It's the BMS, Stupid!

[999zip999]

Those ping pouch cells will look like a popcorn pillow that came out of the mircowave if used at high amps. Maybe lipo or A123 battery. Ping makes a great battery for it's purpose a commuter bike. Like 22amps maybe 25amps to last the test of time. You can buy Pings high output bms 60a. and run a 40 amp controller but it will take some life out of it and the pouches will puff. High demands need a high output cell battery.
With all that you could build your own A123 20ah pouch battery with the kits found here.

Edit: Wow I saw that you do have four 36v 15ah Ping batteries for this. Still mind blowing

 

[dogman dan]

About 60 pounds of battery, 4 of those. No range anxiety I guess.

 

[999zip999]

4 x bms's have to take apart like lipo to charge for the safety of lifepo4. Sorry I hard for me to rap my head around. Yes I know the plus side in ? But.

 

[llile]

Ok, I'd like to put a halt to the speculation:

1. I am still alive, and have not blown up any batteries or houses despite rumors to the contrary. Pirates and brigands have not overrun my car at an intersection nor am I running on LIPO, hydrogen, A123s, lead-acid, etc.

2. I am running two 15AH Ping 36V batteries in series, which as I count 'em makes it a 15AH 72V battery with a 40A current limit. I have room for two more such batteries to make a 30AH pack but have not forked over the cash yet. Yes, I will need that much battery for some long range trips.

3. I am running two motors with two 40A controllers (or at least, until a few minutes ago they were 40 amps....) A surge at the beginning of a throttle twist seems to be the culprit, overloading the PING BMS current limit.

Simple measures first! I took apart the motor controllers, and found the two shunts (quite obviously shunts, to my eye, pics to follow) in parallel. Snipped one shunt, and bent things to they cannot touch. Made a stab at a four wire ohms measurement and came up with 4.6 milliohms, about double the 2.21 milliohms marked on the controller. This measurement is at the limit of the equipment I have at hand, so it is not extremely accurate. Did this surgery on both controllers. This is despite John Robert's generous offer of reprogramming the controllers. He will get a crack at this bike soon enough anyway. He still owes me a wheel truing.

Hooked all the electronics back up to the bike. Voila! No tripping the amp limits on the BMS. I plugged an ammeter in where the battery fuse would go, and set the Cycle Analyst to 1, then 2 then 4 amps, and calibrated the CA milliohm shunt setting so my ammeter matches the reading on the CA within 5% or so. Then I set the CA at 15 amps, which seems to overshoot to 18 or 20 before it comes under control. Bike ran fine up on blocks, no matter how hard I crank the throttle. Took it out for a spin down the driveway (800 feet one way, a good test run) and it made it successfully. Climbed a small hill. floored it, never killed the amp limit in the BMS. Thanks to all who posted various solutions - I now understand some of the components a lot better, and have catalogued this issue for others who may run into it. next post I will show dissecting the motor controller shunts and explain four wire ohm measurements for any who are interested. .

 

[llile]

OK here goes

Two Controllers


Remove this endplate:


Two shunts in parallel. How do we know? They are connected in parallel to the negative trace, and they are nothing more than pieces of nichrome wire. They'd have to be for 40 milliohms resistance.
View attachment 2

Snip one and shave the other a little.


DO NOT remove this endplate. You'll spend the next thirty minutes stuffing the wires, silicon rubber gasket, and all back into place. Ans there are no shunts underneath it.

 

[llile]

What is all this four wire ohms stuff anyhow?

Remember ohms law, no? E=IR. To measure a resistance, you run a current through it, and measure the voltage across it. A little math, and you know resistance. This is exactly what an ohmmeter does. However, every measurement also includes the resistance of the ohmeter leads. "What's a few tenths of an ohm among friends?" you ask. Well, if you are measuring 4 milliohms, the meter lead resistance swamps the measurement. What to do?

Going back to ohms law, we inject a current, and then measure the voltage across it. I have a 200 millivolt range on my multimeter. I have a power supply that can be regulated near an amp. Inject an amp (that is the first two wires) into the device under test, and measure voltage drop across it (The last two wires). A little math - voila! Four wire ohms measurement.

I inject the current by soldering two small wires at the base of the shunt resistor, and use the voltmeter probes to test the ends of the shunt. It matters where you inject the current, and where you measure. Millimeters matter here. The result will not be all that accurate unless you have a really good voltmeter. Still, I can tell that I have about 4 milliohms as a result, twice the original.

 

[dnmun]

i would just solder up the shunt on the BMS to stop it from tripping off for overcurrent. maybe even add another set of shunt wires even to prevent it cutting out on overcurrent.

it will not kill the battery to draw 3-4C for a short period, just not long because the electrolyte gel can only move so much charge. the new pouches have wide tabs so the current is not a problem for the electrodes.

 

[dogman dan]

My experience is that 2-3c spikes can unbalance the hell out of it though. But not just a few of them, I'm talking a spike like that every 20 seconds like I was doing on the racetrack. It's still pretty tough on that size pingbattery, but others have survived 30 amp controllers.

Long as you are only seeing 25 amps combined most of the time, it should be fine. A 40 amp spike at infrequent intervals won't do any serious harm.

 

[SamTexas]

Hi, I have the same problem ..

http://endless-sphere.com/forums/viewtopic.php?f=14&t=46589

im using 45a controller

Are you still around? Look at the previous post by llile.
http://endless-sphere.com/forums/viewtopic.php?f=2&t=47456&start=15#p698800
He basically cut his controller max current by half.

 

[DanSFIV]

thanks Sam.. Yeah my shunt is the same style. I can cut it like the link you sent.

 

[SamTexas]

You're welcome. Be careful to not damage surrounding parts. Cut it but don't remove it. You might want to solder it back in the future should you decide to get rid of the Battery Murdering System or get another battery that can support high current.