Maxwell Ultra-Cap's/Lithium cells hybrid pack
 

Hey guys, just here to talk about putting Maxwell ultra capacitors in with a 5s2p A123 pack. If you do not know the consequences of this setup, it would essentially reduce, or almost relieve voltage droop of a battery, and would take all the peak loading away from the battery. The cap. back would weight in at 480 grams and has a peak voltage of 20 volts at 38.75 farads, which equals to 5275 Joules of energy. This 'type' of power in combination with the battery discharge capability, the output power of this hybrid pack would be phenomenal. It would also get rid of erratic voltage fluctuations...

The voltage under hard acceleration would stay almost at nominal state (16.5v) and would have very, VERY flat discharge curve (that is bursting power). You would have to use 8 cells in series for 20 volts peak (2.5 volts/cell), capacity goes down by C/Sc

The specs for the cap is:

Capacity: 310 farads
Voltage: 2.5 volts
Current: 1500 amps
Length: 62mm
Diameter: 33mm


There is also smaller cells,

120 farads
140 farads
and a 350 farads capacitors

It would be interesting to see how this works. I agree that it would all but eliminate the instantaneous voltage sag providing that the heavy current draw isn't continuous. I have a couple of questions though;

- What are the physical dimensions? If it is too big, it may have a hard time fitting in some already cramped areas.

- How do you charge them? A capacitor that size (even 500,00uF) should NOT be simply slapped on the batteries as the instaneous current would be all but infinite, only limited by the battery capacity. As a matter of fact, it would seem like a dead short until the cap charge level increased. Most of the time, large caps are charged via a series connected resistor. The time it would take to charge is 5*R*C. Whatever charging procedure is used must be used whenever a fresh set of batteries are installed.

In car audio, the general rule of thumb is to use 1F for every 500-1000watts (rms). Of course more is better, but there is a bang-for-the-buck aspect. I would think that 2-3F would be more than sufficient and be much smaller, lighter, and easier to install.

Also, there is an added benefit to the use of a cap; runtime may actually go up a little. Since there won't be large pulses of discharge current, the battery will actually deliver the full Ah rating of the battery.

Yes, you are very correct in your notings. Thats the thing, with bashing the power duty is bursting.

First, the size is overall the same size as my 5s2p A123 pack, although the weight is a little over half the weight of the battery pack.

As for charging, i have my BK Precision power supply that can bring it up to the voltage of the pack at any current rate it set it at, it's completely programmable. (If i simply connected the M1 pack to the caps, the tabs would disappear ;) Then, i simply couple it in with the battery pack, and put a very short connector between it and the controller, ~.5 inch, and maybe shorten up the controller input lead length, to completely get rid of voltage spikes due to inductance/EMF.

I would be able to charge with the caps coupled to the pack.

And yes, i forgot to mention it, you would have longer runtime (if the small added weight would not offset) because of the Ah retention. Most likely longer cell life too, maybe not enough to justify the Cap's cost in the first place; but the caps are rated at 500,000 duty cycles. :)

Im thinking the smaller 120 farad caps would do the job too, and the bank would weight only 232 grams. Each cell is 51mm long, and 26mm diameter. This cap is rated at 500 amps.

I really think this would provide very high power levels, very, how do you say... like hooking up a 12V car battery up to your model. Let's say, very 'ROBUST PERFORMANCE'. I've always wanted to try this.

Oh, BTW, Brian. How do you like the Spektrum radios, i have been wanting to get one of those for a while. Just would like to know how it feels?

Believe me, I'm not trying to argue against using them, I'm just trying to see if there is a feasible way to use them.

That said, you said you can charge with your bench power supply at whatever current you set. OK, that's constant current mode. However batteries are constant voltage sources. So, if I charge the cap up and then install it in my R/C, great, all will work fine. But once the batts finally dump, the batts and the cap will have a lower voltage. So, you stick a new pack in there and connect them and BAM! Those poor Deans will have a nice black mark where the arc hit. I guess I'm thinking more in terms of Mr R/C Joe who just hooks stuff up and goes. I don't think he'll be bringing along a multi-thousand dollar P/S. :)

A way to get around this would be to permanently wire one of those caps to each pack. That way, they are charged at the same time and discharged at the same time, so there will never be a difference in potential. The only time you'll have to be careful is the initial hookup when you build the pack/cap combo.

This would certainly help battery life! All you would need is a battery that can supply the average current required by your setup since all spikes would be all but nonexistant!

Also, no more worries about max C, pulsed C, etc - only the constant draw would be the issue, which is quite low. For example, my Revo with 14XGP3300 cells gets about 13-14 minutes of runtime. I'm sure it pulls 100A bursts (or more), but the average draw is only ~14A. That means a lot more batteries can be used where they couldn't before due to their low max discharge performance, even though their capacity was great (like GP4300's).

