Super capacitors to save battery life?

lwik

100 mW
Joined
Sep 14, 2016
Messages
46
Hi, this is more of a question about it being possible. I searched the forum before posting but didn't find the answer. I do a lot of eBike touring and in March I rode the southern tier 3166 miles across America. Last year I have about 5500 miles of touring. I have 2 52v 15AH batteries (lithium) which I use to help my 475 pound right, me and the fully loaded bike up hills and out run dogs.

So 1 of my batteries started to cut out / BMS safety prematurely when I would open up the throttle at around 49v. The battery has about 300 charge cycles and is now at the manufacture to get repaired. Is it possible to add just enough super capacitors to capture regenerative breaking and also supply that initial boost of power? Going down some of the larger hills and in cities coming to a fast stop because of cars or traffic light I use the region to slow me down. Going down some of these large grades I see my CA3 pulling over 500 watts of regen which I know is not good for batteries. Having some type of super capacitor as a buffer between the battery and controller might save my batteries.

Is there any information out there or ideas for this concept? I bike tour a lot and love it but blowing through a battery in 250 cycles / $1K battery is not fun. Thats about $4 a charge. I included a couple of photos of the trip including the 300 watt solar panel I use to charge in the middle of nowhere. : )

Thank you.
 

Attachments

  • IMG_9832.jpg
    IMG_9832.jpg
    124.5 KB · Views: 5,471
  • IMG_8312.jpg
    IMG_8312.jpg
    140.6 KB · Views: 5,471
lwik said:
Is it possible to add just enough super capacitors to capture regenerative breaking and also supply that initial boost of power?
Short answer - no. Putting supercaps in parallel will not significantly change things, since energy storage is related to voltage change and there won't be much voltage change.

If you wanted to get fancy and add a dual-link system (i.e. your ultracaps power the motor, and a controlled-current switcher links that to the battery) maybe. But 500 watts regen is nothing - it's 10 amps into a 50 volt battery and that's less than a .5C charge into two 15ah batteries, which the battery should not have any problem tolerating.

I'd fix the battery first.
 
your batteries are wrong for the use case. if you charge up to 4.1v and cut off at 3.2 you should get thousands of cycles, not dozens. get the proper cells for the task at hand and make a battery with the proper capacity. a battery within spec is a happy battery.

in order to help you answer these for starters:

what is your peak draw (sub 10s) and what is your average draw in amps? (note that voltage is complely irrelevant here)
what is your desired battery capacity? what do you want your battery to do for you?


do you have a custom battery/drive system or factory stuff? if you have factory i would recommend gutting the battery pack and put in a high end aftermaket programmable BMS you can read out on the road and reprogram at will and have very high end cells rated for the job.

you should not do these kinds of trips with chinese crap cells or a dumb as a bag of sand bms installed. especially on the go you need more.

there are plenty of places that can build a custom battery for your use case. if you already have custom batteries you should get your money back.
 
Short answer: Maybe and sort of.

I've done it, but using high rate LiPo rather than Super Capacitors.

I have a 40Ah bank of LiFePo4, which is really old. Really sags badly, but still maintains over half its overall capacity. Rather than ditch it, I "stiffened" it up with some a small amount of high rate LiPo. What it does is:

1. On hard acceleration, it has the lower resistance, so it provides most of the power. The LiPo lowers its voltage through consumption of power, while the LiFePo4 provides very little power, so its lowered voltage comes through load based voltage drop. When the load stops, the LiFePo4 bounces back to it's "resting" voltage which is now higher than the the LiPo's resting voltage. Large inrush current. 2V difference, 50mOhm battery, boom, 40A current.

2. On regen, same thing happens in reverse. Having lower resistance, LiPo sucks up the power first as it has lower resistance. LiFePo4 gets very little, but its "apparent" voltage is high while the charge is on. When the regen charge current goes off, the LiPo feeds big power into the LiFePo4. Thankfully, (a) regen is short, so we're not talking a lot of power, (b) LiFePo4 is big yeah yeah yeah, LiFePo4's not small, no no no. Sorry distracted. I mean to say that a small battery feeding unregulated power into a big battery is generally not a big deal, because 40A into a 40Ah battery is only 1C, the other way around, 40A into a 16Ah LiPo battery is nearly 3C.

