supercap battery boosting

jimmyhackers

10 kW
Joined
May 11, 2015
Messages
600
i used to do this with regular caps on my old Pb lead acid batts.
with reasonable results. bats were less voltage saggy n acceleration was more punchy.

it seems a few people on the youtubes are doing it to lithium batts now.

is there any merit to doing this to lithiums? (mine seem saggy)
if so is there a formulae/ratio to get the right farad rating for your battery capacity?
 
no, its useless in real world applications.

this has been a topic of discussion for many years now and always with the same result: its stupid to do so.

its like sticking your hand out a window to help you slow down.

its basically never the battery that is the limiting factor. if you have the space and budget for supercaps its simpler and yields better results to just increase the battery size with the space you reserved for the caps. the motor/controller or even traction is more of a limit then the power delivery of the battery.
 
100F capacitor holds just 50mAh for lithium range 4,2 to 2,5V
but capable of high current drain

acceleration could not be noticed.. but there could be some results at getting more range.

The supercapacitor catches the high drains at acceleration due to very low internal resistance. result some conservation of lithium cell life attached to parallel capacitor, and helping pump those amps..

there is only one way to simple test supercap results at ebike, this is the dirty way to do it

Example.
for ebike if you have a 10S battery try to parallel two serial 100F/2.5V (result 2S 5V 50F super cap) to some 1S at battery array for test sample. (remember supercap same cell voltage at parallel moment)
Go for a ride from 100% charge until battery total level 20%.
Check 1S sample cells paralleled w supercaps. voltage should be higher than the other 9S blocks. Means the sample block saved some energy..

this 2 capacitors trick applied to 10S battery means will need 20 serial 100F capacitors
100F 2.5V capacitor cost same as a cell $1.5, so investiment is +$30

extra kms gain maybe only a few percentage.

for the price of $30 u get 20cells 2500mAh
adding it parallel to battery means a extra 5000mAh vs few mAh gain capacity for parallel capacitors with lithium cells..

PS: There is a low cost boost trick.. If you convert ebike 10S to 11S battery ,U will get better performance and feel more acceleration. For a 10S to 11S ebike battery convertion means arround extra 4P cells. $6 (plus new bms) for new extra punch and speed..

think about this
 
i was less thinking extra range.
more thinking better battery performance, potential increased battery lifespan.

more so in sharing the load of the highest peaks of current draw.

https://www.youtube.com/watch?v=_Kikri_Dmk4
in this video the concept is seemingly being used to good effect but with a lead acid batt and an inverter.
again though.....this is only to let the inverter not LVC from the intial "power on" load spike from the appliance.

i only really expect this to help in high load situations....like acceleration from standstill ect.

if anything the convertion loses from the battery having the charge the supercaps after each volatge sag....could mean less overall range.
 
Sorry, it does not work like that in practice.

If you have the space for supercaps you have the space for more batteries. More batteries gives you more everything so you dont need supercaps.
If you have problems with voltage sag you need to change your battery, not bolt on more crap to offset your wrong cell choice.

SLA batteries have no place in this, neither do inverters.
 
https://www.youtube.com/watch?v=NoD5BHOi0dQ

so this guys an idiot? (he kinda is, but most americans come across to me like that)

i get that more or better batteries would be better, and buying better/more batteries would also be more expensive.
however, thats not what im asking.....

its like im asking about wether using an umbrella would help my raincoat keep me dryer, and your telling me to get a better raincoat.

between more batts and a supercap array....the supercaps would be cheaper, and smaller and be transferable between all my 72v rides batteries.

i asked about this for lead acid batteries and on here long ago and again all i got was "nah get better batts" and "it wont work".
i ended up having to try it for myself to see that it was most definately worth it.

i get the feeling history is going to repeat itself.
 
jimmyhackers said:
https://www.youtube.com/watch?v=NoD5BHOi0dQ

so this guys an idiot? (he kinda is, but most americans come across to me like that)
Well, it's Youtube content, not engineering.

