Hello

I'm trying/planning to run a little circuit (basically just an ATTINY) from a supercapacitor. The supercap I want to use is 1F 2.5V and has a low ESR. The idea is that the supercap will remain permanently connected to the circuit with no on/off switch but will be topped up occasionally from a battery. The maximum current the circuit draws is about 10mA v briefly, after which it sleeps, only drawing a few uA.

Given these constraints, do I need to limit the current when charging the supercap. At the moment I'm planning to do it via a couple of AA cells with a 2.5V voltage regulator. Is this adequate? I don't want to stress the capacitor but just as important don't want it to mess up the ATTINY.

 

You need to set the regulator up like it is charging a battery - so yes you need to limit the current to protect the regulator.
Why do you want the added expense of the cap when you could just use the batteries?

 

Is the Attiny going to run off a significantly lower voltage than 2.5V? You know that the voltage on the cap will be droping linearily as it discharges, so you might need to have a step up converter in there. It really may be better to use batteries instead of a cap. Also, have you accounted for the self-discharge of the cap? It might be significant enough to ruin your day.

 

Is there some reason an ATtiny won't run on (2) AA batteries (3volts)?
If so, there is an ATtiny 43U that runs on one AA cell by using a voltage booster.
I think this looks like adding a fuel tank to a sail boat. It merely adds extra expense and energy losses that simply don't happen when using a battery. The fact that Supercapacitors are amazing doesn't mean they are the right part for the job.

 

Hello

I'm trying/planning to run a little circuit (basically just an ATTINY) from a supercapacitor. The supercap I want to use is 1F 2.5V and has a low ESR. The idea is that the supercap will remain permanently connected to the circuit with no on/off switch but will be topped up occasionally from a battery. The maximum current the circuit draws is about 10mA v briefly, after which it sleeps, only drawing a few uA.

Given these constraints, do I need to limit the current when charging the supercap. At the moment I'm planning to do it via a couple of AA cells with a 2.5V voltage regulator. Is this adequate? I don't want to stress the capacitor but just as important don't want it to mess up the ATTINY.

How briefly is "briefly"? You cap voltage is going to be dropping at 10 mV/s, so it doesn't take very many seconds before you are getting down quite a bit.

How long does it need to supply the sleeping current? At 10uA your cap voltage will drop more than a third of it's starting voltage in one day.

 

The discharge rate isn't going to be an issue. If the supercap has very low ESR, that means there is no internal resistor to artificially limit the current, as there is in some supercaps intended for memory backup.

As for charging, you'll have to look at the data sheet for your cap. Chances are, the capacitor will handle any current that a AA battery can put out, but again, check the spec sheet. I know that the 50 farad boostcaps I use have a maximum current spec of 27 amps, either for charging or discharging.

What will destroy a supercapacitor is overvoltage. If it's rated for 2.5 volts, that really means 2.5 volts. Don't let it go to 2.55 volts. You may want to use a circuit like the one here http://pididu.com/wordpress/solarbike/supercapacitor-voltage-limiter/ to limit the voltage. You would change the zener value to 1.8 volts for your application. The disadvantage of that circuit is that it wastes energy by simply shunting extra current away from the supercap when the voltage gets too high. It's basically a voltage regulator. Choice B, smaller and less precise, is just to use a Zener diode alone.

The reason I mention wasteful shunt regulators at all is that series-type regulators like the LM317 will not be able to supply 2.5 volts out with 3 volts in. Of course, you could use four AA batteries, to supply a 5 or 6-volt input, which you could regulate down to 2.5 volts, and you'd be golden.

 

You need to set the regulator up like it is charging a battery - so yes you need to limit the current to protect the regulator.
Why do you want the added expense of the cap when you could just use the batteries?

Battery is not an option because of the small space and the need for it all to be encapsulated

 

Is the Attiny going to run off a significantly lower voltage than 2.5V? You know that the voltage on the cap will be droping linearily as it discharges, so you might need to have a step up converter in there. It really may be better to use batteries instead of a cap. Also, have you accounted for the self-discharge of the cap? It might be significant enough to ruin your day.

The ATTINY can work down to 1.8V and if I need to I can have the brown-out detector on.
Batteries are not an option.
Self-discharge doesn't seem to be a problem so far.

