Reduced power at high PAS on Aventon Level.2

[Jeremy McCreary]

Searched the forum, but no clearly pertinent posts...

Friend's new torque-sensing 2023 Level.2 worked perfectly for the first 100 miles or so. But on yesterday's ride, all higher PAS levels suddenly felt like PAS 1/5, which still felt normal to him. Throttle response and display were also normal.

Rode it myself to confirm. When I pushed hard on the pedals with PAS 5/5 showing on the display, the power would ramp up for a split second, then surge with my power strokes for 3-4 crank rotations, then fall back to the power I was getting in PAS 1. Repeatedly.

Similar behavior at PAS 3 and 4. Not sure about PAS 2.

It all started when his battery got down to 20% for the first time ever. The bike snapped out of this spell once thereafter, but not for long.

Toggling power didn't help. All connectors I could find were tight. No visible cable damage. Brake levers and brakes were functioning normally. Didn't think to check motor temperature.

Thoughts? Thanks!

 

[spokewrench]

How is it on throttle?

Can you swap batteries? Maybe the battery can't keep up with higher demands.
When I charge, I plug a Kill-A-Watt into the wall, a timer into the Kill-A-Watt, and the charger into the timer. With a Rad battery, if the green light comes on the charger and the meter reads 2.0 watts or above, I figure it's balancing and leave it on.

At 60 miles, my Abound is still at 65% after 1 charge, so I'm a little foggy. I think I charged the Abound battery the same way as a Rad battery. Balancing the Level 2 battery might improve performance and extend its service life.

 

[Jeremy McCreary]

Throttle operation's been normal throughout.

The problem hasn't recurred since my friend charged the battery — though it hasn't been below 50% since the incident.

No extra battery to swap. And the bike's little more than 6 months and 100 mi old. Could the battery be unbalanced already?

I'm going to recommend that he contact his LBS/dealer and/or Aventon support and ask if his "brown-out" above lowest assist could be normal behavior below 20% charge?

Certainly doesn't seem normal to me. I can see reducing power output at 10% as a low-battery warning/defense. But at 20%??

His Aventon and my 2023 Surface 604 V Rook are very similar in many ways. I've ridden mine down to 5-10% several times, and it"s never done anything like that.

 

[spokewrench]

Ah, 100 miles in six months... I may be onto something!

In the 7 months after I bought my Radmission, I put about 10 miles on my Radrunner. Both bikes have aftermarket displays that read volts. When I took the Radrunner out in May, the voltage was lower than before. That would be self-discharge.

I hadn't been more than a mile when demanding more power would cause the output and the displayed voltage to drop sharply.

My battery had four "stacks" of 13 cells in series. If self-discharge had been the same in all cells, the static voltage reading would have meant they were all half charged. Cells don't self-discharge equally. The weaker ones might produce 2 amps but couldn't produce 10 amps except briefly. When I wanted 10 amps (350 watts of mechanical output), after a couple of seconds, voltage and power would drop. Pushing current through the cells that couldn't keep up would probably damage them.

After charging to the green light, it took many hours for the charger's input power to drop below 2. I figure that was the balancing process, trickling current to the undercharged cells. The battery ran like new after that. (For riders who charge weekly, a little unintended extra charge time each week may maintain balance.)

Having charged my Abound battery only once, I don't remember if the charger has a green light, but I left it on until the input power was below 2.0 watts, like a Rad battery. I knew the cells could have become unbalanced in the time between manufacture and arrival at my house.

 

[spokewrench]

I finally measured what my timer alone draws: 0.6 watt. Normally, my Rad charger tapers down to a combined reading of 1.7 Watt into the timer, so the charger draws about 0.9 Watt when balancing is done.

 

[Slaphappygamer]

I experience low power when my battery gets low. I usually charge at 46v. Voltage sag can drop to 44v.

For example, if the controller can only provide 20 amps max, then at 44v, total wattage is 880 watts. Where as 20 amps at 54 is 1080 watts. Doesn’t sound like much, but the PAS will all be lower wattages as well.

 

[Jeremy McCreary]

Appreciate all the suggestions. Probably don't apply to my neighbor's problem, though, as throttle and PAS 1/5 stayed pretty normal. The only issue, that one time, was that all higher PAS levels felt like PAS 1. And now everything's fine after charging without rebalancing.

