Loss of Assist as Battery Discharges

6zfshdb

Well-Known Member
I have 3 identical bikes. The 5 PAS settings work normally when the battery is fresh but the level of assist declines gradually as the battery voltage drops. By the time the battery is 70% depleted, PAS 1 is all but useless with little or no assist at all.

I noticed this problem when I bought the first bike. I contacted the manufacturer and they sent me a replacement controller under warranty. The new controller acted the same as the other three.

Considering Ohm's law, this would seem to be normal. However, a well designed controller could compensate for the voltage drop and eliminate this power decline.

This is really an an annoyance rather than a major problem. It happens so gradually I only notice it on long rides. When I do, I simply bump up the PAS level to compensate as the battery depletes.

Have others noticed this issue? Are all controllers designed this way or are my Dapu units just an inferior product?
 

Art Deco

Well-Known Member
I haven't experienced it but was warned of that behavior by my lbs when I bought our bikes. They are Specialized turbo Comos, so not the Dapu controller AFAIK. I guess it is widespread behavior , but you are correct that programming could mask it , if they wanted to.
 

antboy

Well-Known Member
Have others noticed this issue? Are all controllers designed this way or are my Dapu units just an inferior product?
I could be wrong, but I think most hub drives have the same issue. I think some of the "big name" mid drives might compensate for this, but I'm not sure.

My Bafang rear hub performs the same way. If I leave it charging overnight before a long ride to balance the batteries, it's got EXTRA pep for the first few kilometers, enough so that I can use a lower PAS level at the beginning.

It then settles down to what I consider "normal", and the power drop basically follows the typical voltage sag curve, so a very gradual drop between 90 and 40%, and a steeper drop off after that.

A rule of thumb I've been using for longer rides is to head back home when the battery hits about 60ish% (based on terrain/weather - especially wind), to avoid going below 20%, both from a battery health and PAS performance POV.
 

fooferdoggie

Well-Known Member
my bosch if there is a drop on the last mile it is hard to notice. it is consistent through the whole battery as far as I can tell. but it is also battery quality too the lower the quality of the cells the worse this problem is.
 

antboy

Well-Known Member
my bosch if there is a drop on the last mile it is hard to notice. it is consistent through the whole battery as far as I can tell. but it is also battery quality too the lower the quality of the cells the worse this problem is.
I wouldn't blame the batteries in 6zfshdb's case, unless we know that they're not brand name cells. :)

I think it's more about the battery management system and motor type.

Though not EXACTLY the same, it's conceptually like the difference between an old school flashlight with an incandescent bulb is on/off, brighter with a fresh battery that gradually dies when the battery does, vs. a good LED flashlight with multiple brightness settings via a small integrated circuit to essentially maintain the same brightness over the life of the charge.

IIRC (just randomly reading/researching) - not from personal experience), the controllers on mid-drive motors from Bosch etc., regulate the power in a way to maintain a constant draw, and shut down while there's still more of a charge in the actual battery, which would prevent damage, while hub motors will let you run the battery down further, but that last 20-30% is like the dim old-school flashlight. :)

I don't understand enough about the demands of a geared hub drive vs. a mid-drive to speculate why hub drives (generally?) seem to manage the battery differently.
 

6zfshdb

Well-Known Member
Thanks for the replies. I'm glad to hear I'm not the only one who has experienced this issue.

The bike has a hub motor and my batteries contain Panasonic cells. All 5 batteries perform the same. Apparently, mid drive motors use a better quality controller.

I deal with this problem by riding with a spare battery and swap them when the charge reaches 30%. The bike is ride-able below 30% in a pinch but requires using a higher PAS setting.
 

harryS

Well-Known Member
That's normal for batteries. One of my controllers shows 100W in PAS 1 at 40V. When down to 34V, it's about 80 watts, It means I have to supply the missing wattage by pedaling harder. I can feel that.

Yes, a controller could be rigged to have constant voltage, I have done that by putting a voltage booster behind the battery, so it always gets 40V to the controller, even as the supply battery is depleted. You pay for this though. When the battery is down. the booster is sucking extra current out of the battery to keep putting out 40V. Kills efficiency a lot. Probably a 25% loss in total range just to have constant power.
 