 

How briefly is "briefly"? You cap voltage is going to be dropping at 10 mV/s, so it doesn't take very many seconds before you are getting down quite a bit.

How long does it need to supply the sleeping current? At 10uA your cap voltage will drop more than a third of it's starting voltage in one day.

I can't say how long "briefly" is; the sleeping current is less than 1uA; in practice it is lasting long enough between top-ups. I only need it to last a couple of hours, which it does easily.

 

Is there some reason an ATtiny won't run on (2) AA batteries (3volts)?
If so, there is an ATtiny 43U that runs on one AA cell by using a voltage booster.
I think this looks like adding a fuel tank to a sail boat. It merely adds extra expense and energy losses that simply don't happen when using a battery. The fact that Supercapacitors are amazing doesn't mean they are the right part for the job.

Can't use a battery. The device is far too small for that. I'm asking about supercapacitors because that is what I need to use, not because I am dazzled by their amazing qualities.

 

The discharge rate isn't going to be an issue. If the supercap has very low ESR, that means there is no internal resistor to artificially limit the current, as there is in some supercaps intended for memory backup.

As for charging, you'll have to look at the data sheet for your cap. Chances are, the capacitor will handle any current that a AA battery can put out, but again, check the spec sheet. I know that the 50 farad boostcaps I use have a maximum current spec of 27 amps, either for charging or discharging.

What will destroy a supercapacitor is overvoltage. If it's rated for 2.5 volts, that really means 2.5 volts. Don't let it go to 2.55 volts. You may want to use a circuit like the one here http://pididu.com/wordpress/solarbike/supercapacitor-voltage-limiter/ to limit the voltage. You would change the zener value to 1.8 volts for your application. The disadvantage of that circuit is that it wastes energy by simply shunting extra current away from the supercap when the voltage gets too high. It's basically a voltage regulator. Choice B, smaller and less precise, is just to use a Zener diode alone.

The reason I mention wasteful shunt regulators at all is that series-type regulators like the LM317 will not be able to supply 2.5 volts out with 3 volts in. Of course, you could use four AA batteries, to supply a 5 or 6-volt input, which you could regulate down to 2.5 volts, and you'd be golden.

Thank you for answering my question. Most helpful.
Out of interest, what happens if you charge a 2.5V capacitor to 2.55V? I charged one to 3.5V by mistake a few times and it actually seemed to cope with it OK.

 

Hello,

Here are some notes on the abuse of super/ultra capacitors:

Failure under Abuse Conditions

Over voltage

• Loss of capacitance
• Increase of ESR
• Bulging
• Possible venting

Over temperature

• Loss of capacitance
• Increase in ESR
• Bulging
• Possible venting

Mechanical Stress

• Deformation
• Broken lead
• Increase in ESR

This comes from this page:
http://www.tecategroup.com/ultracapacitors-supercapacitors/ultracapacitor-FAQ.php

Bertus

 

Battery is not an option because of the small space and the need for it all to be encapsulated

Check out this little guy:
http://www.powerstream.com/ultra-light.htm

 

Battery is not an option because of the small space and the need for it all to be encapsulated

But if it is all encapsulated, how do you periodically charge it with a battery?

 

Thank you for answering my question. Most helpful.
Out of interest, what happens if you charge a 2.5V capacitor to 2.55V? I charged one to 3.5V by mistake a few times and it actually seemed to cope with it OK.

I may have overspoken. Actually, I don't have direct experience with supercap failure modes. I'm always very conservative with charging them, because (at least for now) replacements are expensive. I know with the boostcaps I use, I can charge to 2.70 volts, but only if the temperature doesn't exceed 65 C. I thought that was kind of weird, as I'm used to the cheap capacitors being rated to 85 C. The spec sheet says that if I have 85 C ambient, I'm only allowed to charge to 2.30 volts. And yes, the spec has that extra 0 of precision on the voltage - I didn't just put it in. Could be that when you charged yours, the room temperature was very cool, or it could also be that you've degraded the capacitance of your part, already.

 

A space constraint argues FOR a battery, not against it. A battery has a much higher energy density, whether expressed by energy per volume, weight or cost. And, a battery has much lower leakage and lower sagging of voltage during discharge.