Capacitor charge trigger

Status
Not open for further replies.

Kisen

Member
Hello,
I am designing with a little renewable energy device that captures vibration and stores the energy in a li ion super capacitor. Super cap is 3.8v 40F.
This supercap/vibration generator is the only source of power to the system.

There will be a buck/boost circuit between the supercap and load that is regulated to 3.3v.

The problem i can for see is the capacitors discharge. The capacitor, as soon as it gets any charge will discharge into the buck/boost and power the load circuit, but only for a short time until the capacitor is empty. This as i see it means that the circuit only powers up when the generator is picking up vibration. There wont be enough time to do everything that the circuit needs to do using this method.

So what i think i need is a trigger. This will allow the capacitor to charge up and the load wont see it. When it reaches a threshold voltage, say 3.5v. Then the capacitor is allowed to power the rest of the circuit. This allows the device to gather its power over a period of time and then trigger the devices functions with plenty of stored power.
Once triggered ideally i need the capacitor to be allowed to drain continuously until either fully depleted or or until a signal is sent to it to collapse the trigger.

Is this possible, do you guys have any ideas?
 
If you tell us what you are trying to do then we may be able to help. I know you think you have told us above but you've described your solution to the problem, not the problem.

Mike.
 
Hi Pommie,

I shall try again.

The project is a tracker with a renewable aspect to it. Basically it has no batteries that need charging or replacing.
The energy source is a vibration generator. This charges up a super capacitor.
What i want to do is collect charge from the generator over a period of time and then when enough charge has been gathered in the supercap, use it to power up the load circuitry, which is a buck/boost regulator powering a uC and a radio module.
When the uC finishes its job it needs to shut off the supply from the cap so as to preserve whatever power is left in it ready for the next cycle.
 
The circuit (and device) you need is called comparator. This has an output that changes state when the input voltage exceeds a reference voltage. You will need *very* low power devices, such as a CMOS opamp and an ultra-low current reference. How much power (watts or amps) is coming from the generator?

ak
 
Hi Analogkid,

Thanks for that, however i dont know how much power the generator will create yet. I havn't built it. One thing is for certain and that is that at times it will obviously generate nothing at all.
Am i correct in that when the generator creates power, it will put majority into the capacitor and a little amount will be used to power the comparator and reference voltage?

So from what you are saying i use a comparator and voltage reference to look for the capacitors voltage. Then its output changes to drive a thyristor to latch the power on. How could this be cut off when the microcontroller finishes it job?
 
Have you written code for the tracker, ie can you add code to the processor to control the system?
If not then you could rig up some voltage controlled power switch maybe using a fet and a zener.
 
A supercap is generally a gold or graphite? based unit, Li-Ion is a rechargeable cell or battery.

However, for your application, a lithium cell protection IC may be suitable.
They are designed to isolate the load from the cell when the voltage is too low, to prevent over-discharge and damage.
They reconnect the load when the voltage increases to a suitable level - typically something above 2V.

As they run from the cell, they also tend to have low standby current.

This is an example of a basic, minimal function one:
http://www.unisonic.com.tw/datasheet/UB2421.pdf

Or a more advanced one that also has an interface for an MCU to configure it and monitor the power level - and it appears to have even lower standby current draw, just one microamp.
http://www.farnell.com/datasheets/2...MIxd7r7_jb3QIVyrTtCh1bfgn5EAAYASAAEgI2efD_BwE

[For info, a tiny lithium cell may be more effective as supercaps often have poor leakage characteristics. It all depends how much power your system will put in to the storage device].
 
Hi rjenkinsgb,

I plan to use a Lithium Ion Super capacitor, They are a hybrid of sorts.

Found here: https://uk.farnell.com/vinatech/vlcrs3r8406mg/cap-40f-3-8v-lithium-ion/dp/2846646

The li ion protection IC does sound like something that will do the job quite well, I shall look into this further. Thanks

The main reason for choosing the supercap over a battery, is due to the lifespan of the battery and charging characteristics. A supercap can take energy in the form of a charge so chemistry doesn't come into it.

Ultimately i may be wrong, but i suppose that all part of the fun.
 