This one-time occurrence happened to occur the one time his battery got below 20%. But temporal correlation and causation being 2 very different things, we don't really know yet that the lowish battery was the root cause.

 

[spokewrench]

I, too, like to charge around 46. It's partly because I like the crisper response and partly to spare the battery. I don't know much about my Abound controller or battery, but another bike has a 35 amp controller and four "stacks" of cells in the battery. That would be nearly 9 amps per cell. I think a cell can supply that with less wear and tear if it's at least half charged. The other bike has a 25 amp controller and a battery with three stacks, so that would be more than 8 amps per cell.

 

[spokewrench]

Appreciate all the suggestions. Probably don't apply to my neighbor's problem, though, as throttle and PAS 1/5 stayed pretty normal. The only issue, that one time, was that all higher PAS levels felt like PAS 1. And now everything's fine after charging without rebalancing.

This one-time occurrence happened to occur the one time his battery got below 20%. But temporal correlation and causation being 2 very different things, we don't really know yet that the low battery was the root cause.

Yesterday you said you wanted your friend to contact Aventon Support. Today you call it a one-time occurrence. You say it happened the one time the battery got down to 20% and never any other time. Do you call that coincidence?

A 20% reading reflects the voltage of a stack of 13 cells in series. After only 100 miles in 6 months, I'd expect the cells to be unequal, meaning some could be near 0%. They might get by at a low current demand (PAS 1) but not keep up with higher demands. I use my throttle sparingly. Did you push his down all the way and hold it there?

The problem could have "fixed itself" because I doubt it's possible to recharge without also doing some balancing. I've found that normal recharging, stopping when I see a green light, isn't always enough, so I monitor the input watts.

 

[Jeremy McCreary]

Yesterday you said you wanted your friend to contact Aventon Support. Today you call it a one-time occurrence. You say it happened the one time the battery got down to 20% and never any other time. Do you call that coincidence?

I call it a temporal correlation possibly but not conclusively related to the sub-20% state of charge. I'd have to see (or read here about) more instances to start thinking there's a causal link.

However, if we were to learn that it's an intentional controller feature, we'd be done. So still hoping he calls.

Main concern is getting caught much farther from home thinking it was a battery charge thing when it's really something else.

 

[Jeremy McCreary]

I call it a temporal correlation possibly but not conclusively related to the sub-20% state of charge. I'd have to see (or read here about) more instances to start thinking there's a causal link.

However, if we were to learn that it's an intentional controller feature, we'd be done. So still hoping he calls.

Main concern is getting caught much farther from home thinking it was a battery charge thing when it's really something else.

Addendum: My apologies for losing track of an important clue mentioned in the OP. My neighbor just reminded me of it:

During the epidode, he'd be in normal-feeling PAS 1 and then select, say, PAS 3. For a moment, the assist power would increase, then ebb and surge a few times in rapid succession, and then fall back to PAS 1-like assist power. Pretty fishy for an intended behavior.

He's going to call support. I'll report what they say.

 

[spokewrench]

Addendum: My apologies for losing track of an important clue mentioned in the OP. My neighbor just reminded me of it:

During the epidode, he'd be in normal-feeling PAS 1 and then select, say, PAS 3. For a moment, the assist power would increase, then ebb and surge a few times in rapid succession, and then fall back to PAS 1-like assist power. Pretty fishy for an intended behavior.

He's going to call support. I'll report what they say.

Did Aventon tell your friend anything?

 

[Jeremy McCreary]

Did Aventon tell your friend anything?D

Sent a message to their support on his behalf 2 days ago. Nothing back yet.

 

[PCeBiker]

During the epidode, he'd be in normal-feeling PAS 1 and then select, say, PAS 3. For a moment, the assist power would increase, then ebb and surge a few times in rapid succession, and then fall back to PAS 1-like assist power. Pretty fishy for an intended behavior.

I think that's exactly how the controller would operate before it goes into limp mode.

That's how my car goes into limp mode.

It tries a few times then after certain parameters are met it goes into limp mode and stays there until new parameters are met or the ECU is reset.


"PCM counts the number of misfires that occurred at 200 crankshaft revolutions and 1,000 crankshaft revolutions and calculates misfire ratio for each crankshaft revolution. If the ratio exceeds the preprogrammed criteria, PCM determines that a misfire, which can damage the catalytic converter or affect emission performance, has occurred."


