Supercapacitor circuit design help please!

[scnoob]

Hi all, this is my first post to a forum, ever, so please excuse any faux-pas I make.  (constructive criticism is most welcome!)

I've got an electronics technician level of training (over 10 years ago so I am a little rusty), and am a little over my head with this personal project I've taken on.  I am building an automotive booster pack out of some supercapactiors I've purchased.  I've attached a JPG of my circuit diagram, as well as the TinyCAD file from which the JPG was created.

First of all, sorry for the poor layout of this circuit diagram!  It started out simple and clean enough but got out of hand as I discovered more and more problems that needed resolution.  I'm looking for help in finding design flaws, incorrect part implementation, and any other issues you may come across before I go ahead and buy the parts to put together a working unit.  Any recommended part substitutions are welcome.  I would be very interested in design simplifications as well...the fewer parts the better!  As long as it doesn't compromise any of the design goals.  Hopefully none of the magic smoke escapes when I first apply power!

Design goals:
1. Prevent overcharging (and damage) of each individual supercapacitor (balancing is not really a concern)
2. Eliminate parasitic supercapacitor loads to maintain charge as long as possible while in storage
3. Charge supercapacitors at a safe rate
4. Prevent boosted load from backfeeding supercapacitors into overcharge condition
5. Protect against booster cables hooked up backwards

Questions:
1. Do I have the + and - inputs on the comparators reversed?
2. Will negative voltage from reversed booster cables fry the comparators or anything else?
3. Will connecting together the Supply and Output grounds of the up/down converter blow it up?
4. Have I correctly wired up the MOSFETs?  Are the Gate resistors necessary/correct?
5. Will the main power relay flicker due to variable load voltage while starting the boosted car?
6. Does anyone know where I can get a 14V fixed linear LDO voltage regulator?  I don't want to have to use an adjustable one.  I don't see any on Digikey, which is where I will be purchasing the parts I don't already have.
7. Have I chosen the correct parts for my application?

Notes:
1. All unlabeled diodes are standard IN4007.
2. 3400F capacitors are Maxwell BCAP3400.
3. The Up/Down converter I am considering can be found here: http://www.ebay.ca/itm/400597430280
4. The solar panel and SLA battery are removable.
5. The SLA battery will have an independant maintenance charger connected to it on occasion.
6. The relay connecting the supercapacitors to the booster cables is rated for 14V 700A. (12V coil)  I already have it.
7. I will of course be using appropriately sized wiring.
8. Ignore the pinout numbers on the comparators...they're set in TinyCad and I can't see how to remove them.
9. Red LED part # LTL-4263
10. Green LED part # LTL-307G
11. Yellow LED part # LTL-307Y
12. Blue LED = 10mA, unknown voltage...I already have it, I'll figure it out and adjust the resistor accordingly.
13. VoltMeter Relay part # HE128-ND
14. All low power MOSFETs are part # FDU3N40TU
15. All high power MOSFETs are part # TK40E06N1,S1X

[Psi]

When faced with the same problem i said, "screw it" and put the caps into the car without any protection, balancing or SLA 

[max_torque]

If supercaps were any good as passenger car "batteries", well, they'd be used for passenger car batteries............

[SeanB]

If supercaps were any good as passenger car "batteries", well, they'd be used for passenger car batteries............

Lead is cheaper to make, store and recycle, plus it has been around for over a century as a battery in traction use. Supercaps are still a lot more per cell, even with the lower capacity and mass.

I was looking to get some, but for the price I can just buy 10 car batteries instead.

[scnoob]

When faced with the same problem i said, "screw it" and put the caps into the car without any protection, balancing or SLA 

Haha!  Considering the difficulties I have run into while designing this circuit, I have been tempted to do the same.  Out of curiosity, have you ever measured the voltage across each individual capacitor?  I'd be curious to know how they are doing. 

Since I'm using only 5 of the 2.85V caps instead of 6 of the 2.7V, I only have 14.25V to play with instead of 16.2, and I've even dropped that down to 14 for insurance.  Since the alternator tends to put out more than my 14 volts, I have to build in some form of safety features.

I'm not actually installing them directly in the car anyways, I plan to use them as a portable booster pack on the farm for the various tractors etc. that get rarely used and inevitably have a dead battery when you need them.  I've gone through several of the commercial booster packs, and they just don't last.  They also don't have the guts to crank a big diesel, especially if the weather is getting a little cold.

