As battery depletes, when exactly does the system "throttle" back throttle? (pun intended)

[Fran Fabrizio]

I've noticed that somewhere below 50% charge the throttle stops responding as snappily. It's a pretty dramatic dropoff. I assume this is for battery conservation. Is this a software-controlled thing (nothing in the menus but just wondering what the underlying mechanism is)? I haven't had the bike long enough to know exactly at what voltage it's happening, but would love to know so I can plan for it. Of course it's usually happening at the end of the day, at the end of the commute home. I'll be at a red light in traffic and when it turns green I'm expecting to zoom up to speed and get through the intersection quickly, and instead I'm surprised when the bike is pretty pokey off of the line.

The other reason I'm asking is just to confirm this is expected behavior and not an issue with my bike. I don't notice a big change in the pedal assist, it feels about as snappy, but the throttle is like, maybe 33% of what it is normally.

 

[harryS]

It's normal. It's not in the circuitry or software. It's your battery running out of power. It's the same as a flashlight getting dimmer as the cells wear out.

You might be interested to known that a typical 48V ebike battery shows 54.6V at full charge and runs down to 40-42 volts when the system shuts down. When it reads 48V, it's actually only about half charged. IF you have a 36V battery, these numbers change to 42V at full charge and 30V when used up.

The throttle is more noticeable because depending how you apply it, that can ask for a lot of current, which the battery may not have when it's low. Pedal assist is a more measured quantity.

 

[Fran Fabrizio]

I am generally aware of how power will gradually drop off as the battery depletes (and I do the 80%/20% charging thing so I do watch the voltages closely), but the throttle seems to suddenly drop off, not gradually. It'll be responsive at one stoplight and totally sluggish a minute later. It felt very much like some threshold had been reached in the software, that's why my thinking went in that direction.

 

[Reid]

Hi Fran, take note of your battery pack voltage? There is a more detailed screen you get to by pressing power and the plus buttons together.

When the pack's resting voltage gets down to...I don't know, don't recall exactly...say, 46V, the performance of the pack falls way off.

Internal resistance and counter EMF are the culprits for the loss of zip that a partially discharged battery offers. Your instantaneous voltage will fall way below the resting voltage. The meter will not register this drop in voltage on demand of heavy throttle.

It will help to set your low voltage cutoff to 40V...but don't actually ever run down the charge to where sudden demands ever drop the pack voltage below, say (I am guessing because I don't recall my own pack experiments' numbers) about 46V.

If this is not the answer then it may be something to ask Juiced Support.

 

[Fran Fabrizio]

It will help to set your low voltage cutoff to 40V...but don't actually ever run down the charge to where sudden demands ever drop the pack voltage below, say (I am guessing because I don't recall my own pack experiments' numbers) about 46V.

Ah, this is a good clue. I believe I set mine to like 42V on Tora's recommendation in his video walking through the control settings. That may be why I'm experiencing the sudden cutoff of throttle. I will reset to 40V on the way home and see what happens. Great idea, thanks Reid!

Did you mean to say don't ever run down the charge to where sudden demands drop the pack below 46V? Because that happens to me with plenty of juice left. I'm pretty sure I've seen it momentarily below 46V at like 50% charge in my experience (what I mean is that resting voltage will be at or near 48V but instantaneous voltage can drop to 46 or less). I will confirm on commute home, though - I could be mis-remembering.

 

[Jeffrios]

Are you keeping the battery inside during the day or on the bike out in the cold?

Even if it is inside, with the pack being worn down and the battery being colder, that may amplify your drainage and cutoff at 42V that you have experienced.

 

[Fran Fabrizio]

The battery has been out in the bike locker (which has been about 32-40 degrees the past couple of weeks). I take it in at night for charging, but have avoided lugging it in during the day at work. I usually experience this several miles into the evening ride home, so even if I took the battery in during the day, it may not make a difference if this is related to temperature. We're going to have a mini heat wave this week (50s feel like 80s after a Minnesota winter!) so maybe I'll notice a difference as the week goes on and it gets warmer.

All good ideas, looking forward to experimentation this evening. Will post observations tonight!

 

[Bruce Arnold]

I notice some loss at 46V, but it's really dramatic at 44V. I just don't go below 46V any more unless I'm on the way home from a long ride.

Took it down to 42V once (with the controller set at 40V minimum like Reid mentioned above.) At that point, the throttle was barely evident, and there was no useful difference between levels 2, 3, or Sport. Eco and 1 still performed pretty much as before.

 

[Reid]

The battery has been out in the bike locker (which has been about 32-40 degrees the past couple of weeks). I take it in at night for charging, but have avoided lugging it in during the day at work. I usually experience this several miles into the evening ride home, so even if I took the battery in during the day, it may not make a difference if this is related to temperature. We're going to have a mini heat wave this week (50s feel like 80s after a Minnesota winter!) so maybe I'll notice a difference as the week goes on and it gets warmer.

All good ideas, looking forward to experimentation this evening. Will post observations tonight!

There you go, an added major factor, battery chemical activity reduces with lower temperature.

One thing you may count on: there is no software limiting of the current other than 20A max and also the brick wall of the low voltage cutoff.

As our pack voltage declines our heavy demands for current cause dips of the pack voltage into the low voltage cutoff area. Which is why I set my low voltage cutoff to the 40V minimum although I never run the pack down to anywhere near that low.

