BMS Cutouts at low speed ???

[Ypedal]

Alright... the good .. the bad .. the ugly .

The Good : 72v of Lithium makes my Chaos GO .. sweet lord it's fun sucking back 1.4 kw up hills !!!!

The Bad : I'm really puzzled, my 2 x 36v 20ah Li packs have a 35 amp limited BMS, both packs are conected in series for 72v on a crystalyte 20 amp controller, also has a CycleAnalyst hooked up with built in conector, set to max of 20 amps.

If i pedal to aprox 10 km/h i can use all the throttle i want without any problems, CA shows 1400w and voltage drops to a nice 77.x v and zoom-zoom.. away i go...

HOWEVER --- If i dare touch the throttle from a dead stop.. POP>. bms shutdown, but from only one of the packs. resulting in one unhappy Y

What gives ?

[Ypedal]

I suspect one of a few things

1- Controller : Is it possible that a 20 amp controller can allow more than 35 amps for any brief moment causing a very sensitive BMS to trip ?

2- BMS on one of the packs

3- Battery voltage: the 2 packs do not have identical operating voltage, one of the packs charges to 41.4v, the other to 42v , both packs are fully charged so i know it's not LVC on either one..


A while back i had a problem with a Pedal First controller, bms would shut down, but the same pack ran fine with an Instant Start controller, so i simply changed the controller...

I need this bike to GO from a dead stop, pedaling to 10 km/h is not going to work for me

[knightmb]

I suspect one of a few things

1- Controller : Is it possible that a 20 amp controller can allow more than 35 amps for any brief moment causing a very sensitive BMS to trip ?

I'm not sure, with the CA on mine I never see more than 25 amps spike out while riding. I've set my CA to the "insanely high update speed" setting just to test, never seen any spikes that high.

2- BMS on one of the packs

Possible, maybe it's too sensitive, etc.

3- Battery voltage: the 2 packs do not have identical operating voltage, one of the packs charges to 41.4v, the other to 42v , both packs are fully charged so i know it's not LVC on either one..

Each pack monitors itself, so that might point back to a bad BMS

A while back i had a problem with a Pedal First controller, bms would shut down, but the same pack ran fine with an Instant Start controller, so i simply changed the controller...

I need this bike to GO from a dead stop, pedaling to 10 km/h is not going to work for me

I've always wondered if these BMS determine amps by total power going through them in terms of watts, then divide out (which would cause a shutdown when putting two in series) or similar to the CA with a little shunt and reading the voltage.

[Jozzer]

Is it the second of the packs? Same pack every time? Try each pack seperatly at 36v? My guess is it doesn't like having 2 packs worth of power running through.

My 20A controllers have never let more than 20A register on the CA....unless I MAKE them

[Ypedal]

Well that's the weird part.

Does the controller feed anything back to the battery when operating ?

I assume that the battery/bms has no idea if i'm going 10 or 15 km/h.. with that assumption, releasing the throttle at 15 km/h and then flooring it would give max amps ie: aprox 25 , only difference is that when i'm rolling the bms don't trip ?

Is there anything i can install in-line between the battery and controller to prevent feedback ? or is that not even happening ?

[The7]

Is it the second of the packs? Same pack every time? Try each pack seperatly at 36v?

Try as Jozzer's suggestion.

[Ypedal]

I tried a pile of stuff last night, so i'll have to re-do the experiment.

One of the packs for SURE works fine at 36v from a dead stop.

The 2nd pack i beleive is the culprit.

However.... Both packs have been used on a different bike with an instant start controller and work fine.... from a dead stop.

[eP]

Check the snubber capacitors.

Try to double them and check the difference if will be any.

[The7]

Some thoughts on current multiplication at low speed.

If the average battery current is limited to 20A; then at PWM duty cycle of
a)100 % (duty cycle), the peak battery current = 20A
b)75 % (duty cycle), the peak battery current = 26.7A
c)50% (duty cycle), the peak battery current = 40A
d)25% (duty cycle), the peak battery current = 80A

So at very low speed (or start), if you turn the throttle and the PWM duty cycle is less than 50% for the battery current limit of 20A, the peak battery current (say 40A) may exceed the max allowable current (say 35A) of your battery pack.

It is quite possible that the BMS of the pack is very sensitive and could not tolerate short-time peak current at very low speed.

[Ypedal]

Oh boy.. something i should also add...

( for those of you not familiar with my ride ) My Chaos is using a dual speed motor ie: crystalyte 406/409

I'm using a single speed 36-72v 20amp controller, and i use a 3PDT relay to swap between 406 and 409.

To energize the coil on the relay, i use a large resistor to drop the 72v battery power down to 24v. I conected the leads for the coil directly to the controller by stripping a bit of insulation behind the anderson powerpoles and soldering them on there.

This means when i plug the battery into the controller, the coil gets energized at the same time.

I ran a small On/Off switch to the handlebars, to either Open or Close the circuit to the coil.. in 409 mode the coil is OFF.. in 406 the coil gets energized by the switch and i'm on Go-Fast mode.

All my tests were in 409.. i have not yet gone to 406....

Easier things first.. i'm going to swap the controller tonight with another one to see if anything changes.... Opening the packs right now is NOT an easy thing to do..

[The7]

However.... Both packs have been used on a different bike with an instant start controller and work fine.... from a dead stop.

For the same battery current limit (say 20A), the current multiplication could be twice if the battery voltage is double at the same speed.