Hi Dr Pepper,
I havnt written any code as of yet.
I didnt want to rely on any code to monitor this voltage as at times there may not be any power to power the uC. Im not sure how it would respond to this sort of power supply.
This is all experimental for me i guess so plenty to play about with.
 
OK, that is something new! Looks to be an interesting device & very good characteristics.
(And tiny for the rated capacity).

Yep, fairly new and yeah tiny for capacity. Farnell has them up to 240F...
I think i might have enough to work with now, i shall draw up a schematic and post it up when im done.
I hope it will be fairly efficient for what its doing.
 
I'd still like to know if it's a shivering mouse or an earthquake that's the source of energy.

Mike.
 
I would suggest you first find out how much power your 'generator' provides, I imagine it will be VERY small - and will probably take a considerable time before there's enough energy stored to do anything at all, assuming that there's no leakage from the capacitor?.

I would also suggest you try googling 'beam bots', which is (was?) an old style of very simple robot, using the principles that you seem to be looking for?. The usual active device was a 1381 voltage detector, which is probably long obsolete now?.

Essentially they wait until the capacitor has charged enough, then use the stored energy to move - and keep repeating this action.

The beam bots though use very little energy, your tracker will almost certainly need a LOT more, and as I've already mentioned, you need to see what your generator can provide.
 
Not relevent now I spose, however code could be used here, moreover microcontroller functions, most have a brownout reset, you could set this up on certain chips to only bring the thing out of reset when the power rail is over a certain value.
 
Well, wireless industrial buttons have been available for a few years that only use the energy from the button being pressed to transmit a signal - so it's not that impractical..

There's not much call for a tracker that only works 'up to 25m away', and I would suggest there's a lot more energy available from a sturdy button press than some kind of 'vibration generator' - but as I said, he needs to first find out what amount of energy is available under the circumstances of it's use, and how often he hopes to be able to transmit.

'Presumably', as he mentioned 'tracker', he's talking about GSM/GPRS, which is pretty power hungry, and VERY touchy about the capabilities of it's PSU.

It's funny how everyone jumps on the name 'super capacitor', where they are generally far less useful than a decent battery - and may well have a shorter life time as well. The introduction of super capacitors in VCR's for clock/memory backup proved shorter lived than the NiCd cells they replaced - but it was anice little money spinner replacing them
 
With a 40 Farad cap, there will be a lot of power available, if the generator parts works.

At that it's 40 amp-seconds per volt, to charge it or discharge it..
 
With a 40 Farad cap, there will be a lot of power available, if the generator parts works.

Only if that amount of energy has been stored in the capacitor, and it all depends on what the generator is capable of - which I suspect is VERY little. Hence my suggestion that finding out should be the first part of any such project.

As always in these types of threads it's all guess work, but presumably he's wanting to place a tracker on someone, or something, and hoping to power it be movements and vibrations.
 
Not relevent now I spose, however code could be used here, moreover microcontroller functions, most have a brownout reset, you could set this up on certain chips to only bring the thing out of reset when the power rail is over a certain value.

This could be useful, i dont have much experience with these functions, but will look at how they may help.



The tracker aspect isnt yet determined, i would love to use a GPRS tracker, possibly with GLONASS built in, but as you say they are power hungry. Another option that seems to be coming in quite well is LoRa. This is low power (dont have any exact figures) and long range. The downside is it requires a network of gateways. These are being created by other companies of which you can use for a fee.

With regard to the super cap vs battery, are you suggesting that a small Li Ion cell, will out live that of a supercap?
I have a bunch of 501225 cells which are almost the same size as the proposed cap, but i felt that charging it up from a trickle source like a vibration generator just wouldnt work.
Maybe i need to do some more research


I'd still like to know if it's a shivering mouse or an earthquake that's the source of energy.

Mike.

It could be either. Depends on where this thing eventually goes. Earthquakes as a source of energy, while massive, doesnt fair well for the population. A shivering mouse may be a little on the weak side.

Intended vibration source at this time would be while mounted to a car/engine etc, or a pet dog, horse that sort of thing.

At this point its just a project with zero commercial value.
 
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…