Problem with super capacitors is that when you put them in series, they lose capacity.

So a 500 Farad 2.7v Capacitor is... 500 Farads. (Thanks, Captain Obvious)
Two 500 Farad 2.7v Capacitors for 5.4v is 250 Farads
Three 500 Farad 2.7v Capacitors for 8.1v is 166 Farads.

You see where this is going right?

If you want near super capacitor performance, try LTO, something like this:

Ewt-lithium-Titanate-battery-18650-14450-26650.png_350x350.png


http://www.alibaba.com/product-detail/Ewt-lithium-Titanate-battery-18650-14450_60677808770.html?spm=a2700.7724838.2017115.123.451b583b0tzByg

These can charge and discharge around 20C (get one of the bigger ones and we're talking 50Amps), and have multi-thousand cycle life.

If you've got money to burn, these babies are the Bugatti Veyrons of batteries, with the price and rarity to match. If you have to ask how much, you can't afford it. If you can afford it, you probably can't find one to buy.

cell-img03.png


100A with no voltage drop:

cell-graph04.png


Charged to 80% in 1 minute:

cell-graph03.png


Sorry. I get excited about cool tech. I just saw they released a 10Ah version of that cell. One cell can discharge at 1500w. That would make 48v pack capable of discharging 30kw. That's just insanity. If Toshiba had a retail outlet for this, I'd be all:

takemymoney.jpg
 
These are old stock lto, but price is right.
Lots of buyers use for super cap replacement.
https://endless-sphere.com/forums/viewtopic.php?f=31&t=91735

s-l400-8.jpg
 
billvon said:
lwik said:
Is it possible to add just enough super capacitors to capture regenerative breaking and also supply that initial boost of power?

If you wanted to get fancy and add a dual-link system (i.e. your ultracaps power the motor, and a controlled-current switcher links that to the battery) maybe. But 500 watts regen is nothing - it's 10 amps into a 50 volt battery and that's less than a .5C charge into two 15ah batteries, which the battery should not have any problem tolerating.
Thanks for the reply Bill. My battery is in LA getting repaired a the ebike shop I bought it at. Over the xmas break we are riding LA to Las Vegas and I will need it back. I was trying to do some more research on the dual link systems you were mentioning? Would that be an automated switching between the battery and Ultra Capacitors for regen? Thanks!

@flippy
This is the battery that I have http://www.hi-powercycles.com/new-52v-15ah-hpc-elite-usa-made-battery-system/
These guys are amazing and have been very helpful and sell high quality gear. They have built up a couple bikes for me that went 3200 miles across the US with almost no problem. The motor is a 2000w Crystalyte and the Controller is a Phaserunner from Grin and CA3 computers with throttles. As far as using bad Chinese gear I went out of my way to buy good gear. Sold my house in Santa Barbara and built a $10K touring rig because I didn't want problems. Photo is of the Phaserunner mounted on my girlfriends bike, mine is the same. I use 2 Satiators from Grin to charge and have a custom charge setting that cutoff at 57.6 and not 58.6. But some spots in the desert between towns I would top the battery off since I wasn't sure if I would reach the next town on that charge.

With all the research I have done, I understand that it is bad to fully charge and discharge the batteries. Also putting extra strain on the battery buy demanding a lot of instant power and also using the regen to slow the bike down. Like riding through a town and giving it throttle at the stop light and then using the region to help slow it down at the next light. That constant charge and discharge can be hard on the batteries from what I have read and with the battery failing at 49 volts is kind of proof that it is hard on the batteries.

Also when the battery was new, and I would do a full charge to 58.6v (before I started to limit the charge to 57.6). I could see it slowly drop from 58.6 to 58.4 and so on. Now if it charge to 58.6 and I ride a 1/4 mile the battery voltage is at 57.6. Maybe it is a bad cell that is causing this battery to fail or BMS, we'll find out.

