Is my controller 'bad'? Severe loss of power at low SoC?

[jeff-jordan]

That's the risk if you set LowVoltageProtect to 58V. In effect, this disables the tortoise mode.

But trying to drive home without recharging the battery is not good at all, because this a) might damage the cell-pack with the lowest capacity even more and b) is severely unsafe.

For longer journeys where there's a risk of the battery dying, I take my "spare fuel can" with me. I even built a sturdy mount for it in my side case.

___

Furthermore:
1. Don't trust the SoC bars! This is only a non-linear voltage display, with a very long time constant (5min or so).
2. Even in the lowest gear, the E-Odin might easily draw more than 70A, especially when you start off with the bike.

[dominik]

Check your voltage level during driving and if you have the possibility also the single cell voltages.
Only with that you know what is your real problem.

I still think you have a too big drop in voltage during acceleration.
Now you have to find out if it is from the battery or from something else like connectors, fuse or even bad wiring.

[bob2.0]

That's the risk if you set LowVoltageProtect to 58V. In effect, this disables the tortoise mode.

I'm not sure. 'Tortoise mode' has never worked properly, anyway.

The first time I hit volt-bottom, it was the battery that cut out, just like this instance.

The controller keeps on pulling down to the battery going dead, whatever LowVoltProtect is set to. The reason I set it low is so the derating curve (that starts from 74V) started later.

The flat battery was my fault, it was totally a 'rider error', not an engineering error. I had done a 'mid-charge' and misjudged how much charge it had taken on.

I just went too far and ran it flat. Only by a few miles, but misjudged it.

Check your voltage level during driving and if you have the possibility also the single cell voltages.

I'll install an instantaneous voltage readout. I was meaning to do it before, I did not get around to it.

The battery is why it cut out, I can't control what is going on in voltage drop terms inside the battery.

I have struggled with this 'derating' that happens in my controller from 74V or so. I think you might be suggesting that this is happening because of voltage drop, but this derate curve is clear and distinct at any power/current drawn.

If you guys do not experience this effect with your controllers, then I think I have to go ahead and replace my controller. Maybe you are right and there is a voltage drop happening, and perhaps it is in the controller itself? Something weak in the bridge transistors, perhaps?

I have now charged it fully back up. It took a long time in the last stage of charging, it was dribbling <1A for several hours, presumably balancing up the cells? I think the battery should be OK, it cut in 'speed 1' and recovered to 68V, so I don't think it dipped that low but maybe the BMS spotted a weak cell getting pulled down. I hope it is not a future problem.

[jeff-jordan]

That's the risk if you set LowVoltageProtect to 58V. In effect, this disables the tortoise mode.

I'm not sure. 'Tortoise mode' has never worked properly, anyway.
...

This/your thread is titled: Is my controller "bad"? Severe loss of power at low SoC?
That's exactly what the tortoise mode does, it severely reduces the power at low SoC.

Now you doubt that it worked properly ???

[bob2.0]

This/your thread is titled: Is my controller "bad"? Severe loss of power at low SoC?
That's exactly what the tortoise mode does, it severely reduces the power at low SoC.

Now you doubt that it worked properly ???

The loss of power is NOT tortoise mode, and I've never said it was.

The power curtailment begins at 80% SoC, ramps down the power limit in a generally linear way, and just gets worse and worse.

It doesn't even get as far as tortoise mode.

I'd expect tortoise mode to actually be faster!!!

As I asked last year, is it meant to do this? I can start a long trip in 'speed 2' and by 70~80% I set it in 'speed 3' just to keep up the same power as I got in speed 2 at the start.

By 50% it is a snail's pace and slows down to 50kph on any sort of steep hill.

By 30~40% it struggles to pull off on a slope unless I push off on the ground.

This is far from where any 'tortoise' has any business to be.

What I have found, the last page of posts, is that the lowvoltprotect seems to feed into this power limitation 'curve' and each volt lower I set it, it delays the onset of this power limitation by a similar amount, like the whole derating curve (that spans the whole battery charge) is pushed down a bit.

This is my point. This is NOT tortoise mode set by lowvoltprotect. But, adjusting that value helps change the effect 'it' has, whatever it is.

[dominik]

Measure your voltage during acceleration and you will see what is going on.

Or order and connect a Bluetooth Adapter and make a screen recording during acceleration.
Not guessing, measuring is the only way to find your problem.

[bob2.0]

I agree, and it was in my plan last year.

"Time flies!"

The only snag is that I cannot find a 3 wire waterproof voltmeter to buy.

[jeff-jordan]

...
The only snag is that I cannot find a 3 wire waterproof voltmeter to buy.

What about a 2 wire waterproof voltmeter (just a volt-meter), or this one (the round, external one is waterproof)?
Or even a real coulomb-meter (with 300A shunt), that allows you to determine much better what's left in the tank (battery).
Of course, installation of the coulomb-meter, with inserting a shunt into the 4AWG wiring of the battery requires much more work.

[dominik]

Or this one.
No need to cut or unscrew big wires.

[jeff-jordan]

Or this one.
No need to cut or unscrew big wires.

Screenshot_20260509_064342_AliExpress.jpg

Yeah. Or this one or that one (200A, a little bit cheaper).
I used such a coulometer with hall sensor for a while, but I always had the problem that it was pretty inaccurate measuring the current (the zero-point always drifted approximately +/- 0.3A, depending on the outside temperature).
Furthermore, it was not waterproof, which meant it had to remain in the "fuel compartment"—which is less practical if you want to monitor the voltage during acceleration.