a LaCrosse Atomic clock. When batteries run down and I change them I have to go through the entire setup all over again from scratch. Have to tell it to find the outdoor thermometer and to find the UCT (Universal Coordinated Time) in Denver CO. Have to set the time zone (mountain), 12 or 24 hr and the dates MM / DD / YYYY. Lately my clock has been loosing connection with the outdoor thermometer. I've replaced plenty of batteries to make sure I am not running on weak batteries. (double A). Every time I disconnect a battery for the slightest amount of time I have to go through the whole setup routine again. But it always makes connection to the outdoor thermometer. The ODT is useful because I like to know how cold it is outside before I go out.

So I want to add a capacitor across the battery, one large enough to hold the clock settings while I slip in new batteries.

Anyone have any suggestions?

[edit]
Somewhere in my stock of mondo-junko I have a small super capacitor, but as usual, I've seen it a hundred times but can't find it when I want it. What about tossing an old cell phone battery across the two AA batteries? Might that be hazardous? Should I limit the recharge current? What ? ? ?

[edit #2]
What about a whole bunch of capacitors in parallel?

 

How long does it take you to change the battery?

How much current does the clock draw while it is operating?

Another option is to just bring out the battery terminals and put another battery across them. Then replace the internal battery, then remove the external battery.

 

I thought about the spare battery for connecting during the change but that's an extra thing I have to build and store somewhere. I know, I know, what's the big deal about that? Well, if you see my storage system you'd be surprised I can even find the system. What I need is an Automated Storage System, an ASS. Of course, if you look at my current system you might think I was an ASS.

As for how long it takes to change batteries, not more than 15 seconds, though I haven't timed myself. As for the current draw - sorry, that I don't know. Only that it'll run on two AA batteries for a year. So the draw MUST be quite low. Just enough to drive the LCD panels (two of them), and to search for the UCT signal every night.

I thought about putting some chargeable batteries in it, but then I thought - what if I want to reset the time? I'd have to open it up and disconnect a battery wire. Having a capacitor that can carry the current for a minute would be all that is needed to change the batteries. And if I wanted to reset the clock all I'd have to do is wait for the cap to discharge.

 

I thought about the spare battery for connecting during the change but that's an extra thing I have to build and store somewhere. I know, I know, what's the big deal about that? Well, if you see my storage system you'd be surprised I can even find the system. What I need is an Automated Storage System, an ASS. Of course, if you look at my current system you might think I was an ASS.

As for how long it takes to change batteries, not more than 15 seconds, though I haven't timed myself. As for the current draw - sorry, that I don't know. Only that it'll run on two AA batteries for a year. So the draw MUST be quite low. Just enough to drive the LCD panels (two of them), and to search for the UCT signal every night.

I thought about putting some chargeable batteries in it, but then I thought - what if I want to reset the time? I'd have to open it up and disconnect a battery wire. Having a capacitor that can carry the current for a minute would be all that is needed to change the batteries. And if I wanted to reset the clock all I'd have to do is wait for the cap to discharge.

How much time are you willing to put into this to save, 5 minutes of setting effort each time you change the batteries? How often do you change the batteries, once every 6 months or year? How long will this thing last 10 years? Let's say you save your self 60 to 90 minutes over the live of this thing. I'm going to nominate you for finding "the ultimate first world problem".

If you do build something, good luck finding it the second time you need it!

 

It would be pretty easy to measure the current draw, but we can come up with an estimate. Are the batteries in parallel or in series? In either case, you are probably down in the 100 microamp range (I'm basing that one a project I did twenty years ago that needed to run off AA batteries (in series) for at least 30 days and the current budget turned out to be a bit under a milliamp). So if you want to keep the voltage droop to less than, say, a quarter volt after fifteen seconds your

Q = CV
I = (Q/T) = C(V/T)
C = I/(V/T) = 100 uA / (0.25 V / 15 s) = 6000 uF

Not huge, but not small, either. You can probably get by with something quite a bit smaller depending on how much voltage droop you can actually handle and if you can improve your changout time. If you can tolerate 0.5 V and can get the change out time to under 5 s (and I bet you can), then that takes you down to 1000 uF right there.

 

How much time are you willing to put into this to save, 5 minutes of setting effort each time you change the batteries? How often do you change the batteries, once every 6 months or year? How long will this thing last 10 years? Let's say you save your self 60 to 90 minutes over the live of this thing. I'm going to nominate you for finding "the ultimate first world problem".

If you do build something, good luck finding it the second time you need it!

But keep in mind that most engineers became engineers because we are lazy. We saw all the things we could make to save ourselves effort -- we just didn't count on how much work it takes to be lazy.

 

Assuming an AA cell has a 2000mAh capacity, a 1yr duration suggests the average current is ~200uA (if my maths is coorect). However, your clock probably draws current in bursts, so your backup source would need to cope with that. That complicates the calculation of any capacitor value.

 

Assuming an AA cell has a 2000mAh capacity, a 1yr duration suggests the average current is ~200uA (if my maths is coorect). However, your clock probably draws current in bursts, so your backup source would need to cope with that. That complicates the calculation of any capacitor value.