@sunder

I am interested in your system? So I am wondering if you did the something I am thinking of doing because of it effecting your batteries? I was looking at replacing the batteries with super capacitors but after a few hours of research that became clear it was not a possibility. That's when creating a buffer that I could use to regen and boost became the idea. I have seen 2.85V 3400F and 6 of those would be 17.1v and 566f which I think is 56WH. If that is boosted up to say 42V maybe that would produce 22WH, these are just rough numbers converting watts, amps. The motor runs from 36-72 volts, if I can boost up to that then that might work? I might be talking out my but here :) I am not an electronic engineer. I program for the most part. But if this is something feasible then I would be interested in testing it out.

Thanks for the replies!
 

Attachments

  • IMG_7817.jpg
    IMG_7817.jpg
    89.7 KB · Views: 5,408
lwik said:
Thanks for the reply Bill. My battery is in LA getting repaired a the ebike shop I bought it at. Over the xmas break we are riding LA to Las Vegas and I will need it back. I was trying to do some more research on the dual link systems you were mentioning? Would that be an automated switching between the battery and Ultra Capacitors for regen?
No.

Think of it this way - your "regular" system (battery + motor controller/inverter + motor) has the battery replaced by an ultracap. So now the only battery you have is the ultracap. Limitless power but limited energy; you can get to full speed very quickly, go about half a mile and then run out of energy.

Then you add a second system which has a buck/boost that draws up to X amps from another battery (like your lithium ion) and uses that to replenish the ultracap.

So let's say you have a 36V ultracap + motor system. You then get a 48V lithium ion battery, use a 48V to 36V (actually 42V) buck converter that's current limited to 20 amps. Then the lithium ion battery never sees more than a 20 amp draw.

This, of course, is cumbersome and expensive, which is why I don't think it's a great idea. But it accomplishes what you want, which is using the ultracap for power (and regen) and the li-ion battery for energy storage.
 
lwik said:
@sunder

I am interested in your system? So I am wondering if you did the something I am thinking of doing because of it effecting your batteries? I was looking at replacing the batteries with super capacitors but after a few hours of research that became clear it was not a possibility. That's when creating a buffer that I could use to regen and boost became the idea. I have seen 2.85V 3400F and 6 of those would be 17.1v and 566f which I think is 56WH. If that is boosted up to say 42V maybe that would produce 22WH, these are just rough numbers converting watts, amps. The motor runs from 36-72 volts, if I can boost up to that then that might work? I might be talking out my but here :) I am not an electronic engineer. I program for the most part. But if this is something feasible then I would be interested in testing it out.

Thanks for the replies!

Yes, I used high rate LiPo cells as a "super capacitor", because unlike super capacitors, you don't lose energy capacity when serialising LiPo cells.

I'm not understanding why you would want to serialise super caps to 17.1v (losing a huge amount of capacity), then boost further. Boost converters aren't all that efficient, and until you go very expensive, most boost converters aren't going to handle surge very well. You're back to square one.

Bill's solution is "technically pure", but in my opinion, unnecessarily complex and inefficient. As I said, even good boost converters only convert at 90%, and unless you get a big one, it may struggle to keep your super capacitors full and at high voltage. You don't need a buck or boost converter to regulate the energy transfer between your super cap side and your battery side. If you design well, we're talking very small voltage differences, and therefore manageable current transfers.

If I were you, I would buy Inwo's LTO pack. For just $150 + delivery and a bit of work on your part, It is rated to deliver 7.7kw for 30 seconds, which triples your current battery's capability. It would also add 60% to your range. Unfortunately it also adds 16.5lbs to your rig.

But speaking of this strategy. You mentioned that you had two of those batteries. If you run them in parallel, you in effect halve the load on each battery both during regen and heavy take off.

And since I love "Money to burn" options... 7 of these in series, would give you 30kw capabilities, and an additional 30% extra range, for 7lbs extra weight:

https://hobbyking.com/en_us/turnigy-nano-tech-ultimate-6400mah-2s2p-90c-hardcase-lipo-square-pack.html
 
When you work the numbers on it, for the same amount of weight, volume and cost you add to the vehicle, more battery vs more capacitor makes for longer life as well as potentially better range for the times you need to use all of it.

