Eliminating battery sag?

[r3volved]

Forgive my electronics ignorance...
If you had a capacitor bank between your battery and your controller, could you effectively reduce or eliminate voltage sag?
My logic is that the capacitor bank would hold and dump a peak load on demand to take that initial load off the battery cells?

 

[Matt Gruber]

better to use a small pack of HIGH POWER cells in parallel, they prevent most of the sag, and at normal amps, your regular pack should not sag much.
problem solved.

 

[dogman dan]

It wouldn't help with the sag you have under continuous load, but it could help keep the battery from seeing as much of a start up spike in amps. Some small cap wouldn't do it though, it would take huge expensive ones.

But since better batteries is so much cheaper..... And they will have less sag under continuous loads.....

Bear in mind, the entire pack need not be a high c rate, It can be part 20c, and then some 40c to help reduce sag.

 

[r3volved]

Ah ok thanks
I'm not really trying to eliminate sag ATM, it was just something that had crossed my mind.
If you're mixing C ratings, would that cause higher c rated cells to discharge quicker and unbalance?

 

[Matt Gruber]

If the various C cells are wired in parallel each cell will stay even.

 

[dogman dan]

Under a heavy spike, the high c rate battery will briefly dish up more of the current, then when the load lessens the low c battery will put it back into the high c battery. This will be controlled by their individual internal resistance. The low resistance cells will give more. As the low resistance cells discharge, they will slow when their resistance climbs. The end result is less sag under a spike, mostly simply because the battery is now larger, with less sag under continuous load. But the small pack of better cells will smooth out the spike on the lesser cells. How much it helps depends on the ratios.

But if you want to carry a very small battery, and have less sag, then the whole thing needs to be higher c rate.

 

[r3volved]

Awesome... That is super good info!
It all makes total sense. I was definitely not aware that was possible.

 

[999zip999]

First we must know what type of battery using what cell ? As not all cells are the same even in lipo, lifepo4 and sometimes you get what you pay for. If lucky...
Plus Sag is what batteries do,

 

[dogman dan]

It can be worth it to just carry more capacity, even if some of it is lame. I got a big improvement, carrying some very tired lead and some NiCad on one bike. Only did it long enough to get more NiCad. But it definitely beat either one by itself.

I had too much sag with either kind, but together it worked great. At the price of being heavy as hell, but the sag was much less.

 

[TheBeastie]

Forgive my electronics ignorance...
If you had a capacitor bank between your battery and your controller, could you effectively reduce or eliminate voltage sag?
My logic is that the capacitor bank would hold and dump a peak load on demand to take that initial load off the battery cells?

Yeah it would work, and we get the odd person asking if its been done.
A few anti-capacitor trolls do jump in and say "NO!" it cant be done etc.
But it has been done by this guy,
[youtube]KW8PF5RiRnI[/youtube]

More here:
http://www.endless-sphere.com/forums/viewtopic.php?f=6&t=7511&start=25#p118675

Thing is that guy was working at a university and was donated his ultracap bank from Maxwell which cost probably at least $20k 6 years ago.
They are somewhat cheaper now, the D size cell seems pretty cheap practical to take the task on ($11 USD per 350 Farad cell), still for the cost its still a geeky project. It should make your battery pack have more total cycles (who knows how many though) due to considerably lowering your takeoff discharge current/C rate as takeoff uses by far the most power from your battery pack but if I try to think maximum American style this gain is not good enough to be acceptable. NO ONE on here rides an ebike to help save the environment its all about saving money


If your asking has anyone done this type of stuff recently the answer is no, no one wants to spend the required money, or they at least don't want to post on this ES forum about it if they have.
If your wondering if there are some miracle cheap ultracapacitors hiding out there that can do it other then my spreadsheet below that answer is double no, the Maxwell ultracaps are as cheap as it gets, at least for something that would work.

I have wondered about it my self as the "cap maths" adds up easily for a takeoff boost as long as you ignore the cost.
I posted a bit more about it on this thread
http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=59306&start=25#p887751
I also did a ultra-capacitor spread sheet to sum up cost and weight etc.
https://docs.google.com/spreadsheets/d/1sLe0sqyq33GyAzIIoaE8JWZOhbOO28GxlnNDwWlad3w/edit#gid=0

 

[r3volved]

First we must know what type of battery using what cell ? As not all cells are the same even in lipo, lifepo4 and sometimes you get what you pay for. If lucky...
Plus Sag is what batteries do,

No particular chemistry (I use lifepo4, but i'm just trying to wrap my head around the theory)
It peaked my interest when I was watching a bunch of interviews with Elon Musk and he started going on about his interest in ultra-capacitors.

Seemed logical with my limited understanding. I really appreciate the feedback...several good points I just learned!

 

[Punx0r]

No. There isn't a single reason to attempt to use electrochemical capacitors to stabilise the voltage of a low C-rate battery.

LiCo batteries are available at power densities of ~5000W/kg and are cheap.