Because a battery pack inherently has what is termed internal resistance,

and current through this resistance inherently causes a drop in voltage,

therefore, as our at-rest voltage steadily declines, at some point (OK, thanks Bruce!) like 44V resting voltage, heavy current demand may trigger low voltage cutoff, just when the most power is wanted. Instantaneous dips in to LVC territory really damp the power, shutting off current till the current-turned-off voltage rises again.

Additionally, let's also say that cold effectively increases the battery pack's internal resistance and thus, the voltage sag.

 

[Fran Fabrizio]

Ok, I experimented all the way home and have some answers!

Short version: It is software limited, and Reid was right, lowering the low voltage limit resolved the issue. There might be normal battery charge degradation and cold weather also contributing, but it's clear that the issue I was describing is due to the low voltage software limit.

But something interesting may be going on... read on for long version:

I was in boost level 1 the entire time. I started with the low voltage cutoff at 42V.

At the start of the ride, resting voltage was 48.0V (50% charge). When I started pedaling, the pedal assist consistently caused voltage to dip about 1V, to around 47V. If I then applied the throttle, it would drop another 1.5-2V, to around 45.5-45.0V. Basically, if I used both pedal assist and throttle, I saw a drop of about 3V.

This behavior was consistent until resting voltage got down to around 46V. Then, when I started pedaling, the pedal assist remained very consistent, causing about a 1V drop, to 45V. However, if I tried the throttle it became noticably less powerful, and would only cause about a 1V drop, to around 44V, instead of the 1.5-2V drop it was doing prior. So, total voltage drop with pedal assist and throttle was around 2V, and throttle was sluggish.

Then, when I hit a resting voltage of around 44.5V, I could start pedaling and get the same 1V drop to 43.5V, but the throttle basically had no effect, and the voltage didn't drop any further at all.

Then I set the low voltage cutoff to 40V (resting voltage still 44.5V). At this point, I could pedal assist for a 1V drop to 43.5V and the throttle was instantly restored to full effect, causing a drop to around 41.5V.

My half-baked theory: With my low voltage cutoff at 42V and a resting voltage of 46V, I would have still expected full throttle (pedal takes me down to 45V and throttle only to ~ 43V, well above the cutoff). BUT if the low voltage setting of 42 is really acting like a 42.X cutoff (meaning anything below 43V would trip it), then all of my observations start to make sense. With a resting of 46V, pedaling taking me to 45V, that throttle would cause a momentary dip just into 42.9V territory, cutting it out, and maybe it's cutting in and out intermittently as I hold throttle, which gives me the overall effect of "half throttle" or sluggish throttle.

Then, when I got to resting 44.5V, it makes sense that the throttle basically wasn't working at all, because pedaling takes me to 43.5V, and then at that point I'd only be a short distance from the 42.XV cutoff so the throttle would basically instantly cut out.

The fact that I could set the low voltage limit to 40 and immediately the throttle would cause a dip to 41.5V and give me full power confirms this is primarily a software issue.

So, my plan is to set the limit at 40, which should give me full throttle down to a resting voltage around 44V, and I usually don't get that low on my normal commute. A resting of 44V is > 30% charge, so this seems ok in terms of battery life.

As the weather warms, I'm curious to see if pedal assist causes a less than 1V drop and/or throttle causes a less than 2V drop.

Thanks all!

 

[Bruce Arnold]

@Fran Fabrizio This accords so well with my experience as mentioned above. Thanks for the confirmation.

 

[Chris Hammond]

Great info Fran, Reid, and Bruce. This issue with the LVC is one I hadn't even considered. I really don't plan on using the throttle on my bike. I am planning on setting my LVC at 46V (I orderd the 52V battery and that is just over 20%). It will be interesting to see if I run into performance issues as my battery gets depleted. If so, at least I now know how to fix it easily. Thanks for the info guys.

 

[Bruce Arnold]

Chris, I think if you set the LVC at 46V, you'll start to notice degraded performance at 48V or so, which is only half the battery. However, I am not confident that this is the case. I'm not certain why I think this would be the case. Something along these lines:

With the controller set at 42V, Fran started noticing performance degradation at around 46V. With it set at 40V, I started noticing marked changes at 44V. There seems to be a threshold of gradually decreasing performance at some level above the LVC. Perhaps to protect the battery from too rapid a discharge. Like I say, I'm really not certain about this. Just a guess, not even an educated guess.

So please let us know what your results are.

 

[Fran Fabrizio]

The software developer in me thinks there could be a simple explanation, a bug in the low voltage level implementation code. Voltage is a floating point number, but the low voltage cutoff you enter in the menu is an integer. To compare the voltage to the cutoff, you have to make them the same data type. If you accidentally do something like:

if int(voltage) <= low_voltage_cutoff
  # kill the motor

with a low_voltage_cutoff of 42, a voltage of 42.999999999V would kill the motor and explain why I was experiencing throttle degradation a full volt earlier than expected.

Probably not it, but I want that to be the reason. And it really does fit the observed pattern exceptionally well. Probably just a coincidence. If I'm right I win a free bike, right?

 

[Dunbar]

I believe it’s what they referred to as ‘chug’ on the original Cross Current. As you start bumping into the low voltage cutoff the controller cuts power and you get a lurching sensation to the power delivery. It’s most noticeable on throttle because that puts the biggest load on the battery. You would probably notice similar behavior on assist level S if you were pulling a lot of power from the battery.

Also, you have to realize that 48V resting pack voltage is around 25-30% battery life remaining. My Cross Current gets extremely sluggish in general on the higher assist levels once I get to <25% battery remaining. To the point that climbing even a modest incline becomes challenging.