Hey everyone:
I am looking at the sudden drop in pack voltage when I start from a stop on my nice 15 degree slope streets in my neighborhood. This is the XM3500 with the 40AH TS cells which, from what I can find and also tell from my own measurements, they have about 5 mohm internal resistance each. This along with the wiring and copper straps is about 120mohm total resistance so about a 9V drop with 2C from the stack and takes away the accelerating/starting torque. I have found when my pack is below 50% SOC and I try to start from a dead stop going up the 15 degree slope, the torque cuts out and the bike sometimes shuts down (I assume from UVC).
So, I started thinking about putting a super capacitor in parallel with enough capacitance and low ESR to give me a few seconds of hold-up voltage to get started on these slopes in these situations (after a couple of seconds the bike speed is enough that the back EMF reduces the inrush current needs so it is fine and the capacitors are then discharged during this event and the full current is from the pack again). However, I figure I need 10F minimum and it has to go across 72V peak.
There are the Maxwell boostcaps but I would need a BMS on them and run a string broken up piecewise as you can not put one cap per cell since they are 2.5-2.7V and the cell can go to 3.65V during charge. Also, they are expensive ($50 each for the 650F and I would need 28 and then a BMS just for these).
So, I have found car audio boost caps that can run 30F each and they are designed (some) to take 24V with balancing done internally. They also have ESR ratings of 1.5mohm. I can get three of these across cells to keep 24V on each, have 3 of the 30F in series to get 10F, and the cost is about $100 each. Has anyone ever considered this and do you see anything wrong with this?
Thanks
Your money is better spent buying more storage...
Jeff K.
Indeed. Even the biggest audio caps around are only capable of running a motor for measures in the fraction of a second.
Money would be better spent on a small A123 pack to put in parallel with the existing one.
The author of this post isn't responsible for any injury, disability or dismemberment, death, financial loss, illness, addiction, hereditary disease, or any other undesirable consequence or general misfortune resulting from use of the "information" contai
You could buy these http://www.dontronics-shop.com/cap-1000uf-35vx5.html at a good price. You would have to connect one per string of 7 batts. It wouldn't change your requirements for a BMS.
It would boost your range a bit by countering Peukert's effect, and keep the peak demand loads of the battery down when accelerating, which might help your batteries survive more cycles.
I tried once connecting smaller caps to my XM-3000 to see if it would boost my acceleration- it didn't. The batts can supply the peak power requirements, the motor just isn't wound to draw so much current. It would help in situations when the batteries can't provide peak power.
It would be an interesting experiment to try. But I'm leaning towards the other Jeff's answer. Better to spend your money on more lithium.
-=Jeff=-
XM-3000...
-DC-DC converter replaced with a Dell D220P-01 power supply.
-72V mod
-Expensive bank charger until I come up with something better... Still trying.
-
You'd need a LOT of those caps. Since a Farad of capacitance is the ability to deliver 1A for one second, you'd need 100,000 of those capacitors to even run a 100A motor for one second. Plus, since their voltage drops as their stored charge does, only about 1/4 of that is going to be at a useful voltage.
The author of this post isn't responsible for any injury, disability or dismemberment, death, financial loss, illness, addiction, hereditary disease, or any other undesirable consequence or general misfortune resulting from use of the "information" contai
1 coulomb is the amount of electric charge transported by a current of 1 ampere in 1 second.
1F*35V=35 coulombs. = 35A for one second. Your numbers look way Link, did I miscalculate?
The major voltage drops will be during acceleration for most of the discharge curve, although towards the end the caps won't help.
XM-3000...
-DC-DC converter replaced with a Dell D220P-01 power supply.
-72V mod
-Expensive bank charger until I come up with something better... Still trying.
-
I am actually after having a lower ESR drop when first accelerating from a dead stop on a slope where there is 80A in-rush current. I only want this for about a second or two to help boost the stall torque.
Buying more batteries does not lower the ESR significantly. Putting more batteries in series would help but then I am pushing the stack voltage up and exceeding the controller and motor voltage ratings (plus charger). So then I have to buy all new components (money).
The nominal running speed and torque are fine so I don't care once I have started moving. Thus, I want a smaller ESR for a couple of seconds only and this is what a capacitor does well.
P.S. sparc5 - linkofhyrule is correct. The formula is I=C*dV/dt, so you get a dV over a period of dt as I/C, so 1A/1F is 1V per second. The situation here with ESR considered makes it more complex so this helps me to relax my requirement down to 10F total across the whole stack.
Green electric power and use thereof; what more do we need?