Oh, and I like my Spektrum a lot. The receiver is small and fits anywhere. No glitches ever. I only wish the transmitter would have more model memories. If I had to do it over again, I'd get a Nomadio and still use the Spektrum receivers where small size is needed (the Nomadio receivers are kinda big IMO) simply because it can store more models.

Yeah, i sooo want to try this. About the setup, ya if you have more than one pack, the deans connector would disappear if using A123 cells, and would be very dangerous using LiPo's. But, i only have one pack, that can be charged in 25 minutes ;) (the BK power supply is $330) So, for me i would have the pack and caps permanently hooked together, well not really, just a deans but they would never be disconnected.

So, would you know the duration of the 100+ amp bursts in a brushless setup like Quark and an XL motor? Just so i can do more calcs.

This would be a kick in the face for the motor! :)

Also, the self discharge of the 120 farad cap is .15mA which is VERY low, no worries about it discharging the battery pack anytime soon...

Yeah, if you are set up for that kind of thing and understand the possible risks if not done right, it should work great. Doesn't seem to be something that will be available for the masses anytime soon though.

The bursts doen't last very long; basically however long it takes for the motor to reach the rpm of the signal it's being fed, which is mostly on take-off out of the hole, and any hard acceleration. I'm sure there are a couple graphs of an EagleTree plot around here somewhere.

150uA delf discharge; not bad. According to my calculations, it would take about 3.5 years to totally drain a 4.6Ah pack not including the self-discharge of the battery itself.

All of the bursts I have recorded using my Eagle Tree MicroPower have been more of "spikes", where the high current point is maintained for something like 1/10th of a second, and the current tapers off on either side of the high point. The total burst might occur over a second or two, depending on how I use the throttle before and after the spike.

Yah, so preferably getting the smallest pack possible to take care of the spikes and bursts. Enough to keep the voltage up at nominal.

However, even a small voltage 'difference' from nominal voltage would equate to large currents, again, like you said, a cap of this size is basically a short circuit...

So you say 1/10 th of a second eh? :) That sounds good then...

Do you think 2300 Joules is enough, that is a 8s 140 farad (17.5 F.) cap bank. @ 232 grams with 500 amp

Or the larger one at 480 grams, 8s 310 farad (38.75 F.) cap bank with 5300 Joules with 1500 amp

2300 joules enough?

a 5S 8000mA battery delivers a total of 532800 joules and at 50C there is 62000 joules of available energy per second. (at 3.1V per cell)

I am in the process of testing 58F version... :)

D-cell to compare

Artur

What is your opinion on them?

Did you test them a bit yet? i see you shortened them, i pressume where empty.. ? :p

They seem to have a relative high internal resistance?

Not that bad, specs can be downloaded here

Very promising...
I was tracking them for long time and happy that you can get one. Hopefully later their would be wider selection.
In RC application lower internal resistance would be more important the capacitance, but power ratings should be still high.

Preliminary testing, shows regular capacitor behavior, but full-time testing bed is under construction. Stock E-Maxx-Hacker-MambaMaxx-12 cheap cells-eagletree...

Artur

But with the relative high internal resistance, they won't deliver too much current, but i am looking forward to the test results. Can you please post them if you have them?

Thanks!

I will, here simple simulation I did for 25A/75A transition circuit and graph
Over discharge batteries internal resistance goes up, so caps help will be more noticeable and with low-end batteries as well. At least in theory :)

Artur

Serum, 62,000 Joules per second! LOL You must have done calc error, that battery pack can deliver up to 3000 Joules continuously (per second) @ ....3000 watt (apparently) 62,000 Joules per second would be around 3300 amps

An RC car would/could never use this much power, it's more instantaneous power pulses. For those that don't quite understand, the capacitors energy is directly proportional to the instantaneous voltage of the cap, not taking into account the very small voltage drop between the two terminals.

energy formula: 1/2 (CV) C - capacity in farads, V - voltage in volts

Any questions or comments welcomed

Sorry, my bad. it does 6100 joule. (6100 watts in bursts)

total amount of energy is 532800 joule.

I personally don't see the advantage of these cells, their internal resistance is too high, and with the right cells (read good quality) you wouldn't need them (which was my point) they would add nothing but weight.

Once they reached a lower internal resistance they will be more interesting in replacing batteries, with the ability of very quick charging.

This cells will never replace batteries. They will boost batteries performance and replace them in applications with short-term power requirements.

Tens of mOhms is pretty low, not every battery can have it.

Self-discharge rate kill them as batteries
Charge-Discharge cycles rating put them ahead of batteries
Also wider operational temperature range...

Artur

But what is the use them in rc vehicles if you have got 6000 watts of peak power in your cells? (proper lipo cells, 5S 8000 mAh)

It doesn't need this boost, it will only add weight. You are going to test them with cheap cells, for that they can work miracles, i agree. but for good cells?

