am i correct in calculating this amount of surface ?

considering the falcon's ev is a "regular" triangle bag where it's base is 46 cm and it's height is 28 cm, that would means it's surface would be 644 square cms, now taking the 26650 battery since it's diameter is 2.6 it's surface should 5.3 cm. which would mean that the bag should be able to hold 644/6.3= around 121 cells, which would mean 112 cells if using 16s7p configuration which would 16ah. am i right?

also if have an ecospeed controller, do i really need a bms?

I would draw out the dimensions of the bag, then cut disks to mock up the cells to see how many could really fit. It's just the way my brain works, I grab stuff and manipulate rather than calculate.

As for the bms, yes, you always need a bms. No, it doesn't have to be an automatic bms. What I mean is, if you don't put a bms on the pack you are the bms.

We do it all the time with RC lipo, but sometimes we screw it up. Sometimes a bms board fails. But generally, humans can be a bit unreliable.


If you do your own monitoring, start with a decent way to read your pack voltage so you don't overdischarge, and run a conservative dod so one cell doesn't get run too far.

Wire the pack with some Jst plug balance wires for 8s each. That way you can easily monitor the cells with a cellog 8 or similar device. Then you will also be able to use a balancing charger (8s RC charger) to balance the pack, or use a single cell charger to bring up low cells one by one when you need a balance.

If your controller has an LVC that matches your desired lowest possible discharge voltage, that can help prevent pack damage. But if you are running without a bms to cut off the pack when a single paralell cell group gets too low, I would advise stopping before the controller lvc gets close to cutting off. Whole pack lvc is just too crude for running a battery to the very end of a 100% DOD discharge. To discharge than deep, you need individual cell group monitoring.

this is from the ecospeed's controller's 2 reference on how to change settings in the controller via the serial cable.

batt:

Sets 3 values:

1 Minimum battery voltage, in millivolts. If battery drops to this voltage under load, motor power will be reduced to keep battery voltage above this level. This is used to avoid straining a weak or cold battery. Higher values than the default reduce battery strain, but can cause power loss if the battery can't stay at the requested voltage.

2 Battery cutoff voltage, in millivolts: If battery drops to this level, the controller will shut down to protect the battery. Lithium batteries also have a BMS (Battery Management System) that performs this same function. So, this is just a backup to that. The default value is set well below what a BMS would typically use for a cutoff voltage. You can set this higher than the BMS cutoff voltage which will reduce the useable capacity of your battery, but extend its life.

3 Maximum voltage drop, in millivolts. This is similar to one, but specifies a maximum voltage drop under load. So if the battery is at 54 volts, and this value is 6 volts, the battery will be limited to 48 volts. As the battery discharges and its resting voltage drops, the minimum voltage will be reduced to maintain the 6 volt drop. So at 50 volts, we will go down to 44 volts. This keeps up until the minimum battery voltage, 1 above, is reached. This parameter serves to limit strain even with a fully charged battery.

what's why i was asking why do i need a bms (which according to what i understand does exactly that).

no of course you don't need a BMS. it costs so much to pay for the cells why waste even more money on a BMS? just let us know how it turns out. i have no idea how you come up with these numbers but if it works for you then keep on keepin on.

Many of us do ride with no bms, other than our own monitoring. And sometimes we screw up and ruin batteries overdischarging them, or overcharging them.

But the key thing here, is your controller is not monitoring the battery at the cell level. So if your battery is perfectly balanced you are ok. But if it's not having cells that are asolutley identical in performance, then you can easily overdischarge one cell group long long long before your LVC which only knows the total pack voltage stops it.

The bms is the solution to using a battery with slightly less that perfectly matched cells in it. If one cell gets low early, the whole pack shuts down.

Those of us practicing the human monitoring rather than use a bms typically practice a very very very conservative discharge plan, never getting close to cell damage by overdischarge. Very often a wattmeter is used so we know exactly how many watthours we have used. We know our pack intimately, and know exactly how many watthours we have in the pack. Hop on a bike with no bms, then ride till the controller stops it could work with conservative settings, but it's no substitute for monitoring at the cell level.

If you do want to reset your controllers lvc, 16s lifepo4 should be stopped by 44v. But if you wanted to set it conservative, yet not half your range, you could try setting it to 48v.

so basically i need a bms, hmm, a123rc doesn't sell one i think, not the 48 volt kind (they do sell pcb's spec i think, don't know what that is)..

yes, do that. build up a battery with no BMS from scratch and then use it.

manually monitor it to death like it is a sick baby, always make sure you use only about 40% of the battery capacity to be safe because that is what people think is safe, and then let us know how it turns out.

most of the people who do ruin their packs never tell because they don't wanna let on. others always blame it on the accident of leaving the controller on or not knowing that the charger did not charge it up when they thought it did.

maybe you can guess how much current the cell logs use to keep track of the voltages and then you can watch all 16 channels like a hawk until you see one go low while you are riding.

sure, you don't need to waste your precious money on a BMS because all the smart people call it a battery murdering system, and they know because they are experts on batteries and know how the BMS works better than anyone else.

it is just a waste of money and the reason that the manufacturers build batteries with a BMS is just in order to raise the price. i am sure you know better than they do.

