Motor shuts down due to low voltage cutoff. Do I need a new battery?

[robotron2084]

Bit of info on my system:
48v 15Ah Hailong battery pack from Amazon. About 11 months old.
CycleAnalyst v3 Head Unit
PhaseRunner Controller
TSM-A5 Motor Direct Drive Rear Hub with built-in torque sensor.

I have the CA configured to 1000w max. In general I'm doing 200-400w. Bike has been doing good for the past year.

My battery pack shook loose from its mount the other week. I reattached it with some nylock nuts, and went out for a ride and noticed seriously degraded performance, and a tendency to shut off completely. The CycleAnalyst was showing a flashing 'V' which, according to the docs mention that this is likely a low voltage cut off. My current theory is that the battery is bad. I don't have another battery for purposes of testing, so i can't easily rule it out as a problem. Could this be something else?

 

[gromike]

Battery connection and/or wiring?

 

[robotron2084]

Battery connects to the PhaseRunner via XT60 connector.

 

[gromike]

Since your battery displaced, and you have since have had problems with battery output you should do a detailed inspection of all components and wires between the battery and controller.

 

[neptronix]

What's the voltage at right before it cuts off?
Is it random, does it happen during bumps, or during full throttle from a stop?

 

[robotron2084]

What's the voltage at right before it cuts off?
Is it random, does it happen during bumps, or during full throttle from a stop?

I shall try to see what the voltage is, but it is hard to see. I wonder if I change the screen to the diagnostics display it might give me more info. I'll try this again tomorrow and see if i can tell.

It is not random as far as i can tell, doesn't happen on bumps or anything. And if you saw the road i live on, we got some bumps! It can cut out if I push down too hard from a stop. I think I see what you're getting at: if its a lose wire or bad solder maybe it would be more intermittent?

Since your battery displaced, and you have since have had problems with battery output you should do a detailed inspection of all components and wires between the battery and controller.

Good point as well! It is literally a wire coming out of the battery that connects to an xt60, then that xt60 goes directly to the phaserunner. I checked my connection points and they seem solid.

 

[e-beach]

How well balanced is your battery pack?

 

[gromike]

It's everything, the run of wire, the wire connection to the connector, and the connector. Not saying that's the problem, but because of your malfunction you have got to pick up every stitch..

 

[robotron2084]

How well balanced is your battery pack?

How would I determine this?

 

[neptronix]

Checking the balance would involve opening the pack.

An easier way to rule out disbalance is to leave it on a charger for at least 4 hours to make sure it balances itself
..if at a certain voltage the pack consistently cuts out early, like way above LVC ( at least 39 volts on your pack ), and there's some consistency to it, you probably do have the balance issue.

A balance issue will cause the bike to run for a while, then the battery cuts out way early.

If the pack drops >5v of voltage at full load ( IE from a stop ), it is probably worn out.

I wonder if this pack has no-name chinese cells. Those tend to wear out pretty early.
Cheap packs also tend to have a BMS draw power from 1 cell.. which kills 1 cell group of the battery early.

Good thing you have a cycle analyst, you have a full time voltage display to help diagnose this!

 

[docw009]

How would I determine this?

When the motor cuts out with your big V, is it down for good? Or can you power it back up and ride for a long time? Need to distinguish between an intermittent loss of power and a systematic low voltage shut off.

When you recharge the battery, see what the Cycle Analyst is reporting as full charge, Also what is the voltage when it goes out on you. Full charge on a 48V battery will be 54.6V. It should run all the way down to 40-42V, An unbalanced 48V battery usually charges up to less than 53V, and shuts off way early.

 

[robotron2084]

When the motor cuts out with your big V, is it down for good? Or can you power it back up and ride for a long time? Need to distinguish between an intermittent loss of power and a systematic low voltage shut off.

When you recharge the battery, see what the Cycle Analyst is reporting as full charge, Also what is the voltage when it goes out on you. Full charge on a 48V battery will be 54.6V. It should run all the way down to 40-42V, An unbalanced 48V battery usually charges up to less than 53V, and shuts off way early.

When it cuts out, I can immediately turn the CA off and on and Im back up and can keep riding. I just plugged it in and I'll check the voltage level when its done. I will say that I did notice that the battery level lights on the battery itself does show 5 green dots and one red dot when full.

Checking the balance would involve opening the pack.

Kinda had a feeling that might be the 'case'.

If the pack drops >5v of voltage at full load ( IE from a stop ), it is probably worn out.

I believe I'm seeing some pretty big voltage drops but its really hard to see in realtime, riding my bike and the CA display is kinda laggy. I'll give it a test ride today after its charged.

I wonder if this pack has no-name chinese cells. Those tend to wear out pretty early.
Cheap packs also tend to have a BMS draw power from 1 cell.. which kills 1 cell group of the battery early.

Ugh, yeah. This pack was well rated when i bought it. Now there are lots of reviews talking about this problem, so i guess i got suckered.

Good thing you have a cycle analyst, you have a full time voltage display to help diagnose this!

Its a bit user unfriendly but I don't mind getting my hands dirty a bit. Overall the experience has been good from Grin Tech, they've been helpful. Of course the one component I didn't buy from them was the battery...