Let's say the benfits are:

-More consistent power (very little voltage droop under burst loading) the controller would always be getting the same voltage...
--Longer runtime (capacity retention because of less strain)
-Safer, especially for LiPo
-Longer cycle life
-If power input leads shortened to esc, much better for the controller...


I can say this, BUT nothing will compare to real world testing.

True,

But they can extend batteries life and help ESC.
Is it worth the added weight?

Artur

What do they weight then? I think it would be wiser to gain more cells (with the same weight) and choose a lower KV motor, to simple unload the controller/motor/leads from the high current and gain real runtime instead. (total available joule)

@zero;

i don't know what you think of this, or if you knew it at all; but a flighpower pack delivers 7V under a 100A load for their 2S 4900 mAh pack.. I don't think a more consistent cell is available at this time.. only dropping .4-.5V under a 700 watt load.

420g

It is ~5mOhms and at room temp...

Artur

420g is quite some stress for the motor/drivetrain/speedo to toss around.

I would prefer a lower kv motor, with a higher cellcount for sure! It will turn out more efficient as well. (and the added runtime is nice too)

If you ran 4S lipo, you can get an 8S setup, (gaining the same weight) double the runtime and roughly halve the current.

the 140 Farad caps weight 29 grams each, you need 8 cells.

the 310 Farad caps weight 60 grams each, you need 8 cells.

Serum, putting more batts in the car is completely irrelevant, we are talking about a specific battery, compared...

Those Flightpower batts sound super sweet, however i will never buy a Lithium Polymer pack ever. They are too dangerous, can't be charged fast, and need special attention. And they cost alot more than A123 cells. Im sticking to A123 cells for now, until AltairNano comes to market with their Nanosafe battery.

You are talking about taking stress from the motor/speedo, and all of a sudden my sugestion is completely irrelevant? what is your point/goal??

I am suggesting something different, by adding cells, get the same weight and gain runtime, AND run more efficient.

The capacitor idea is nice, providing the weight and size could be reduced. If the caps are the same size/weight (or more) as simply adding a parallel set of batteries, you'd most likely get more bang for the buck with the added batteries. This is why I said use something like a 2-5F cap; it still will offer substantial help with the high current pulses while remaining relatively small and lightweight. Even 1F will help a lot for 1,000 watt loads.

5 mOhm is actually not bad. That's only a 0.5v drop at 100A. Don't forget, that's just the voltage drop of the cap. The batteries will reduce that 0.5v drop to a lesser value because they are in parallel.

Also, if these caps are hooked in series, it add the voltage values, but halves the capacity. So, if you took two 2.5v 10F caps and hooked them in series, you'd get a single 5v 5F cap. You put 10 of those in series, and you get a 25v cap rated at "only" 0.5F.

I am absolutely sure using these is good for the batteries and provides a more stable voltage with substantially less sag at high current, but when used in an R/C, it's going to become a compromise of cap/battery performance vs total size and weight.

ya, it's all a matter of energy density for the caps. a group of MIT scientists have made ultra caps with energy density of 60Wh/kg, NICE.

Better yet EStor has a capacitor with if im not mistaken, 120 Wh/kg? Don't quote me on that, i'll have to look it up again.

For my setup, 232 grams for a 38.75 Farad, 20 volts cap bank is pretty fair. i've emaile Maxwell Technologies for a price quote on the caps.

I think it will be cheaper to add a few cells..

Perhaps you can request a sample..

I did this with a few ballancing IC's

Yes, you can 'always' add more cells, but thats not the point. This is a redundant add-on, it will take away all the strain from the batt, WHATEVER the battery pack may be. THeoretically it will pay for it's self.

-Cells will last alot longer
-safer for esc
-better performance
-guaranteed rated battery cell life (you will never get rated cell life in RC apps)
-it's more of a one time buy, they last 500,000 duty cycles, they are not wearing out anytime soon, they will be there new pack after pack...

The future of energy storage will most likely have to do with capacitor type storage.

And only then they are interesting, this is what i said in a previous post. (fast charging, fast discharging, long lifetime)

All the possitive things you name are also a fact with adding more cells, but the biggest point is that you will gain more than double the runtime..

A reason for me NOT to use them (in case you didn't noticed my opinion on them)

you think that capacitors would really help?... a long time ago.. I use to use the 10, 000 uF caps and I didn't really notice any difference... they where 1" x 1" x 1"

AQWUT, you were using 'capacitors', you had '10,000uF'

I am talking about 'ultra-capacitors', let say '140,000,000uF' for the same size.

QUITE a big difference, they are basically super concentrated capacitors, they are double layered(electrode plates), and have a few other tricks.

and they are heavy!

230 grams for 8 cell pack?! I wouldn't call that heavy!

:031: what happened to the 480 gramms you talked about in your opening post.?