Daffyh

Member
Both my electric scooters do this and as mentioned you need a cc/cv controller to eliminate it, neither of my Bosch mid drive bikes do it. Same with my LED flashlights the better ones are fully regulated.
 

Alex M

Well-Known Member
I "think" that declining state of charge not only lowers the voltage but also affects battery ability to generate current. This is why manufacturers don't install CC controllers, to avoid stressing the battery too much.

Bosch motors are low-wattage, they draw very low current, so controller can increase the draw to compensate for declining voltage without much damage.
 
Last edited:

Sierratim

Well-Known Member
I "think" that declining state of charge not only lowers the voltage but also affects battery ability to generate current. This is why manufacturers don't install CC controllers, to avoid stressing the battery too much.

Bosch motors are low-wattage, they draw very low current, so controller can increase the draw to compensate for declining voltage without much damage.
Lithium based batteries have such a low internal resistance that drawing more current to maintain motor torque isn't really an issue. You can still arc weld with a lithium pack near its cutoff voltage, though this can damage the battery in terms of life!

Our mid-drive Specialized ebikes don't demonstrate a drop in motor torque as the battery voltage drops. They must be compensating with more current to maintain motor torque as we ride that last steep hill on the way home, i.e. constant power controllers,

As an aside, many mid-drives are rated 250W nominal power output. Their actual output peaks at well over twice that. This graph has been posted on other threads.

mceclip10.png
 

CdnShaun

Active Member
I started out the last few years with Bionx 48v motors/batteries and like others here, less go-go juice as the battery depletes.

That said the display on my Bionx bikes show me this - Battery level 100% to 40% will give me up to the full 100% assist. Battery level 40% to 20% will limit only to 75% assist. Battery 20% and below limits to 40% assist at most. Like others said, enough to keep riding if no other option, I simply plan ahead and carry extra batteries to offset the distance I am going that day.

My new bikes this year will be 52V configurations. I have read the 52V setups have a lot less sag towards the end of the battery life on a ride - will report back in a few weeks when I can compare the 52V custom build to my original 48V propriety setup.

Of course anyone who can speak to 48V vs 52V would be appreciated in this thread as well.

Cheers
 

Timpo

Well-Known Member
Considering Ohm's law, this would seem to be normal. However, a well designed controller could compensate for the voltage drop and eliminate this power decline.
This issue was addressed by Tora from Juiced Bikes 3 or 4 years ago.

Maybe your controller is fine, but the battery may have low discharge rate. (Max drain current)

What kind of cells do you have?
Each cell is different. Just because they're all 18650 lithium ion, the drain characteristics are different.

SAMSUNG 35E
  • Nominal Capacity: 3500mAh
  • Continuous Discharge Rating: 8A
  • Pulse Discharge Rating: ?
  • Nominal Voltage: 3.6v
SAMSUNG 30Q
  • Nominal Capacity: 3000mAh
  • Continuous Discharge Rating: 15A
  • Pulse Discharge Rating: 25A
  • Nominal Voltage: 3.6V
SAMSUNG 26F
  • Nominal Capacity: 2600mAh
  • Continuous Discharge Rating: 5.2A
  • Pulse Discharge Rating: ?
  • Nominal Voltage: 3.6V
 

Timpo

Well-Known Member
Exactly why us flashaholics pay attention to the cells we use in overdriven LED flashlights, 30Q being a popular choice for 18650.
yeah, I used to have Samsung 26F pack and switched to 35E.

Made a huge difference, despite both being 48V.

I used to think that as long as max current on battery could suffice the max current on controller, there shouldn't be a problem.

But I learned from ebikeschool.com YouTube channel that, realistically, the battery's max current should be far exceeding the controller's max current, and you would not want to be using battery's max current, ever.
 

Daffyh

Member
yeah, I used to have Samsung 26F pack and switched to 35E.

Made a huge difference, despite both being 48V.

I used to think that as long as max current on battery could suffice the max current on controller, there shouldn't be a problem.

But I learned from ebikeschool.com YouTube channel that, realistically, the battery's max current should be far exceeding the controller's max current, and you would not want to be using battery's max current, ever.
Indeed as it will drain super fast and *usually* but not always a rise in tempertaure. There are some cells that can do high drain and retain temperature stability.