OK I got the difference between a bms and a controller in that the bms has access to all the cells (wait, ALL the cells as in 112 wires to pull ,on top of the ones that do the discharge and charge?)
but aren't bms like motor controllers in that they only come with the pack they work with? I don't know what properties to look for..

for example is this: http://www.bmsbattery.com/bmspcm/323-17s26s-24a-max-discharge-current-bms.html considered reliable? i understand if i'm making 16s7p i can actually connect the 7p as one battery (as each paralleled configuration is considered one battery).

I would just get a signalab 16s bms myself. I've come to like the blinky led's that show you when it's fully charged and balanced. If you want to really pull huge power like 100 amps, I would not use a bms per se, but rather some kind of low battery warning device. That way the humongous power won't have to pass through fets on a bms. But a signalab bms from ping would be fine for 30-40 amps. Just email ping and tell him what level of amps you want to be using. He has stock bms, or higher rate ones for 40 -50 amps.

But more and more, I do practice at least some minimal monitoring of my pingbattery. While it was new, I used a cycleanalyst with it so I always know pack voltage and how deep I discharged it. Now that my ping is getting old and tends to unbalance a lot, I even added some 8s jst plug ports to the wiring so I can check every paralell cell groups voltage with a cellog 8 in just a few seconds.

I try to dilligently type cell or paralell cell group depending on my meaning. But it gets tedious, so often enough many of us just say "cell" when refering to a group of paralelled cells in a battery pack. Once paralelled, the group more or less behaves as one cell. So you got it right, no matter how many p cells in the pack, you only need 17 wires to monitor a 16s pack. One negative, one positive, and 15 that are both + of one cell, and the - of the adjacent cell.

Best approach of all is use a bms, but still monitor the battery. Using the bms allows you to practice deeper discharges much less risk of overdischarge of a single paralell cell group. But still monitoring the battery manually allows you to smell a rat if the bms fails to stop the discharge when it should. Kind of like having a belt and suspenders on your pants.

Running without a bms is really best only for applications that require huge amps, and using the "fun" budget where lots of time spent fiddling with montioring, and even battery destruction could well be part of the fun. Transportation budget, you want batteries to last a long long time by the use of conservative discharge rates and depth. Then the bms is highly recomended.

but i've allready checked pingbattery site, they have no bms there. unless i'm missing something. cellman does have http://em3ev.com/store/index.php?route=product/product&path=42&product_id=66 and i don't predict i'll go over 40 amps.

although i don't have a cycle analyst i do have speedict.
when you say monitor the battery do you mean just plug in one of them led voltage meter? cause i already have those , since i requested ecospeed to open up a the voltage in on the throttle output i connected it there (splitting it from the lines to the throttle itself, thus i have 4 wires, 2 of them splitted).

He'll sell you one. Just pop him an email, and tell him what you want, regular or higher current. Same thing with spare cells, etc. He doesn't put it on the page, but he can get you parts. He would rather just send you the whole pack of course.

Nothing wrong with cellmans bms either. His choice works fine with a123's of course.

Monitoring can be anything you want it to be. Mainly I just meant a simple voltmeter or better still a wattmeter.
But it could be full blown individual cell monitoring if you wish.

Somtething so you know about it before it's too late if a LVC fails to stop discharge in time. You see 44v resting, time to stop! Or you go to ride off full charged and see 48v, WTF? obviously it's not charged after all.

Well I've ordered from cellnans I like the design of metal cover, plus I always try to order first via online, less confusing. For example way back where I searched for a triangle battery I exchanged a few emails with cellman and when it came to a bag he refered me to falcon, back then I decided not peruse ,but, months later I decided to order , but I ordered the battery from falcon as well cause I confused him with cellman. I didn't even know he had a triangle battery.

I'll order don't 20 cells from a123rc but what about wires? what awful would I need?

It's only possible to fill 78% of a square area with a circle. (A circle in a box with the side of the box equaling the diameter of the circle will only be 78%).

You can stagger fill the circles to improve it slightly better.

Dig way back in the battery section for many long threads discussing how to connect many round cells.

Definitely, Absolutely, be sure you get cells with the tabs on them. Soldering directly to the end of the cell is tricky.

i think it's too much to ask for original a123 cells WITH tabs, i've ordered a roll of soldering soldering tabs. what's so tricky about it?

The M1 cells are "hard" to solder to directly. You need high heat and high thermal mass to heat the end quickly and solder the tab to it.
If you heat it more than a few seconds, the insides of the cell get too hot (some of the internals are plastic) and they melt damaging the cell.
The damage might just be reduced capacity but unless you are able to test the capacity of each cell *after* soldering, you will never know.
It is still possible to find original A123 M1 cells with tabs.

Phil

emaayan said:

i think it's too much to ask for original a123 cells WITH tabs, i've ordered a roll of soldering soldering tabs. what's so tricky about it?

Click to expand...

can i ask a123rc? to "pre-tab them?

Possibly. If you don't get them tabbed, spot welding tabs yourself is the other option besides soldering.

All the discussion of this stuff is in those old round cell construction threads.

Unfortunately I dunno how to weld