Awesome trip! I've been thinking about trying to hit the south pole on a fatbike with a portable solar array. Did you have much trouble with wind wanting to blow the panels around? Were you happy with the performance and durability of those units?
 
Not following the lost capacity in series connected capacitors.
Energy storage is the same. Series or parallel.
Look at each capacitor discretely.
Can be charged to full capacity. 500 Farad @ 2.7v.
 
If you have a 1,000F 2.7v cap, and you put 10 in series you have a 100F 27v capacitor.
 
Sunder said:
lwik said:
@sunder

I am interested in your system? So I am wondering if you did the something I am thinking of doing because of it effecting your batteries? I was looking at replacing the batteries with super capacitors but after a few hours of research that became clear it was not a possibility. That's when creating a buffer that I could use to regen and boost became the idea. I have seen 2.85V 3400F and 6 of those would be 17.1v and 566f which I think is 56WH. If that is boosted up to say 42V maybe that would produce 22WH, these are just rough numbers converting watts, amps. The motor runs from 36-72 volts, if I can boost up to that then that might work? I might be talking out my but here :) I am not an electronic engineer. I program for the most part. But if this is something feasible then I would be interested in testing it out.

Thanks for the replies!

Yes, I used high rate LiPo cells as a "super capacitor", because unlike super capacitors, you don't lose energy capacity when serialising LiPo cells.

you dont loose it (energy stored) with super caps either - you loose capacity (in Ah or whatever unit you're using) but the stored Kwh/Joules etc goes up linearly whether you ad caps in series or parallel. Look up the formula for energy in a cap - its relation to its voltage is square, not linear. So while capacity (Ah) halves, the voltage doubles, and given energy stored (kwh/J) is proportional to capacity*voltage*voltage... you end up doubling the energy stored, either way (paralleling or series'ing). Edit - worth noting that this doesn't account for losses, so while energy stored is theoretically the same, there's some losses to account for that make it a loosing game... tldr the basic theory says 2x the capacity, but the usable? not quite so much.

That said, still not a good idea imo. If you want the cheap option, using some high C rate cells in parallel should help a great deal. Just make sure your peak power draw/regen is not beyond the capability of that high rate pack on its own, because it'll be doing 90+% of the work.

A better option is building a new pack (or getting yours repaired as it may have a dud cell) capable of delivering/absorbing the power levels you demand. Keep well within its rated C rates and capacity, and it'll last far more than 250 cycles - likely more than 500, even if its cheapo HK lipo. Even super caps dont have the energy density to justify their use (for the space they take up/ammount they weigh) compared to just adding some more batteries.
 
Sunder said:
Short answer: Maybe and sort of.

http://www.alibaba.com/product-detail/Ewt-lithium-Titanate-battery-18650-14450_60677808770.html?spm=a2700.7724838.2017115.123.451b583b0tzByg

These can charge and discharge around 20C (get one of the bigger ones and we're talking 50Amps), and have multi-thousand cycle life.

If you've got money to burn, these babies are the Bugatti Veyrons of batteries, with the price and rarity to match. If you have to ask how much, you can't afford it. If you can afford it, you probably can't find one to buy.

cell-img03.png


100A with no voltage drop:

cell-graph04.png


Charged to 80% in 1 minute:

cell-graph03.png


Sorry. I get excited about cool tech. I just saw they released a 10Ah version of that cell. One cell can discharge at 1500w. That would make 48v pack capable of discharging 30kw. That's just insanity. If Toshiba had a retail outlet for this, I'd be all:

takemymoney.jpg

So I just looked at this battery and it was saying had up to 15,000 charge cycles and it seems like it was also charge and discharge fast. That seems to be the better solution for me then the Super Capacitors. They seem to be sold through eBay but I am not sure I trust eBay. If it was eBay and other established retailers then I would be more inclined.

So I have toured from the Canadian boarder to the Mexican boarder. Then the pacific to the Atlantic ocean this year and in 2 years Im planning my around the world trip over 3 years. These batteries sound like the solution to what I am looking for and can take a beating. Thanks for this info!
 
