Ultra Capacitors

[injin man]

Does anyone here have a working knowledge of these
devices. I'm going to use them to smooth out surge loads
on the Inverter. There are 12 of them in series for 30 volts.
The Ultracaps are 3000 Farad each but it works out to be less
in series, they should deliver extra power when needed.
My system is 24v nom, the Battery is on the way, it's a 24V
Nickel Iron Battery built by Exide.

I need to know what to make the connections with between
the cells.

[Mad_Labs]

Sorry I don't have an answer for ya, I've only ever used supercaps ~ 1 F. Your bank would be 250 F! One thing I do know is to derate the voltage specs considerably, if that cap bank fails it ain't gonna be pretty. What I want to know is where did you get these beasts and how much? I'm thinking rail gun / gauss cannon extrodinaire!

Jonathan

[mike90045]

> I'm going to use them to smooth out surge loads on the Inverter.

No, you need "ripple" caps, to reduce the 120Hz that the batteries otherwise have to supply  (peak & valley of the 60hz)  under high loads.

[injin man]

Mike

These are being used as they are intended, these are not caps in the same sense. These will act
like high performance batteries, except they can accept charge rates many times faster the LA
Batts. They use these to 'boost' assist Diesel engines. They also drive the motors in Buses and
Hybrid Cars.

Check out these guys, they are the manufacturuer of the ones I got:http://www.maxwell.com/

Also Ebay has a good selection to choose from. The ones I got came from Buses, they are rated
a million cycles.

[injin man]

Here is a good clip on Ultra caps:http://www.youtube.com/watch?v=s4B1aViSKgc

[injin man]

So, no one here has looked at these recently? It's NOT a new technology
but it appears to have made some significant progress in the last 2 years.

You will have to ignore the media hype and stick to the facts about these,
I've seen some reviews that are little hard to swallow, but the fact is they
work in some applications and not in others. Small power to me looks promising
and I'll share what I find here.

[mike90045]

...., the Battery is on the way, it's a 24V  Nickel Iron Battery built by Exide.

I need to know what to make the connections with between the cells. 

Between the battery cells, or the caps ?

I've had to develop a adapter terminal from a standard crimp lug, to the M20 bold on my NiFe batteries.

My batteries came with nickel plated steel links (shown laid in place, not bolted in yet)

pics
http://tinyurl.com/LMR-BigLug
http://tinyurl.com/LMR-NiFe

[injin man]

Mike

I need to connect the Caps together in series, although I've seen some
put together no one says what the flat bar is, I'm assuming Aluminum
would be a good medium?

The NiFe Cells are connected together with about a 3/8" round nickel
plated copper rod. I got some spares to fool around with and some extra
connectors so that shouldn't be a problem.

[mike90045]

.....I'm assuming Aluminum would be a good medium? ...

For starting a fire - yes.

Aluminum develops an insulating oxide coating in seconds after it is cut. It's BAD for electrical conductor, unless you use wire alloys for wire, with anti-ox goop and high pressure split nuts.

I've seen tinned copper strips, or buss bars used to wire cap banks with. Even large circuit board with thru holes for screw terminal capacitors, and the screw connects to copper trace.

But forget trying to use aluminum.

[DanG]

Wow, such negativity  

You realize 250F at 28.8 volts is 103680 joules... completely released in one second is 103700 watts or 140~ horsepower?

Transient power boost to the inverter should cover the first 40-80 milliseconds. I think you've gone a little overboard  

I gloomed on to my own supercap for my planned 48V inverter too - its out of a bus also. Only weighs 38 kilograms; 2.33F at 270V rated so at 57.6V it is 3865 joules - 3900W or 5HP for the one second discharge. I feel safer with with huge voltage rating being used at 1/5 voltage capacity - I think* I can protect it from anything but a direct lightning strike. I would not want to be you with 250F and inducted surges from nearby lightning!  \\\\shiver\\\\

For aluminum construction design...

I've maintained 700V 1000A cap bank series-parallel networks... Yank forty twenty~pound modules, steam clean them and replace. Works well and lasts long time. There should be a doomsday fast acting fuse on the cap string itself. The cap bank should have some sort of blast enclosure. You should also have a discharge resistor built in to cross the bank at all times - 3 or 5 watt 500k to megaohm values. Look it up and do the math. If you start a fire, blow out the windows, or send a wrench off at 1/10th lightspeed shorting out that homemade electromagnetic railgun its not my fault.

The inrush current for supercaps will be extreme - to bring them online you need to limit charging by adding a 24V load with a known constant current - an incandescent lamp might work, when the lamp dims out one would manually switch it out of the circuit. Taking three of five minutes to initially charge the cap string is not a bad thing since its constant duty afterwards.

A series cap balancing scheme probably should be used especially if the caps are random pick and not matched.

There are aluminum contact tabs already on capacitors... Aluminum would work if you follow industry practices. Buss bar must be heavied up at bolt area to keep long term creep - deflection movement at a minimum; an oversize cable mount pad with anti-ox paste required, washers under both machine thread bolts & nuts and precision torque. Note that a single bad connection can have a higher resistance than the internal resistance of the supercap.

