Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Pard]

Limp mode...

I have been getting Limp mode battery low voltage messages when below 30% lately.  It clears quickly back to normal if I take it easy on the throttle.  My riding must be causing some cells to sag.


Never used to happen. 

Then again, I am riding the bike less frequently these days, so balancing charge events occur less frequently.

I store it between 70% and 80% between rides.

I Start the ride at 100% charged and balanced. 

Millivolt difference between high and low cells in the teens after the full charge. (2023 Eva Rebelle, data from engineering screen)

7k miles on the bike.

Any strategies to make this issue less prevalent?

[PWM]

You potentially have a weak cell that sags under load and limp mode is a protection measure as you already know...a weak cell has high internal resistance and nothing you can do about it...

This cell will charge the fastest out of the bunch so when the BMS senses it is full, the BMS will invoke balance mode...it is very important you monitor when balance mode starts.

On iterative charge cycles, allow system to balance and compare the starting point (in SOC %) when balancing starts.

If this is simply extreme out of balance condition only, then the cell should recover and the balance start should delay further into the cycle...something in the 92-96% balance start seems to be normal.

If the balance start doesn't budge and is consistently <90% when it starts then this confirms a weak sister exists in you pack.

[Pard]

You potentially have a weak cell that sags under load and limp mode is a protection measure as you already know...a weak cell has high internal resistance and nothing you can do about it...

This cell will charge the fastest out of the bunch so when the BMS senses it is full, the BMS will invoke balance mode...it is very important you monitor when balance mode starts.

On iterative charge cycles, allow system to balance and compare the starting point (in SOC %) when balancing starts.

If this is simply extreme out of balance condition only, then the cell should recover and the balance start should delay further into the cycle...something in the 92-96% balance start seems to be normal.

If the balance start doesn't budge and is consistently <90% when it starts then this confirms a weak sister exists in you pack.

Excellent info, and thanks!

I guess I can't use my extended battery warranty anymore!   BUMMMMER

[Pard]

My balance charge always kicked in at around 90%.  Even the last charge before last ride.

Engineering screen showed Mv imbalance of 12.  Same as usual.

Now while charging, at 75% currently, the imbalance is close to 40 Mv at 15A charge rate.  Imbalance drops to 10Mv at a 5A charge rate.

Question is if the weaker cell is pathologically weak at this point as you indicated.

Will charge at a slower rate from 75% to 100% and see where the balancing kicks in and what the final imbalance seems to be.

Thanks gain @PWM

[PWM]

Indeed...run your charger at much lower rate than normal...as you approach 90% throttle down, you will gain improvemt on pulling up the weak cell...5Amp setting versus 15Amp in other words...keeps the delta balance mV voltage lower hence more charge is able to feed the weak cell.

[Pard]

Indeed...run your charger at much lower rate than normal...as you approach 90% throttle down, you will gain improvemt on pulling up the weak cell...5Amp setting versus 15Amp in other words...keeps the delta balance mV voltage lower hence more charge is able to feed the weak cell.

Excellent!  Thanks!

[Pard]

Observations from the engineering screen data:

Balance charged to 100% SOC before a ride today. 

Start of ride, voltages after balance charge to 100%,  4.233/4.250.  Typical imbalance since new.
I kept the engineering screen open for a 100 mile ride down to 18% SOC.

Upon getting underway, voltage between high and low split anywhere from .5-1 tenth of a volt.  4.1/4.2 at 100% charge when I fist started riding.
At 1 mile, SOC dropped to 99% SOC with that pattern continuing throughout the ride, about 1% drop per mile ridden

The splits between high and low cell voltage imbalance would widen at the extreme power demand scenario, up to 2 tenths of a volt when 4.0/4.2 and would narrow to 50Mv when at cruise

SOC seems to be determined by an average of the high and low cell voltages while underway at cruise throttle, with 4.2 being 100% and 3.2 being 0%. 

I confirmed this general formula to be true at 100% down through 90%, 80%, 70%, 60%, etc.20%  SOC while riding.

In other words, an average voltage between high/low of 4.0v while riding will show an 80% SOC on screen.  The bike’s computer does not seem to use the SOC reading when the bike is at rest, because that voltage is generally anywhere from 1-2 tenths of a volt higher than while underway at cruise.  For example, 4.0v while underway will likely show 4.1v when stopped.

The SOC calculation and display by the computer acts with a lag. 

Like the workings of fuel float gauge in a fuel tank showing average fuel quantity over a time period, as opposed to instantaneous fuel level. The SOC display is smoothing out the dips and spikes in voltages over a period of several seconds before it changes.

I did not attempt to trigger a limp mode event, and did not have one today.

I did see voltages drop as low as 3.1 volts while accelerating when the battery was below 20% SOC. 

I suspect that if hard acceleration or high power demand is attempted while the battery SOC is close to 3.2v on the low cell reading, it could drop below 3.0 which must be the threshold to trigger a limp mode condition.

I noticed that medium regen will add a tenth of a volt, but only temporarily, to the reading on the screen when the throttle is released from cruise. I did not notice any situation where the regen introduced enough charge to move the reading on the battery cell voltage higher and stick there once power was reintroduced.

At the end of the ride, with an 18% SOC on the screen, bike on, motor off with kickstand down, my battery balance was excellent 3.42/3.53 V.