I think that his controller tried to pull the necessary amps a few times and the voltage sagged enough to go into limp mode.

Of course, I'd want to know for sure so I'd want an answer from Aventon.

 

[Jeremy McCreary]

I think that's exactly how the controller would operate before it goes into limp mode.

That's how my car goes into limp mode.

It tries a few times then after certain parameters are met it goes into limp mode and stays there until new parameters are met or the ECU is reset.


"PCM counts the number of misfires that occurred at 200 crankshaft revolutions and 1,000 crankshaft revolutions and calculates misfire ratio for each crankshaft revolution. If the ratio exceeds the preprogrammed criteria, PCM determines that a misfire, which can damage the catalytic converter or affect emission performance, has occurred."


I think that his controller tried to pull the necessary amps a few times and the voltage sagged enough to go into limp mode.

Of course, I'd want to know for sure so I'd want an answer from Aventon.

Very interesting take on this. But limp mode at 20% charge? Seems overly conservative.

 

[PCeBiker]

Very interesting take on this. But limp mode at 20% charge? Seems overly conservative.

My BMS shuts me down at 40 volts but it will kick in when my battery is above 42 volts because of voltage sag when I'm only using ~200 Watts from my 48V 25AH battery.

I have a 25 Amp controller that can put out over 1000 Watts at 42 volts.
I'm sure my BMS would start shutting me down closer to 44 volts if I tried to pull 1000 Watts out of it.


Having the Aventon put you into limp mode at 20% doesn't just help protect the battery, it helps insure that you make it home.

It's like when your low fuel light comes on in your car only the ebike does something about it and stops you from wasting gas.

 

[Jeremy McCreary]

Having the Aventon put you into limp mode at 20% doesn't just help protect the battery, it helps insure that you make it home.

It's like when your low fuel light comes on in your car only the ebike does something about it and stops you from wasting gas.

Understood, but I'd be pissed if my bike pulled that nanny stuff on me at 20% charge.

Granted, a simple warning at 20% would be nice, but my non-Aventon ebike goes down to 5-10% charge with no such limp mode, and I like it that way.

 

[PCeBiker]

Understood, but I'd be pissed if my bike pulled that nanny stuff on me at 20% charge.

My ebike does something a bit similar in that both the speed and Watts drop slowly along with battery charge.
I can see and feel the bike getting tired.
My e-bike doesn't quit and home.

,.. my non-Aventon ebike goes down to 5-10% charge with no such limp mode,

I've taken my 's down to 0% until my BMS shut me down.
No such limp mode. It just died.

Try it it's fun and you get to study the limits like calculus.

and I like it that way.

I kinda like the way my e-bike gets sleepy .

I can feel it getting tired and know better when to head home.

 

[spokewrench]

Sent a message to their support on his behalf 2 days ago. Nothing back yet.

I take that to mean your friend couldn't get through by phone. I've read that Aventon has fewer than 50 employees in America. Radpower's tech support must be much bigger, but they ignored me when I asked where to get spokes.

After 6 months of very little use, my Radrunner behaved like the Level, suddenly losing power although the state of charge seemed to be about 50%. A watt meter guided me to leave the charger on until the BMS finished balancing. That solved the problem.

I recharged my Aventon battery from 60%. As before, the green light came on at 20 watts. As before, there was a long taper until the BMS shut off, leaving only the 0.4 W used by the charger. I think the charger is designed to show a green light at 20 watts (maybe 90 ma per cell) because a lithium cell will have a longer service life if not charged to 100%. OTOH, the BMS can't top off low cells before other cells are fully charged. In the future, I guess I'll shut off the charger when I see the green light, unless I'm concerned about imbalance, as in the case of a bike that hadn't been charged much in a year.

I think its the BMS and not the controller that caused the cutbacks. The controller could measure the voltage of several 13-cell stacks in parallel and estimate 20%, but the BMS could monitor individual cells.

 

[Jeremy McCreary]

I take that to mean your friend couldn't get through by phone.

No, he just got busy with other things, and I'm impatient. I still think his bike needs service, we ride often, and I don't want this to recur when we're miles from home again. He kept falling way behind during the incident — even though I was at lowest assist.