With conductivity twice that of copper, even with the reduction of alloying away from pure aluminum, Al is still an improvement - but physical size increase is required to get the weight up to half or more that of copper for a safe & equal conductor. A two-thirds cross section increase over copper is probably minimum.

http://www.stormcopper.com/design/Buss-Bar-Ampacities.htm is a good source for data - and purchasing if you've got deep pockets - but using Aluminum may only give a 35% discount from copper once all the design constraints are added in.

http://www.glyn.com.au/downloads/documents/Maxwell/ultracap_product_guide.pdf

Sounds like a great project - though there are pre-fab modules from decommissioned systems out there that would've been turn-key.

[injin man]

Wow, such negativity  

You realize 250F at 28.8 volts is 103680 joules... completely released in one second is 103700 watts or 140~ horsepower?

Transient power boost to the inverter should cover the first 40-80 milliseconds. I think you've gone a little overboard  

I gloomed on to my own supercap for my planned 48V inverter too - its out of a bus also. Only weighs 38 kilograms; 2.33F at 270V rated so at 57.6V it is 3865 joules - 3900W or 5HP for the one second discharge. I feel safer with with huge voltage rating being used at 1/5 voltage capacity - I think* I can protect it from anything but a direct lightning strike. I would not want to be you with 250F and inducted surges from nearby lightning!  \\\\shiver\\\\

For aluminum construction design...

I've maintained 700V 1000A cap bank series-parallel networks... Yank forty twenty~pound modules, steam clean them and replace. Works well and lasts long time. There should be a doomsday fast acting fuse on the cap string itself. The cap bank should have some sort of blast enclosure. You should also have a discharge resistor built in to cross the bank at all times - 3 or 5 watt 500k to megaohm values. Look it up and do the math. If you start a fire, blow out the windows, or send a wrench off at 1/10th lightspeed shorting out that homemade electromagnetic railgun its not my fault.

The inrush current for supercaps will be extreme - to bring them online you need to limit charging by adding a 24V load with a known constant current - an incandescent lamp might work, when the lamp dims out one would manually switch it out of the circuit. Taking three of five minutes to initially charge the cap string is not a bad thing since its constant duty afterwards.

A series cap balancing scheme probably should be used especially if the caps are random pick and not matched.

There are aluminum contact tabs already on capacitors... Aluminum would work if you follow industry practices. Buss bar must be heavied up at bolt area to keep long term creep - deflection movement at a minimum; an oversize cable mount pad with anti-ox paste required, washers under both machine thread bolts & nuts and precision torque. Note that a single bad connection can have a higher resistance than the internal resistance of the supercap.

With conductivity twice that of copper, even with the reduction of alloying away from pure aluminum, Al is still an improvement - but physical size increase is required to get the weight up to half or more that of copper for a safe & equal conductor. A two-thirds cross section increase over copper is probably minimum.

http://www.stormcopper.com/design/Buss-Bar-Ampacities.htm is a good source for data - and purchasing if you've got deep pockets - but using Aluminum may only give a 35% discount from copper once all the design constraints are added in.

http://www.glyn.com.au/downloads/documents/Maxwell/ultracap_product_guide.pdf

Sounds like a great project - though there are pre-fab modules from decommissioned systems out there that would've been turn-key.

The guy I got these from apparently works on Buses that are being run with these caps! If I'm not mistaken these would have to have someplace for the power to go
in order to discharge them in one second, I've seen video of them welding with them for instance. They appear to act like high energy batteries rather than wizz bang
capacitors at high voltage. You are right however that they do store a lot of power in this arangement. I'll ask the guy about the connections and se what he recommends.

[mike90045]

With conductivity twice that of copper, even with the reduction of alloying away from pure aluminum, Al is still an improvement -

That's what conductivity?
 Not thermal, not electrical.  (per volume)  Silver is the only better thermal & electrical conductor than copper.  I use all three, and understand all 3 quite well, and all I can say is there's a typo in there.

Every thermal and electrical conductivity table lists silver first, then copper. (of mostly plain alloys)  gold and then aluminum follows.

http://www.tibtech.com/conductivity.php

[DanG]

spot on, I'm corrected and you found the memory leak
Electrical conductivity is 65% that of copper, but 210% by weight.
Just trying to write too much too fast and tossed the baby out with the bath water.

[injin man]

Dan

Thanks for the Maxwell link, I'm going to go thru it tonight.
That looks like an Aussie link? didn't see that particular set
of docs on their site, but may have simply missed it.

Correct me if I'm wrong here, but these are being used now
to replace batteries and don't/won't dump their charge all
at once like a true capacitor. The output I'm seeking to counteract
is the droop of a large load on the battery. I'm no EE but it
should help?

[injin man]

Dan

This is from the Maxwell white paper:

Since ultracapacitors have low internal
impedance they are capable of delivering
high currents and are often times placed in
parallel with batteries to load level the
batteries, extending battery life. The
ultracapacitor buffers the battery from
seeing the high peak currents experienced
in the application. This methodology is
employed for devices such as digital
cameras, hybrid drive systems and
regenerative braking (for energy recapture).