You might be able to get the problem fixed for nothing by employing a skilful manoeuvre instead of skilful soldering:
Head downhill for a second first, then hang the tightest turn you can manage and keep going uphill! With a bit of practice you can make it look like it's just for fun...
This information may be used entirely at your own risk.
There is always a way if there is no other way!
:)
Out of curiosity, is this the norm for electric bikes to have difficulty starting up a steep slope when the SOC is low? Being an electrical engineer is makes sense now but not something I considered when buying. An ICE bike can handle this since you have slip in the auto tramission mode or clutch playing in manual so you can rev the motor to have torque and then supply it to the wheels without (hopefully) stalling the motor. However, with electric hub motors and the DC/DC motor driver you are the mercy of the electronics.
Green electric power and use thereof; what more do we need?
Green electric power and use thereof; what more do we need?
Most electrical motors are very inefficient at 0-a few RPM. Getting started up a hill using the motor will waste a lot of energy and produce a lot of heat. On top of all the torque going up a hill normally uses. But at least use your pedals to get moving before throttling, and start with a half throttle before going full.
Anyways, there are super capacitors with some crazy capacity. Digikey.com for example has some in the 100-1000 Farad range. Been considering them to spread out my regenerative braking so I don't have to waste so much energy in dynamic braking. Also to save the batteries. Let's say you have 100 farads at 100v. You can then feed that back into your system 10amps / second for 10 seconds instead of just pouring it in. It'd take a custom controller though. And I'm not sure how much that much capacitance would cost.
Just a quick look at mouser...
90 v and 100 F of capacitance would cost like $750. Not worth it at all just for more efficient regen or bursting.
Edit: Here is a better deal, but still not worth it IMO
50 F 2.3 V $5.25 @ purchase of 50
That would be $262.50 for 50F at 115V, or 5750 Watt Seconds (1.6 Watt Hours)
Edit 2: Just did some calculations, to stop a 450 lb vehicle (+person) traveling 30 mph it would take 18252, so to get full regen braking capabilities with that it would be like $750. Although with bulk comes significant discount, so maybe it would be more like $600.
No, 1F*35V means nothing, since the voltage declines as the cap is discharged, and it's the F alone that dictates the number of electrons in the cap and therefore how many it can supply for how long. The voltage adds its energy by giving some electrical potential to those electrons.
A 1F 35V cap can deliver 1A for one second and be charged up to a maximum of 35V. They can't be related properly for total energy storage without the E = 1/2C * V^2 formula.
The author of this post isn't responsible for any injury, disability or dismemberment, death, financial loss, illness, addiction, hereditary disease, or any other undesirable consequence or general misfortune resulting from use of the "information" contai
I see my error now, thanks for pointing it out.
Did you see on Tracy's site he's posted instructions on how to enable regen on the XM-3000, XM-3500. Thanks Tracy if you're reading this. The part where he says "0k lol" makes me think this is the spot to enable a PWM or digital potentiometer scheme to control the amount of regen. Once you get a feel for the scooter you don't need econ mode anymore, so a flick of that switch to turn the throttle into a break would be a very sweet mod.
Any chance you professional EE's want to take a stab at what sized inductor you'd want to balance 12V batteries?
XM-3000...
-DC-DC converter replaced with a Dell D220P-01 power supply.
-72V mod
-Expensive bank charger until I come up with something better... Still trying.
-
Hey sparc5 - I would consider 5-10mH with an oscillator frequency of a couple hundred kHz.
Interesting instructions on the regen mod. I don't think I will try this until I have a backup controller ;)
Green electric power and use thereof; what more do we need?
The problem is that if the hill is 15 degrees, or 27 percent, it is not an ordinary hill. That would be the kind of hill where sidewalks have to be stairs. A 50cc gas scooter would probably have trouble starting up such a hill too.
Another comment/question. I think those "copper" straps that came with my Thundersky cells are actually cartridge brass, and may be a lot more resistive than copper. I was tempted to use short pieces of 8 gauge stranded wire and soldered-on ring terminals - but it looked like a lot of work.
Np. ;)
Fechter also came up with an interesting method of regen: a brushed controller. By hooking up a brushed controller backwards (with diodes for protection) and activating it, the way the PWM collapses the motor's field induces a high-voltage spike every time a cycle is completed, effectively boosting motor voltage above the batteries regardless of RPM.
A 3-phase rectifier could be used for brushless motors.