[SeanB]

Get the battery booster Photonic Induction tried out ( look for the channel on YT) and get one of those. Should start the diesel tractor at least once.

[Trees]

I can't help you with your circuit, but I give the following for what it's worth..



If you're just storing the booster pack on the shelf for weeks or months at a time, it will last longer the lower the voltage is (voltage and/or heat is what ultimately kills ultracaps). You really do not want more than 2.5 V on each cell for long term storage, 2 V or less would be even better. They might reach an acceptable voltage level via natural self discharge, or it might be worth forcing it to happen. Or you could add an extra cell or two to your system so the volts per cell is never too high to begin with. Maxwell cells have a fairly high self discharge rate so you might be ok without putting any effort in. The problem would be if the cells had different self-discharge rates, and then you charged them all up (some cells would be over-charged). But to emphasize the point, you definitely do not want to maintain constant 2.85 V per cell for long periods of time, especially not with a Maxwell cell.


You can't really hurt the caps themselves by charging them fast, the only thing that might be damaged would be your power supply (including if that is a battery) or cables etc if they are not rated to the power or current that you are using. Even charging at 400 A would not be too much for those caps, you'd just need a little time at constant voltage to finish the charging.



The converter on the ebay link has a 3A max input current. Since it takes 100s of A to crank an engine, isn't that a problem? Or was that part related to a solar panel trickle charge feed..? 



I've thought about doing this myself, and what I would do is what scnoob did Keep it simple. I'd just occasionally check the voltage on each cell to make sure everything was about right. Using some small resistors to maintain balance isn't a bad idea, but so long as you don't have too much voltage per cap to start with, this kind of system should have some abuse tolerance. Also, 3400 F is a lot of cap per cell, so even with some degradation over time the unit should still be capable of cranking an engine so long as the cell resistances don't get too high.

[Psi]

Haha!  Considering the difficulties I have run into while designing this circuit, I have been tempted to do the same.  Out of curiosity, have you ever measured the voltage across each individual capacitor?  I'd be curious to know how they are doing. 

Yes I wondered if balance would be an issue too so I checked it a few times.

After 1 month in the car the biggest difference was <5mV
Checked it again 6months later and biggest difference was around 20mV
Last time I checked was maybe 1.5 years after install and one cap was  70mV out. I should check again  now (2 years)

I do remember that 4 caps were within ~20mv at the 1.5 year check.
It was only 2 of them that were getting a bit far out compared to the others

The self discharge gets much worse as you get closer to there rated voltage.
So higher charged caps tend to self discharge more which helps keep some balance.

My caps fall from 2.5v to 2.3v within a month but after 6 months they are still above 2v

[Psi]

If supercaps were any good as passenger car "batteries", well, they'd be used for passenger car batteries............

Nope, why would a car maker fork out extra money for super caps that would last the life of the car when people are happy to pay for a new battery every 4 years.

I got my caps 2ndhand for 2.5x the cost of a new batt

[mikerj]

If supercaps were any good as passenger car "batteries", well, they'd be used for passenger car batteries............

Nope, why would a car maker fork out extra money for super caps that would last the life of the car when people are happy to pay for a new battery every 4 years.

I got my caps 2ndhand for 2.5x the cost of a new batt

They wouldn't put supercaps in because they just aren't equivalent to a lead acid battery, as well as being expensive.

[Psi]

Most people wouldn't know the difference with supercaps alone. (except for alarm systems which are overly power hungry)
If you left your lights on the car would go flat 20x faster. but that's about it.

Ideally the car maker would have a smaller battery as well.
Car would isolate the supercap to keep maybe 5seconds of cranking available even if the user runs the main battery flat.


They could expand this protection from a flat battery even further.

Add a very small supercap (so cheap), just enough to provide cranking current for one or two engine starts.
Because the cap is small it wouldn't take much energy to charge up from a very small solar panel array.
Quick calc says just 4 of the newer 125x125mm sunpower cells could produce enough energy to start a car in ~10min under full sun.

[scnoob]

I can't help you with your circuit, but I give the following for what it's worth..

If you're just storing the booster pack on the shelf for weeks or months at a time, it will last longer the lower the voltage is (voltage and/or heat is what ultimately kills ultracaps). You really do not want more than 2.5 V on each cell for long term storage, 2 V or less would be even better. They might reach an acceptable voltage level via natural self discharge, or it might be worth forcing it to happen. Or you could add an extra cell or two to your system so the volts per cell is never too high to begin with. Maxwell cells have a fairly high self discharge rate so you might be ok without putting any effort in. The problem would be if the cells had different self-discharge rates, and then you charged them all up (some cells would be over-charged). But to emphasize the point, you definitely do not want to maintain constant 2.85 V per cell for long periods of time, especially not with a Maxwell cell.