PWM of 25% (duty cycle) at 72V battery would give the same ebike speed as PWM of 50% (duty cycle) at 36V battery.

But at that speed, if you floor the throttle, then for a short period of time the peak current from 72V battery would reach 80A and that of 36V would reach 40A.

So for one battery pack, the sensitive BMS may not cut-off.
But for two battery pack, the BMS will cut-off if it is too sensitive.

[eP]

Some thoughts on current multiplication at low speed.

If the average battery current is limited to 20A; then at PWM duty cycle of
a)100 % (duty cycle), the peak battery current = 20A
b)75 % (duty cycle), the peak battery current = 26.7A
c)50% (duty cycle), the peak battery current = 40A
d)25% (duty cycle), the peak battery current = 80A

Controller can't work this way !

It need big capacitors to cut the battery current peaks.

High current peaks would cause high voltage peaks (long wires) so fets would be dead at higher input voltage.

[Ypedal]

Check the snubber capacitors.

Try to double them and check the difference if will be any.

Gladly, but.. my lack of electronics knowledge/skill somewhat impedes on this activity !!!

I can follow directions very well tho.. pictures help. i know what a capacitor is.. but wich is the " snubber ".. i have no clue. Sorry.

[eP]

But for two battery pack, the BMS will cut-off if it is too sensitive.

BMS must be sensitive.
Any sensitivity is too sensitive.

[The7]

All my tests were in 409.. i have not yet gone to 406....

If it cut-offs in 409, it will certainly cut-off in 406.

[xyster]

Easier things first.. i'm going to swap the controller tonight with another one to see if anything changes.... Opening the packs right now is NOT an easy thing to do..

Ypedal, ypedal....ystrugglewithaboxyoushouldn'tneedto? Perhaps cutting a few corners is literally in order so you can just unbolt-bolt one lexan side or the other when you need to?

[fechter]

I'd guess there's a very short time when the current goes high enough to trip the BMS. Ideally, you'd want to add a capacitor or something to the BMS to delay the response a bit.

Another possible thing to try would be adding a humongous capacitor across the controller battery input. The cap could supply the surge current and prevent the BMS from tripping.

Or just bypass the BMS current limiter altogether and only use it for charging

[eP]

Check the snubber capacitors.

Try to double them and check the difference if will be any.

Gladly, but.. my lack of electronics knowledge/skill somewhat impedes on this activity !!!

I can follow directions very well tho.. pictures help. i know what a capacitor is.. but wich is the " snubber ".. i have no clue. Sorry.

Very big caps. 1000uF/100V at least, but 10 000 uF/ 100V would be fine.
Fechter will show you all details.

[eP]

Or just bypass the BMS current limiter altogether and only use it for charging

I wouldn't recommend that option.

It sound like invitation for problems.
Caps. are not so expensive - more caps >> cooler batteries,wires and connectors.

[The7]

Some thoughts on current multiplication at low speed.

If the average battery current is limited to 20A; then at PWM duty cycle of
a)100 % (duty cycle), the peak battery current = 20A
b)75 % (duty cycle), the peak battery current = 26.7A
c)50% (duty cycle), the peak battery current = 40A
d)25% (duty cycle), the peak battery current = 80A

Controller can't work this way !

Thanks.
I should state with more technical correctness as:
c)If the PWM duty cycle is 50%, the battery current would be 40A for 50% of the period to obtain an average battery of 20A.
d) If the PWM duty cycle is 25%, the battery current would be 80A for 25% of the period to obtain an average battery of 20A.

[eP]

Some thoughts on current multiplication at low speed.

If the average battery current is limited to 20A; then at PWM duty cycle of
a)100 % (duty cycle), the peak battery current = 20A
b)75 % (duty cycle), the peak battery current = 26.7A
c)50% (duty cycle), the peak battery current = 40A
d)25% (duty cycle), the peak battery current = 80A

Controller can't work this way !

Thanks.
I should state with more technical correctness as:
c)If the PWM duty cycle is 50%, the battery current would be 40A for 50% of the period to obtain an average battery of 20A.
d) If the PWM duty cycle is 25%, the battery current would be 80A for 25% of the period to obtain an average battery of 20A.

Why you assume no caps there ?

It should be cap's peak current not battery peak.


What will happend (at wires) when 80A goes to zero in 100 nanoseconds ?

[OneEye]

This may be a stupid thing to do, but would an in-line inductor help stabilize and reduce the peaks?

When Neodymics was tinkering with the DeWalt 36V batteries for their bolt-on contraption, they added an inductor (250uH?) in the line because the BMS was tripping on the high inrush current. The inductor slowed the initial inrush current and kept the BMS happy.

[Ypedal]

Like this ?

[Ypedal]

I'd guess there's a very short time when the current goes high enough to trip the BMS. Ideally, you'd want to add a capacitor or something to the BMS to delay the response a bit.

Another possible thing to try would be adding a humongous capacitor across the controller battery input. The cap could supply the surge current and prevent the BMS from tripping.

Or just bypass the BMS current limiter altogether and only use it for charging

How Humongous ?

I really.. REALLY don't want to open the Plexy boxes up.. like.. one month's worth of all my spare time really..

[xyster]

How Humongous ?

I really.. REALLY don't want to open the Plexy boxes up.. like.. one month's worth of all my spare time really..

<p align="center">"Here's my big capacitor." -Fechter

</p>