I thought of the bursty nature, but that likely actually makes thing better -- it means a smaller capacitor would do almost all the time and if you got unlucky when changing out the batteries then it just means you have to swear under your breath and set the damn clock like you used to have to do every time. Sometimes the penalty for being wrong is mild enough to just endure it and move on.

And like GopherT indicated, you are probably going to spend more time designing and implementing the solution than the solution will save you over the life of the clock. But that doesn't mean it isn't a fair trade if the time spent is fun and the time saved isn't.

 

True to the mantra, why not spend 3 hours building something that will save you (in a lifetime) 2 hours.

Here's the thing, it takes easily 5 minutes to go through resetting everything. No exaggeration, it must be about 5 minutes, not to mention the time looking for the instructions on how to select different parts of the programming, so 5 minutes may be on the low side. 5 minutes every year, in 10 years I save myself less than 1 hour. But the nice thing isn't saving time it's about not having to deal with the aggravation.

So far, testing using a 3300 µF cap at 35 VDC (electrolytic) (it's what I have on hand), the display will remain active for 3 seconds, charged to 5 VDC. It's either 5 V or 12 V. The funny thing is that after about 35 seconds the clock makes a last ditch sonic alert (beep). So there must be something retaining something in there. Maybe that'll do it. I'm going to build it up on the breadboard and see. So far I've been using my power supply.

 

Why are you charging the cap to anything other than the battery voltage? Are you trying to connect it to some point deeper in the circuitry after some DC-DC converter?

 

No. I have a 5 v / 12 v power supply. I'm about to solder the 3300 µF to the battery terminals and plug in the batteries and then go through the setup procedure, then change to new batteries.

Keep in mind the reason for being inside this thing at all is because the darn outside thermometer drops out, and I have no accurate estimate of the temperature outside my door. So I've been through the circuitry looking for bad solder joints and anything else that could cause the problem. I found some cotton fibers bridging between two wires. M A Y B E ( I said "Maybe" ) humidity and other contaminants present on the cotton may have been causing the problem. Again - M A Y B E.

So while in here I decided I don't want to go through all the messing around with reprogramming every time it loses the temperature signal. I don't know if this will solve my problem but I'd like to give it a try and see.

 

OK, here's the last word on this project. A 3300 µF cap across the battery terminals will allow at least 15 seconds of disconnect time without losing the programming. After 20 seconds all programming is lost. But that's with good batteries being disconnected. However, literally, if I disconnect the batteries, after 12 seconds the display is still visible, and it can take a second to change a battery. The OLD way, the instant you disconnect, you lost the programming. The good news is that if I truly want to reset the clock all I have to do is disconnect the battery for 30 seconds and I'm ready to do a full reset.

Oh, and the outside temp sensor is responding nicely. So far. Sometimes it's good for a few hours, sometimes it's good for a few weeks. Can you see my frustration? Having to reset the thing just to know what the outside temp is?

Anyway, I'm ready to put all the screws back in and call it a finished project. 3300 µF @ 35 VDC does well enough. Now, only time will tell if the cap drains the batteries.

But now I'm a pro at programming it. At least for today. Who knows what I'll remember of it next year.

Thanks all.

Happy Festivus (for the rest-of-us)

 

I did this about a year ago on my remote thermometer receiver, it's an LCD screen that unpaired itself with the outdoor sensor everytime the batteries we're changed. I think I put a 2200uf capacitor across the batteries (2 x AA) and it works great, the screen stays on for about a minute and even when it's faded to nothing it still stays paired with the sensor when I replace the batteries. There has been no impact on battery life.
I've got a remote power monitor on the incoming house supply that I'm going to do the same to when I get round to it.

 

I'm going to do the same to when I get round to it.

Here you go.

 

Here you go.

I had one of those. Once. A long time ago. They're hard to come by (I guess). I heard a fable, square, rectangle and triangle Tuit's exist, but I've never seen one, let alone gotten one.

 

Anyway, I'm ready to put all the screws back in and call it a finished project. 3300 µF @ 35 VDC does well enough. Now, only time will tell if the cap drains the batteries.

That IS a concern. Large electrolytic capacitors are known for leakage current issues and it wouldn't take much leakage current at all to cut your battery life in half or worse.

 

it wouldn't take much leakage current at all to cut your battery life in half

Well, I guess I'll see soon enough. But still, I don't have to reprogram it every time I disconnect a battery.

 

It's a pretty simple matter to measure the current draw of the clock from the batteries with and without the capacitor in the circuit. If you don't have the capability to make those measurements (all you need is a cheap VOM, which you can usually pick up for under $10 -- and which you can often get for free from places like Harbor Freight) then you should invest in one (or two), since they are very good things to have around the house anyway.

 

Measured something. Don't recall what it was. I think it was something like 2.4 mA. And I'm not going to pull the clock off the wall and open it up again.

 

Measured something. Don't recall what it was. I think it was something like 2.4 mA. And I'm not going to pull the clock off the wall and open it up again.

I understand not wanting to open up the clock again.

If what you measured WAS 2.4 mA (and assuming that this was after installing the capacitor) then you are not going to be very happy. You are probably looking at a battery life on the order of a month or less. Hopefully it was more like 0.24 mA, but it doing a quick look at 3300 uF / 35 V caps, it appears that leakage currents in the >1 mA are not uncommon.