You can't really hurt the caps themselves by charging them fast, the only thing that might be damaged would be your power supply (including if that is a battery) or cables etc if they are not rated to the power or current that you are using. Even charging at 400 A would not be too much for those caps, you'd just need a little time at constant voltage to finish the charging.

The converter on the ebay link has a 3A max input current. Since it takes 100s of A to crank an engine, isn't that a problem? Or was that part related to a solar panel trickle charge feed..? 

I've thought about doing this myself, and what I would do is what scnoob did Keep it simple. I'd just occasionally check the voltage on each cell to make sure everything was about right. Using some small resistors to maintain balance isn't a bad idea, but so long as you don't have too much voltage per cap to start with, this kind of system should have some abuse tolerance. Also, 3400 F is a lot of cap per cell, so even with some degradation over time the unit should still be capable of cranking an engine so long as the cell resistances don't get too high.

Thank you, that is valuable information!  I didn't realize that maintaining the capacitors at their full charge would be bad for them.  I did expect them to self-discharge with time, but so do lead-acid batteries to some extent.  I have a battery tester, which is basically a giant resistor with a volt meter attached that I could use to drain the caps when I don't foresee using them for a while.  When I read the datasheet,
http://www.maxwell.com/images/documents/k2_2_85v_ds_3000619en_1.pdf
it appears to me that they prefer being stored at full charge, as they list a 10 year life at full charge, but only 4 years when discharged...?  Maybe I am reading it wrong.

The up/down converter I plan to use is there solely to provide a steady, predictable power supply to the electronics in the circuit.  It does not feed any power to the capacitors or the load.  I will take some time tonight to make an annotated version of my circuit diagram, which will hopefully give some more insight into it's operation.

Thanks!!

[scnoob]

Yes I wondered if balance would be an issue too so I checked it a few times.

After 1 month in the car the biggest difference was <5mV
Checked it again 6months later and biggest difference was around 20mV
Last time I checked was maybe 1.5 years after install and one cap was  70mV out. I should check again  now (2 years)

I do remember that 4 caps were within ~20mv at the 1.5 year check.
It was only 2 of them that were getting a bit far out compared to the others

The self discharge gets much worse as you get closer to there rated voltage.
So higher charged caps tend to self discharge more which helps keep some balance.

My caps fall from 2.5v to 2.3v within a month but after 6 months they are still above 2v

I think I will give my bank a charge and just watch what happens.  There's a fair bit of manufacturing variance in the model I got, so I am interested to see how they do on their own. 

I think I am still stuck with having a protection circuit in my case, mainly because I'm not willing to spend the extra cash on another capacitor to put into the bank to up the voltage, which would ironically lower the overall capacitance at the same time

I don't like the idea of frying my $$$ supercaps haha!  Plus I like the challenge of working out the circuit problem

[max_torque]

If supercaps were any good as passenger car "batteries", well, they'd be used for passenger car batteries............

Nope, why would a car maker fork out extra money for super caps that would last the life of the car when people are happy to pay for a new battery every 4 years.

I got my caps 2ndhand for 2.5x the cost of a new batt

Off course, silly me, i forgot than all the OEMs are interested in is making sure their cars fail so they can get some more of our money.  yes that must be it.

Hint.  I design and develop cars for a living, i have done so for 25 years, and have worked for just about every car maker in the world at one time or another.  And the total number of times i have been asked to design a component to "fail" so that the owner must buy a new one:  ZERO TIMES.

The reason OEMS use chemical energy storage is because the energy density is sufficient to meet there requirements, super caps do not meet them.  It's as simple as that.



PS:If you'd bought a 2nd hand normal battery that would have been even cheaper than your cheap 2nd caps btw!

[scnoob]

Just FYI, I have added some notes to the circuit diagram and attached them so that hopefully it will be a little easier to work with.

Also a note on the charging...if the charge source voltage is below 14v, the supercaps will never gain a full charge.  They will come up to the source voltage (minus the diode drop) and just sit there.  The Up/Down converter does NOT provide any power to charge the supercaps!

Thanks!