Just because you can do something, doesn't mean it's a good idea. You can propel an office chair with a carbon-dioxide fire extinguisher, but that doesn't automatically make it an affordable or practical means of propulsion.

 

[cal3thousand]

Matt Gruber already nailed it earlier in the thread

DONE

 

[Kingfish]

Did anyone think of the other solution – that is to upgrade the Battery Bus? That’s a lot cheaper than buying batteries. Do that first, then evaluate if you need special capacity.

FWIW – I am running 15S6P (63V/30Ah) Zippy FlightMax at 15/2C. The main battery harness is 6-AWG, and I branch to Paralleled units with 8-AWG. With my 2WD, each controller also gets 8-AWG, and in turn drive each wheel at 2 hp.

I don’t see sag. But then I’m not racing. Are we racing or just trolling normally?

Unencumbered, KF

 

[r3volved]

I won't even know if I have sag until I get my CA hooked up. Not really any issue with my current setup.

I just heard Elon Musk discussing his longing to discover real super capacitors capable of replacing car batteries. I was just think if possible to sort of combine the two ideas to get some perks from both elements. Seems some people have tried.

 

[dogman dan]

Hell yeah use caps if they are free, or that cheap. But a small amount of better cells is cheaper than caps, AND, it increases your range.

Good point about upgrade the wires from battery to controller. I bet I see more sag than I should. But for my needs, 12 to the controller is adequate. I'm only running a 30 amps controller.

But our main point has been, add more batteries to lower sag. Your whole pack need not be high c rate lipo, but 5 ah of it will stiffen up a laptop lico pack nicely.

 

[John in CR]

If battery sag is a problem for you, then so will reduced voltage at lower SOC's. Lifepo4's flatter discharge curve helps with that, but the real answer to both is to first get the most powerful cells you can afford. That means lower internal resistance, so less sag, but also less heat in the battery pack and that means greater efficiency.

Even the best most powerful cells are subject to lower voltage later in the discharge cycle. That makes the one true solution being to run a high enough voltage with a powerful enough motor, so you never ride around at WOT. My bikes have been that way for years, so I ride the speed I choose without sag being a factor at all. I really don't have the foggiest idea why people settle for such piss poor performance that they have to ride around at WOT. That means not only does sag or status of charge affect your speed, but even the slightest incline or tiniest puff of wind in your face will slow you down too. That would never be acceptable for gasser cars or moto's, so why accept it with ebikes?

 

[Matt Gruber]

Case study in sag reduction.
I shortened a razor e300 scooter by 9" to fit in my corvette, to ride around at car shows (scooter illegal on streets)
the stock controller was 29a at 24v.
i wanted to use as few cells as possible.
ended up with just 2 makita packs! 2p10s. this is the runt pack john said i was wasting my time with.
solution:
1. new $10 controller limits amps to 15a. (has 20a switch but it goes to fast so i dont use HIGH)
2. throttle stop, thumb adjustable, limits throttle to where ever i set it, about 7a.
3. 36v up from 24, reduces sag from copper wires.
4. 80T sprocket, was 55T, greatly reduces load, gives perky acceleration. Reduces sag.
So with a tiny junk pack i have a fun scooter for shows, and i still have my better packs for my ebike.
And the big surprise is that at 7a it goes FASTER than it did stock with 29a24v. 13mph up 1 mph, was 12 mph.

 

[Tommy L]

Hey guys!

I've got 2 cents for some here......... It's my Pet Peeve.

In my opinion...... one should think in VOLTS with Amps rather than Amp's alone when wire sizing.
The Higher the VOLTAGE the less resistance as electrons flow faster
I'm very efficient running 144v HOC and always above my nominal voltage of 128v on 40s LiFePO4 and using a 28mm 10wind motor
Even with my lower voltage controller at 88volts @ 76.8volts nominal, I'm finding that a 10wind motor gives great inefficiencies.
That being said, I'm also current limiting to 30amps in both scenarios about 4000watts at 128v and 2400watts at 80v. Also, sizing
Cells/batteries is paramount in order to reduce the possibility of damage to Cells/Batteries. You won't have SAG if you SIZE!

Just keep in mind.... the lower the system (controller/battery voltage) the larger the wire.
If you look at a Car, the 12v starter gets the Big Wire. Sometimes as big as 2 gauge for larger engines. At 12v and 100amps, that 1200Watts of WORK!
Then look at house wiring.... 15amp circuit at 120v on a 14 gauge wire (1800 watts).
This is why I love Higher Voltage Electric Propulsion
This chart below is for reference and of short runs. Step it up a size if longer wire runs are need.
This is not a perfect chart. Do your RESEARCH before attempting any electrical ventures!

Tommy L sends.......