Linky:
http://endless-sphere.com/forums/viewtopic.php?f=2&t=583&st=0&sk=t&sd=a&hilit=fechters+regen+mod
The author of this post isn't responsible for any injury, disability or dismemberment, death, financial loss, illness, addiction, hereditary disease, or any other undesirable consequence or general misfortune resulting from use of the "information" contai
I went out and remeasured and the slope is 14.4degree/25 percent and it is a steep hill. The sidewalks are not steps but from what I understand if the slope is much more they would have to be (steps become a requirement somewhere around 30 percent).
Now that you mention it, I have a neighbor down a ways with a 49CC scooter and they always start out going down the hill to the bottom before turning around to come back up (and they don't go very fast going up). When I have a good SOC, I can get started going up but just barely. If I am up to speed before the hill I end up slowing down to about 20MPH which seems pretty good considering (this is with the new 21 cell setup I have).
Green electric power and use thereof; what more do we need?
So how does this E = 1/2C * V^2 work?
E is joules, I assume. What is C... coulombs? And V voltage? Are coulombs farads * volts?
So an array of super caps that's 50 F 100 V.... would that be 500 coulombs? So 250 * 10000 = 2.5 megajoules? Did I do that right? Seems like I got something wrong, that would be 692 WH.
I think that when 50 caps are string in series the total capacitance is 50 less .
Therefore the 50 50 FARADS Cap in series will give you 1 Farad at 125 volts (2.5 V X 50)
It is possible that I incorrectly remenber the formula for capacitor or was it resistance or inductor ?
HEAT & LOSS & INEFFICIENCY RELATED DIRECTLY to IMPEDANCE
Capacitors in parallel are an increase in farads, capacitors in series are a decrease I think... doesn't make sense why that is though, why would you lose capacity by putting them in series? I guess with the increase in voltage though which exponentially increases that part of the formula, you are still gaining total energy stored even if putting them in series.
So if you have 50 in 50F in series, that is 1/total farads = 50*(1/50)??? So it is essentially 1F of 115 volts, using the source I had above.
That means that for the $260 example above....
Energy = 1F*.5*13225 = 6.6 kj = 1.84 wh
1.84 wh for $260.
Edit My math was way off, I was using coulombs for C but I should have been using farads. I edited this post to correct the math, so 1.84 wh is the correct answer I believe.
Just wanted to add one more thing as to why supercaps aren't ready for regenerative braking: It takes 10 seconds to fully charge a super cap. Since braking should happen in a shorter time (4 seconds?), you will need more capacity (double or so) for power storage than you would for energy storage. So let's say you want to put 6 wh into your cap during regenerative braking. You would probably need over capacitors of over 12 wh storage capability.
A Farad is the amount of capacitance that results in a 1V change in potential when 1 coulomb of stored electrons.
This does, in fact, mean that V * F = C (where C is coulombs). Rereading my other post, I edited it really weird so it came out sounding like it didn't. Sorry, Sparc. My point was that you couldn't figure out how many amps it could supply at a constant voltage (assuming a booster) without the E = 1/2C * V^2 thing.
E is energy in joules. C is capacitance (in Farads). V is for voltage. ;)
A 50F 100V cap bank would be:
E = 1/2(50) * 100^2
E = 25 * 10,000
E = 250,000J
Since a Joule is one watt-second, it takes 3,600 of them to make a Wh. 250,000 / 3,600 = ~70Wh.
It makes sense if you know that the distance between the electrodes in a cap (and a few other things) determine its capacitance. By putting two identical caps in series, you're effectively doubling that distance and halving the capacitance.
I'm going to have to disagree with you there. Not all supercaps are created equal. Some have stupidly high power densities. Like Maxwell caps have something like quadruple the power density of A123s by weight (something like more than 12kW/kg). And they can be charged at the same rate, so definitely no issues with regen.
The author of this post isn't responsible for any injury, disability or dismemberment, death, financial loss, illness, addiction, hereditary disease, or any other undesirable consequence or general misfortune resulting from use of the "information" contai
Why bother thinking about the technical superiourity of supercaps when they are so expensive? I'm not going to be spending $1000 to have 100% of a wheel's braking be regenerative on a 300 lb vehicle. Unless you know of a good source for supercaps for the consumer. I had trouble finidng anything cheaper than digikey or mouser that lists prices.
For the LULZ.
And, currently, they're only superior in terms of power delivery, recharge time, and cycle life. They're about 1/7th as energy dense as SLAs last time I checked.
The author of this post isn't responsible for any injury, disability or dismemberment, death, financial loss, illness, addiction, hereditary disease, or any other undesirable consequence or general misfortune resulting from use of the "information" contai