[scnoob]

I've noticed the circuit diagram has been downloaded numerous times, I was just wondering if there's anybody out there taking a good look at it?

I've seen a lot of debate about the merits of supercapacitors vs lead-acid batteries, and that's fine.  I feel they are separate technologies, each with their own benefits and pitfalls.  I have chosen supercapacitors as the technology that best fits my application.  If you don't agree with me on that stance, that's fine too.

What I came to this forum for is help with circuit design, and to benefit from a broader knowledge base than my own.  Any help would be much appreciated, even if it's only a single answer to one of the questions in the opening post.

Thanks!!

[Psi]

If supercaps were any good as passenger car "batteries", well, they'd be used for passenger car batteries............

Nope, why would a car maker fork out extra money for super caps that would last the life of the car when people are happy to pay for a new battery every 4 years.

I got my caps 2ndhand for 2.5x the cost of a new batt

Off course, silly me, i forgot than all the OEMs are interested in is making sure their cars fail so they can get some more of our money.  yes that must be it.

Hint.  I design and develop cars for a living, i have done so for 25 years, and have worked for just about every car maker in the world at one time or another.  And the total number of times i have been asked to design a component to "fail" so that the owner must buy a new one:  ZERO TIMES.

The reason OEMS use chemical energy storage is because the energy density is sufficient to meet there requirements, super caps do not meet them.  It's as simple as that.



PS:If you'd bought a 2nd hand normal battery that would have been even cheaper than your cheap 2nd caps btw!

I wasn't suggesting they design them to fail. Just that its not economical to save the customer money at the cost of the car makers profits. And this is why they don't do it, not from a technical perspective.
In fact it works great.
I should know, I've had only supercaps in my car for 2+ years

[eas]

If supercaps were any good as passenger car "batteries", well, they'd be used for passenger car batteries............

Did that make sense to you when you posted it...in an electrical engineering forum.

[Psi]

Reminds me of getting car its WOF cert.
-: Sir its a legal requirement that your battery is mounted securely.
-: I dont have a battery, that's a capacitor.

[SeanB]

Funny thing is the motor vehicle registration paper here has a block for power source, petrol, diesel, electric, nuclear, other.......

[yash101]

This might be slightly off-topic.
A supercap holds like NO energy. However, it's capable of releasing the energy it has stored faster than 88 MPH. This will allow the alternator to start the car really fast.
I think that in the ideal situation, a car should be built with a lead-acid or LiPO4 battery to hold the actual energy and a supercap to actually start the car. I'm sure that you could just place a small lead acid battery parallel to your booster pack.

I'm sure that in this setup, a tiny (5AH) battery should be plenty as long as you don't use your car without turning it on. It'll keep the cap bank topped up. However, the capacitor would do the actual starting. A current-limiting resistor would be used to charge up the supercap without exploding the battery!

[Psi]

A supercap holds like NO energy.

It's not that bad, they hold a fair bit. It's just quite a lot less than a battery.

A supercap the physical size of a car battery holds the equivalent of ~1.5Ah
The capacity remaining once the cap is below 6V is a bit useless from a car starting perspective, so lets say 1Ah of useful capacity.

A typical car battery is like 30Ah,  so 1/30 or ~4%

[tggzzz]

I'm a bit late entering this conversation, but it is good to see someone "thinking" before "doing".

Your first port-of-call should be the manufacturer's data sheets, application notes, and any other forms of training available (including wiki/blogs/users etc). Personally I do choose manufacturers based on the quality of those; it is at least as important as the quality of their product.

The better the literature, the more the manufacturer understands and has characterised their product, and the more they understand that their success depends on your success.

[Circlotron]

I think that in the ideal situation, a car should be built with a lead-acid or LiPO4 battery to hold the actual energy and a supercap to actually start the car. I'm sure that you could just place a small lead acid battery parallel to your booster pack.

I just put the supercaps across the existing battery. The trouble is now though, it will crank over just as normal as can be (actually quite a bit better) till the the day the battery has deteriorated to the point that it has only several seconds of cranking left in it. Then it will stick me up somewhere.

[SeanB]

Just buy one of the ultra cheap voltmeters off ebay and put it in. When a lead acid battery dies one cell will go dead first, and the sudden drop to 10V before cranking ( will still start with the supercap boost) will tell you to change the battery before the second cell dies and the battery drops to 8V.. Most cars will still start at 8V across the battery terminals, but at 6V they will not.