I'll report back here, but this conversation has really helped me get a feel of where this problem is. This is my first post and I appreciate the help. I feel a bit more confident to drop some cash on a new battery and know that it will probably fix the issue. Still wish I knew someone local with a compatible battery so I can test the rest of my system tho...

 

[99t4]

I believe I'm seeing some pretty big voltage drops but its really hard to see in realtime, riding my bike and the CA display is kinda laggy. I'll give it a test ride today after its charged.

Good way to test this is to start out at the base of a steep hill. Note the voltage before you start. Then watch the voltage as you apply full throttle from the dead stop. Subtract that voltage reading from the before-start reading. NOTE: If your bike is wheelie-happy you will have to manage that.

 

[amberwolf]

If the battery actually fell and hit the ground, then it's possible that some of the interconnects between cells are damaged and some of the cells in various parallel groups are no longer connected at one end or the other, and no longer contributing to the power the pack provides, so those groups sag in voltage more than others.

That often causes the BMS to shut the pack off early especially under load, but it can just cause greater votlage sag on the entire pack than before--if your system is setup to reduce power demand as voltage drops (both the PR and the CA can be setup to do this, depending on the firmware version, IIRC), then it might not push the pack hard enough to cause it to shut off in these events.

But it does require opening the pack to check for this problem.


Another problem that can happen with impacts is physical cell damage, which is dangerous, because you can't predict how a damaged cell will behave over time--there is a risk of sudden catastrophic failure (internal short, fire, etc).


If the pack has a fuse in it, in a typical fuse holder mounted to the case, it can also be jarred just loose enough to no longer make good contact, and have high resistance causing voltage drop there under load, keeping the rest of the system from seeing that part of the pack voltage.

 

[robotron2084]

Ok, I took some time out this evening to try to figure out what's up and...I think I might have fixed it? This is one of those 'did random things' fixes and one of them seemed to resolve the issue. Time will tell if its completely fixed.

tldr: I think by filing down my m5 screws they stopped putting pressure on the battery case and that made a difference. Alternatively I jiggled something into place when I took apart and put back together the battery pack.

Fully charged voltage was 55.2V and 4.24Vc. A few hours later that was 55.0V and 4.23Vc. I spent some time going over all the wires and seeing if there were any weak connection points and couldn't find any.

So I did the whole 'go up a hill' thing and I had no issue getting up to 1000w (my cap) without it cutting it out. I did a few laps up and down the hill next to my house and it was fine. In general I was seeing maybe 5v of voltage drop tops? 55v going down to 50v. I was a bit stumped so I rolled around a bit more and got it to cut out in the areas I normally ride. These roads are bumpy and aint great. So at this point it was looking more like the issue was a loose wire on bumpy roads, not the battery. I did a couple miles on smoother roads without issue.

I then took apart the battery pack and looked inside to see if there might be a loose wire. I measured the voltage via multimeter at 53.37. The cells looked, to the best of my knowledge, like standard samsungs and didn't seem to be noname. But I couldn't see much of the cells as they were wrapped and heat shrinked etc. Then I tried squeezing and torqueing the battery pack while out of its case and couldn't get the multimeter to report a dropped voltage. The BMS was under heat shrink and really I couldn't make out anything useful without taking off the heat shrink and protective papery bits. So I shurgged and put it back together.

I then realized that the M5 screws I use to connect the downtube 'plate' to my custom mount were a bit proud...and I wondered if maybe this was the issue. I could see it wearing marks on the bottom of the battery. Long term that's not going to be good anyway, so I took some measurements with my calipers and filed them down 2mm.

Sure enough, I'm having zero problems after re-installing. The issues seemed to start when i tightened up these bolts, and I wonder if somehow this was causing some weird pressure on the wiring or BMS. I test rode it really hard to break it again but it seems resolved for now! Fingers crossed...if this holds for the week I will be shocked (in a good way! ).

 

[robotron2084]

Since pics are probably more helpful, here's some pics of the battery and the mount its on, and the now shorter M5 screws. I know a lot of this aint pretty, but it works and all my sins are hidden inside the pannier. You can see the wear marks on the battery...probably not great.

 

[docw009]

OK, it was a connection problem somewhere, not a battery issue. I see that you took out the Hailong pins and wired a long cable (with XT60) to the internal battery connections?

I am familiar with those Hailong cases. They're pretty crappy, at least mine were. I had one where the battery was shipped new with the BMS balance plug loose in the connector. Every time I hit a bump, the battery shut off and came back on. That was not a case problem, but the crappy part was that the battery cradle is only held by two screws, and that the battery's locking pins to the cradle only cover 40% of the battery's length,

I've never seen a Cycle Analyst, nor do I want one. I am surprised that its voltage readings appear high.

 

[amberwolf]

FWIW, the CA has a place in settings to calibrate the voltage readings; probably in the Advanced settings that take an extra step to get into. Before changing those, make sure your multimeter is really reading correctly by testing something else that has a known voltage (like some 5v USB chargers or the USB ports on your computer, etc, which *should* be very close to 5.0v; if they all read low on your meter it may mean the meter isn't well-calibrated, and the CA could be correct already). Also, note down the original CA V/V value before you change it, so you can change it back later if you find you need to.