Inwo said:
These are old stock lto, but price is right.
Lots of buyers use for super cap replacement.
https://endless-sphere.com/forums/viewtopic.php?f=31&t=91735

s-l400-8.jpg

Thanks for the Reply, I am just now looking in to and researching these LTO. I think the capacitor path is a bad bath at this point. The specs on LTO is amazing around 10K charge cycles depending. I looked in the other other poster who recommended the Toshiba SCIB but those don't seem to be sold to individuals. I found these guys with a 30AH LTO https://www.ev-power.eu/LTO-technology/Lithium-Titanate-Oxid-Battery-Cell-LTO-2-4V-30AH.html 21 of those would get me 50.4 nominal voltage or about 1500 WH. After reading I am not to sure what the differences for bike application that the SCIB batteries and the basic LTO like I found different that much? But I think your guys suggestion push this project in a different direction.
 
in you case i would recommend going to a specialist and get a custom battery made for your bike with regular high end 18650's.
form factor can be made exactly to spec instead of huge square blocks and under 2 inches wide if needed.
having the abillity to closely monitor and change settings of your battery and controller depending on the conditions (maximize range or power for example). lifespan is also considerably longer.
it will also be a lot cheaper then getting some unicorn batteries wich are impossible to service. if a single cell dies with monobloc you are SOL, with 18650s you just lose 10w in capacity if a cell dies.

if you need details for what components you need (like the programmable BMS) or the battery builder needs to use you can just ask. i have built dozens of custom batteries so far.
 
The Tesla car has an eight year warranty to provide at least 80% of the new battery range, so they are doing something right. Even if you are unsure about which brand and model of 18650 cell to use, you can actually buy Tesla cells from a wrecked Tesla. Plenty of guys are buying the whole pack and then splitting it up into batches of 100 to sell.

^That being said, adding super capacitors to a battery pack to help capture more of the downhill regen has been tried. You could say that...technically...it worked. But in the final analysis, the extra captured energy compared to the bulk, weight and cost of the super caps was a bad tradeoff.

I agree with Luke on this one, get a bigger battery, use quality cells, charge to 4.0V per cell like Tesla (or at the very least 4.1V)...if you charge to 4.2V and leave the pack charged up to full while you sleep, you are throwing away some of your packs potential life every day.

Also, a bigger battery will run cooler at the same amp-draw, and heat is another factor in eroding pack life...
 
tesla charges to 4.05v and if you enable range mode you get 4.1v. but you need to enable that every time you charge.
and why buy tesla cells when you can just buy NCR18650B's from most retailers?
 
I agree. I spent the last couple days doing lots of research and answers I have gotten back from you guys has pushed me in different directions. Super Capacitors are not the way to do this.

After a few people mentioned LTO batteries in this thread and reading up on it I don't understand why these are the standard? Fast charging and discharging, 10K or so cycles and can take more abuse then what people are using now 18650 Lithium Ion?

Thanks
 
lwik said:
..... I found these guys with a 30AH LTO https://www.ev-power.eu/LTO-technology/Lithium-Titanate-Oxid-Battery-Cell-LTO-2-4V-30AH.html 21 of those would get me 50.4 nominal voltage or about 1500 WH. After reading I am not to sure what the differences for bike application that the SCIB batteries and the basic LTO like I found different that much? But I think your guys suggestion push this project in a different direction.
One difference might be the "Energy Density" at 45Wh/kg.
21 of those cells would weigh 34 kg ! ...a little more than the 8kg total for your current 2 packs. :roll:
No problem for a car , but not ideal for a Ebike.
You may also find the size a little less convenient .
Lesson.:-...There is no free lunch ! :wink:
 
liveforphysics said:
Awesome trip! I've been thinking about trying to hit the south pole on a fatbike with a portable solar array. Did you have much trouble with wind wanting to blow the panels around? Were you happy with the performance and durability of those units?