jimmyhackers said:
i get that more or better batteries would be better, and buying better/more batteries would also be more expensive.
however, thats not what im asking.....

its like im asking about wether using an umbrella would help my raincoat keep me dryer, and your telling me to get a better raincoat.
It's more like asking if a walking stick would help your raincoat keep you drier. They're largely unrelated devices for largely unrelated uses.

jimmyhackers said:
between more batts and a supercap array....the supercaps would be cheaper, and smaller and be transferable between all my 72v rides batteries.
Cheaper and smaller than what? Bricks? They have to provide some measurable benefit to enable such a comparison.

jimmyhackers said:
i asked about this for lead acid batteries and on here long ago and again all i got was "nah get better batts" and "it wont work".
i ended up having to try it for myself to see that it was most definately worth it.
There's no mechanism, let alone math, to support this claim. There are plenty of (recent) threads explaining this in detail.

The only serious technical discussion with regards to supercaps on EVs is for regen -- you should be able to understand why before diving deeper. If your controller limits regen to some fraction of discharge, then supercaps have no application.
 
Supercaps are probably the wrong tech to do what you're suggesting. High power density LTO might be better.

The Toshiba SCiBs configured for 48v can deliver 10kw for 10s at just 3kg. (6.6lbs).

Does that make sense on an eBike where the main battery is only 5kg and max draw is 1kw? Of course not. But on an eMotorcycle with a 60kg battery with aging or low power density cells, oh hell yes.
 
the benefit in lead acid was due to lessening the peukert effect.

essentially shaving off some of the peak load on the battery and offsetting it onto the capacitors.

this also meant for high rush current loads more initial current was able to be delivered.

the peukert effect is only really applicable to lead acid batteries....they say.

but im pretty sure it happens to a lesser effect in all battery technologies/chemistries.

looking at supecap prices.....im sure they are half a scam in themselves.... as by the time youve added enough up to get a high enough voltage, the capacitance is so low its close to regular caps.

i could just rack regular 80v caps in parrallel across the "battery in" untill i feel a difference in punch/torque and leave it at that.
 
Instead of expensive supercapacitors, use a relatively large, low C-rate battery to trickle charge a relatively small, high C-rate battery. Connect the high C-rate battery to the load.
 
advancedelectricbikes said:
Instead of expensive supercapacitors, use a relatively large, low C-rate battery to trickle charge a relatively small, high C-rate battery. Connect the high C-rate battery to the load.

Sir that's a bad recommendation!! never plug some battery to other battery charge port!

if main battery that U call " high C-rate battery" is discharging for example 30A at load this means the parallel "relatively large, low C-rate battery" is following this tendency at trying to put 30A charge to compensate main battery discharge.. maybe bad result..

need to do parallel at power lines after bms!

allways remember to have same voltage at two batteries when doin parallel connection
 
batteryGOLD said:
advancedelectricbikes said:
Instead of expensive supercapacitors, use a relatively large, low C-rate battery to trickle charge a relatively small, high C-rate battery. Connect the high C-rate battery to the load.

Sir that's a bad recommendation!! never plug some battery to other battery charge port!

if main battery that U call " high C-rate battery" is discharging for example 30A at load this means the parallel "relatively large, low C-rate battery" is following this tendency at trying to put 30A charge to compensate main battery discharge.. maybe bad result..

need to do parallel at power lines after bms!

allways remember to have same voltage at two batteries when doin parallel connection
example:

24v, 20ah, 3C <-----> 24v, 5ah, 100C -----> load
  • batteries are series-connected
  • direction of energy flow>>>>>>>>>>>>>>>

As the load is presented to the batteries, current flows from the 100C battery to the load. As the 100C battery is depleted, current flows from the 3C battery to the series-connected 100C battery.
 
Here is my real-world experience.