Gauge ..110V ....12V............Watt's
22 ........5A ......5A...............550 vs 60
20 ........7.5A ....8A..............825 vs 96
18 ........10A ....10A.............1100 vs 120
16 ........13A ....20A.............1430 vs 240
14 ........17A ....40A.............1870 vs 480
12 ........23A ....60A.............2530 vs 720
10 ........33A ....100A...........3630 vs 1200
8 .........46A ....150A..............5060 vs 1800

 

[Punx0r]

In my opinion...... one should think in VOLTS with Amps rather than Amp's alone when wire sizing.
The Higher the VOLTAGE the less resistance as electrons flow faster

This isn't correct - voltage drop on a cable is only proportional to current and resistance, with the resistance being a physical characteristic of the cable. The applied voltage is irrelevant.

V(drop) = IR

You are correct that higher voltage allows more power to be delivered on the same size cable.

Wiring charts for maximum current on a cable are variable, depending on the application. Mains wiring is usually limited by a maximum allowable voltage drop, whereas chassis wiring is usually thermally limited (this obviously depends on the location and routing of the cable and the insulation temperature rating).

 

[cal3thousand]

Hey guys!

The Higher the VOLTAGE the less resistance as electrons flow faster

This seems to violate Ohms law...

I think you mean higher voltage (at a given resistance) means higher current (ie electrons flowing faster). V=I*R

 

[Hillhater]

Just keep in mind.... the lower the system (controller/battery voltage) the larger the wire.
If you look at a Car, the 12v starter gets the Big Wire. Sometimes as big as 2 gauge for larger engines. At 12v and 100amps, that 1200Watts of HEAT!...

No !.. thats not heat dissipation,..its just the power transmitted ( to the starter )
You would have to measure the volt drop between the battery and the starter terminal , to figure out the power lost (as heat)

 

[Tommy L]

ouch! Seems to be some very schooled individuals here
I'm just a math teacher and I work well with numbers.

I'd much rather have the pressure of 140v going down my 10 gauge wire at 30amps (4200 watts) then having
50v going down a 10 gauge wire at 84amps for the same 4200watts of work. At which point I'm thinking that
the 10 gauge wire is acting more like a light bulb filament then transferring energy form source to load.
I would believe that the end result would be less power at the wheel.

Do correct me if I'm wrong in my thinking, as I do not know everything. What I do do, is Experiment - Observe and Take Notes
I have shoulders...... but please be kind when throwing text books at me.
I'm here to learn and share what I've learned

Tommy L sends.....

 

[Punx0r]

I'd much rather have the pressure of 140v going down my 10 gauge wire at 30amps (4200 watts) then having
50v going down a 10 gauge wire at 84amps for the same 4200watts of work. At which point I'm thinking that
the 10 gauge wire is acting more like a light bulb filament then transferring energy form source to load.
I would believe that the end result would be less power at the wheel.

Giving your 10awg wire an arbitrary 0.1 ohms resistance:

@ 140V/30A:

voltage drop on the wire = 30x0.1 = 3V
power lost to heat in the wire = I^2R = 90W
power delivered to load = 4200 - 90 = 4110W

@ 50V/84A

voltage drop on the wire = 84x0.1 = 8.4V
power lost to heat in the wire = I^2R = 705W
power delivered to load = 4200 - 90 = 3495W

We would also need to consider that as the wire heated up it's resistance would increase, further increasing heating (a vicious circle).

0.1 ohm is greater than it would be in reality, but the example is just a comparison (you may wish to check my arithmetic )

 

[Tommy L]

Giving your 10awg wire an arbitrary 0.1 ohms resistance:

@ 140V/30A:

voltage drop on the wire = 30x0.1 = 3V
power lost to heat in the wire = I^2R = 90W
power delivered to load = 4200 - 90 = 4110W

@ 50V/84A

voltage drop on the wire = 84x0.1 = 8.4V
power lost to heat in the wire = I^2R = 705W
power delivered to load = 4200 - 90 = 3495W

We would also need to consider that as the wire heated up it's resistance would increase, further increasing heating (a vicious circle).

0.1 ohm is greater than it would be in reality, but the example is just a comparison (you may wish to check my arithmetic )

No need to check a British Chaps Math!
So, I may not have said a text book explanation earlier but my knowledge through Experimenting is also correct
Nice to see the numbers here and the known fact that increasing heat a vicious circle.

I have said it here on this forum many times...... Size it right! Which goes for wire and battery "C" rate.
Many people here talk about the sag on Headways or other LiFePO4 cells/pouches.
My Lawn Tractor Conversion is all LiFePO4 cells. 51.2v nominal and 55v after charge and resting for a week. Never goes below 50.5v
during the whole cutting process. But I sized the 5C pack to operate in the 2C range, so never a drop in voltage and 3C more Headroom for wet or thick grass
(3.1v per cell which through my experimenting I noticed that this is the normal under load voltage per cell for LiFePO4, people should know that 3.65v is the final
charging voltage and surface charge and should not consider .5v drop per cell as Sag)

Tommy L sends....


[youtube]ygu8bV423wU[/youtube]