Solar panels worked great and they lay flat on the ground once unfolded wind has never been an issue. I keep them in a custom bag and frame that I have attached to my front bike rack, you can see it on the first photo of the finish line of my trip attached to the bike. The 300w panels would produce about 275 watts around noon. They are good for bulk charging but take forever to top off the battery. So days off when camping I could easily charge both of my 52V 15 AH batteries. Plug one in in the morning and then at peak day I would swap them. Solar panels weight around 15 pounds and the aluminum rack is 8 lbs.

The panels with a DC/DC boost charger 24V in 58V out and the 300 watt panels were $2K, not cheap but I am not sorry I got them. It allows me to eBike in to remote areas and not worry about power. One photo show the top of the frame I built for the solar panel. I designed it to com to the exact hight of the rack to extend the surface space. Hope that info helps you.
 

Attachments

  • IMG_5749.jpg
    IMG_5749.jpg
    114 KB · Views: 5,237
LTO are sometimes called supercapacitors, because some models (mostly high end) behave more like capacitors than batteries, the capacitance issue aside.

I've seen some that can charge at 60C without damage. LiPo can discharge at that rate, but try to charge it at that rate and you're asking for a fire.

They also have very long lives. Some can get over 100k cycles on 80% discharge.

They can sit flat for years and still be good. So that to any other lithium and you'll puff.

They are more tolerant of overcharge, (much like an overspeced capacitor)

They have very flat discharge curves.

The down side is that they are expensive and heavy. But thats like capacitors too.
 
Sunder said:
LTO are sometimes called supercapacitors, because some models (mostly high end) behave more like capacitors than batteries, the capacitance issue aside.

I've seen some that can charge at 60C without damage. LiPo can discharge at that rate, but try to charge it at that rate and you're asking for a fire.

They also have very long lives. Some can get over 100k cycles on 80% discharge.

They can sit flat for years and still be good. So that to any other lithium and you'll puff.

They are more tolerant of overcharge, (much like an overspeced capacitor)

They have very flat discharge curves.

The down side is that they are expensive and heavy. But thats like capacitors too.
The weight, I kind of don't mind. My touring rig with me was around 475 lbs. The biggest obstacle when riding across country is charging. I would normally stop and a McDonalds for lunch and if I was camping the night before without a way to charge them I would have to sit in the Mcd's for 2 hours with a Satiator charger and a standard 6-7 amp charger.

If I could build a 20 cell 10 x 11ah to crate a 52v 22ah that weighed 32 pounds that would be fine. I am already hauling around 2 52v 15AG that weigh 16 pounds, another 16 on top of that is fine is fine.

I found this charger on this forum, https://endless-sphere.com/forums/viewtopic.php?f=14&t=71139 that seems to be compact and can handle 1800 watts. Even pulling in to a liquor store for a 20 minute break if I could add 600 watts during that time it would be perfect, if possible. But the biggest help would be knowing that I am not doing any damage, or much when I am slowing the bike rig down going down a 7% grade for 4 miles. The disc breaks heat up and this feature really helps. If this is all possible I will start ordering parts soon.

Thanks.
 
If you're talking about Inwo's cells, they can handle 100A.

Very safe for any charger you can think of, but not impossible to put that much out on regen braking when you are:

1. Carrying that 475lbs + you
2. Going high speed.

Regen braking from 110km/h on my eScooter which weighs about the same, I noticed 80A @ 150v for about a second or two. That was nerve racking. Of course, there is a velocity squared in that formula somewhere, so if you don't get anywhere near that speed, then it's a moot point.

Do you know how much power you are putting out through regen?
 
You should try this, to save battery life.

1) Actually pedal, 2) try to reduce weight, 3) get in an aerodynamic position like a lower tuck position or semi/full recumbant position
 
markz said:
You should try this, to save battery life.

1) Actually pedal, 2) try to reduce weight, 3) get in an aerodynamic position like a lower tuck position or semi/full recumbant position

1. I do peddle, always.
2. Lost 18 pounds on this last trip across country, 265 to 247.
3. Love my Surly Ogre mountain bike and enjoy sitting more upright, than laying on my balls. :)
 
Back
Top