I have 27AH 36V battery (9S10P) of AliExpress HG2 cells (I made up the battery myself) 50amp BMS, on a Kelly controller KBS72101X running to a My1020 motor. This is on a child's quad.
At standard (no caps), under the initial full acceleration, I was seeing the volts drop from 41 (fully charged) down to 36v with my fat 80KG ass on its already considerable weight. Under full load, the V recovered to 38ish after a few seconds as I hit max speed.
Yes, more cells are on their way, but I also ordered two 50V 47,000 uF caps to test. At a cost of £15 for the pair, it was worth finding out. The total capacity of these caps is only about 0.1 seconds' worth of full power. BUT the idea was to help with that starting initial high draw.

The caps sit after the BMS and before the controller. They are outside of a battery box.
Now the initial drop is only to the 38V and recovers to 40v after a few seconds. So I'm getting 1 extra volt per cap.
This makes the quad accelerate better and reach a higher top speed.

I wasn't expecting this to work or make any difference, but having wired it in a way I can swap between them connected or not, it's noticeable with them added to the system.

edit - yes I could have paid WAY more for A graded HG2s, but at £1.7 a cell delivered, I am quite happy with the HG2 copies I have found.
 
advancedelectricbikes said:
example:

24v, 20ah, 3C <-----> 24v, 5ah, 100C -----> load
  • batteries are series-connected
  • direction of energy flow>>>>>>>>>>>>>>>

As the load is presented to the batteries, current flows from the 100C battery to the load. As the 100C battery is depleted, current flows from the 3C battery to the series-connected 100C battery.

Sir, U mean when U connect a 24V to other 24V battery main power lines is a series ? or a parallel?

saying direction of flow ">>>>>>>"
this maybe theoretical becomes a serial line B2 to B1 to load, directional flow

do you have any lab experiment to demonstrate your theory?
60Abattery supply a 500A 5Cbattery? 5C battery has less voltage drop , so maybe possible 3C supply a 100C battery at acceptably charge rate.. remember can't put example 30A charge for a 5P , 6A charge to each cell.. if only few minuts maybe possible.. or if using high charge rate cells
 
GoldenBunip said:
Here is my real-world experience.

I have 27AH 36V battery (9S10P) of AliExpress HG2 cells (I made up the battery myself) 50amp BMS, on a Kelly controller KBS72101X running to a My1020 motor. This is on a child's quad.
At standard (no caps), under the initial full acceleration, I was seeing the volts drop from 41 (fully charged) down to 36v with my fat 80KG ass on its already considerable weight. Under full load, the V recovered to 38ish after a few seconds as I hit max speed.
Yes, more cells are on their way, but I also ordered two 50V 47,000 uF caps to test. At a cost of £15 for the pair, it was worth finding out. The total capacity of these caps is only about 0.1 seconds' worth of full power. .....
Now the initial drop is only to the 38V and recovers to 40v after a few seconds. So I'm getting 1 extra volt per cap.
This makes the quad accelerate better and reach a higher top speed.
Are you sure its a 9s 10p pack ?
If so ??.. how do you get 41 v on a 9s pack ??....4.55v/cell ? :shock:
I have yet to do the math on how much energy those caps give discharging from from 41 (?) to 38v, but i have a feeling its not much..
Try checking again without “supercharging” your 9s pack beyond any useful max voltage , because beyond 38 v your HG2s are giving you nothing ...its just “surface” voltage....not energy...and a shorter , high risk, useful life.
 
got that the wrong way around. :shock: it's a 36V pack.
10 cells in series, so 42.5v is the absolute max charge, but I tend to take it to just over 41ish.

9 batteries in parallel at 3ah each cell, hence the 27AH capacity.
 
seems like there is merit to this. thanks for the info goldenbunip

its worth noting the output of a brushless esc is essential 3 phase ac. meaning over 3 times the pulses and potential load spikes on the battery compared to brushed escs etc.

i guess there is a point of diminishing returns. cost vs weight/size vs effectiveness.

i even wonder if supercaps are better in this respect as they are such low voltage that by the time ive made up one for a 72v battery. the cost would probably make regular 100v caps in parrallel a better option.
 
You can get 100v caps. Always make sure the voltage rating of the cap is way above the voltage of the system.
 
Back
Top