I kept the engineering screen on while I started a recharge at 15A and noticed the cell imbalance jumped to 40Mv between high and low while charging.

Conclusion:

I don’t think my battery is bad, but I do think it may have degraded somewhat over the past 2 years and 7K miles.

 I never had to even think about limp mode or being judicious with throttle application below 30-40%, but I will be mindful of it from now on.

[PWM]

Excellent report...good baseline procedure we should all follow...I think you hit all the important elements in your assesment of battery pack.

For NCM chemistry, industry norms are 4.2V for high-side cutoff (charging) and 2.75V for low-side cutoff and likely where Energica sets the threshold for limp mode.

[Specter]

my bike sat for ehh 3 months w/o being ridden or charged,  I did not turn on lpr, probably should have.
first day back at the track after that and it went to balance at like 82 percent.  i ran it all the way down to 12 percent took it to 80 then slow charged it overnight at 2 amp of 115vac to bring it up.  It's sitting at 99 percent when it finishes the charge.  I call it finish because it says its still balancing but will sit there for HOURS and no real current is going into the bike from the wall.

Ill probably have to cycle it a few times to get it happy again but I did not notice any real change in performance.  It still wants to lift the front end at 120 if I romp it too hard, so there is that......    the software on a race bike is a lot different than a street bike however I am not sure if there are any major changes to the balancing side of it or back end of it, but just the 'use' side of it.

aaron

[SBK74]

@Pard

Compared to pre-2023 Energicas, yours has a wider voltage range and more capacity (I have 54Ah max). Mine stops balancing at 4,140-4,152 V. Lowest seen voltage around SOC 0% is around 3,4xxV. I ride a lot and have low imbalances, around 12mV full to 10mV mid-SOC. After summer holidays, inbalance was worse and needed weeks or riding to recover to old levels. Maybe due to more infrequent riding, your battery was now colder than before and (IMO), gave more voltage sag under load, causing the early limp mode.
 

[Pard]

Excellent report...good baseline procedure we should all follow...I think you hit all the important elements in your assesment of battery pack.

For NCM chemistry, industry norms are 4.2V for high-side cutoff (charging) and 2.75V for low-side cutoff and likely where Energica sets the threshold for limp mode.

Interesting.  Did not know the voltage could drop down that low.

[Pard]

@Pard

Compared to pre-2023 Energicas, yours has a wider voltage range and more capacity (I have 54Ah max). Mine stops balancing at 4,140-4,152 V. Lowest seen voltage around SOC 0% is around 3,4xxV. I ride a lot and have low imbalances, around 12mV full to 10mV mid-SOC. After summer holidays, inbalance was worse and needed weeks or riding to recover to old levels. Maybe due to more infrequent riding, your battery was now colder than before and (IMO), gave more voltage sag under load, causing the early limp mode.

Your theory may be correct.

BTW, if you keep the bike powered on while charging and let it sit after charge is complete, while set at a 1amp charge rate, you can get the battery to have extremely low imbalances.  I was able to get the imbalance to be 2mV by keeping it  connected for 24 hrs.  The engineering screen shows the bike drawing down the high volt cells and charging the low side very slowly.  I wonder how much closer the values could get, but did not have the patience to keep the experiment going, and I suspect it does more harm than good.

[Pard]

Happened again today.  Start of ride the battery was balanced at 100% SOC.

At 29% SOC passing a car on the highway,  70 mph uphill, so a solid demand on the battery, the bike went into limp mode.  Recovered seconds later when I eased off the throttle.

Now I am thinking my battery has degraded in a material way.  Only thing I can do is be more careful with the throttle as I get to lower SOC%.

Such a fun bike to ride.  No service options to get a battery diagnostic test.  That extended battery warranty not much good either. 

Read a few reports (FaceBook) of similar issues that resulted in a warranty battery replacement, back in the day when Energica was in business...

[flynnstig82r]

I had similar issues with my SS9 13.4, especially in cold weather (45 deg F or colder). I never went into limp mode at 29%, but I did at 17% and also once had the bike suddenly flash 0% SOC from 8% and shut off. It came back on after a while and I carefully rode at 25 MPH the last few miles in the breakdown lane. It also shut down during overnight charging a few times, which made me late to out-of-town appointments by adding a charging stop. I always figured it was just the earlier bike's BMS not being as well sorted-out as the newer ones, but maybe I had some weak cells.

[Specter]

Pard, energica usa IS in business doing what they are allowed to do.
They will still fix bikes as they have the parts to do so.

It's a bit more complicated than that.  Ok lets say you have a bad cell.
What you going to do?  replace that one bad cell with a new one?
The rest of the cells are a few years old and have their degradation on them as well,
now you'll have a hot cell and the rest of them will have issues.
No matter the technology it's a really bad idea to put a brand new battery mixed with older batteries, especially in a series arm.
You might get away with it on a purely paralled circuit as voltages will tend to 'fix themselves' as to which battery draws how much but in series in a pack.  bad juju

The best way to fix it wouild be to find a cell that is about as old as the rest are, ie the pack is, and ideally been abused the same way.  You could run a charge / discharge curve and see the characteristics of it, find one that is as close as possible and put it in there.

It wouldn't hurt to call them and see if you are still under warranty.

worse case, they have incredible deals on experias now.  IM really thinking of getting one.

aaron