Print

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

[princec]

Useful info here: https://batteryuniversity.com/learn/article/elevating_self_discharge

It seems that an average of 1% / day might be par for the course. Interesting how it discharges faster the warmer it is and the more fully charged it is.

Cas

[sharagan]

Hi Princec,

thanks for the article, it is interesting. It seems that you understood it a bit wrongly as it states: "5% in 24h, then 1–2% per month (plus 3% for safety circuit)" which is no way near to 1% a day or a drop from 60% to 50% in more than a week.

[princec]

Of course one problem here is that the SOC as shown on the Zero dash is not absolute number, it's been put through a couple of filters, algorithms, and lookup tables and so on until it arrives with a figure. Were I tasked with writing such a bit of firmware I'd be looking at the filtered measured voltage over the last several minutes, the time elapsed since the last known charge of any reasonable duration, the battery temperature, and a great big bunch of lookup table data collected from lab testing.

Cas

[sharagan]

Even if there are any claimed algorithms and filters in play to show an approximate or absolute number, it does not change anything about the fact that the SOC is decreasing more rapidly than it is normal for this type of battery.

[Auriga]

For the 2020+ models Zero had to add an always on OBD power supply and sometimes a second OBD port to meet European regulations. If you have one of those, that may increase battery drain. I don't know if they changed all bikes or just European ones.

[princec]

Even if there are any claimed algorithms and filters in play to show an approximate or absolute number, it does not change anything about the fact that the SOC is decreasing more rapidly than it is normal for this type of battery.

Well, that's exactly the opposite of what I'm saying... all you're seeing on the dash is a number that the software is coming up with. Think about it: how can the bike carry on going for a little while even when this number reaches 0%, indicating that it is totally empty?

Cas

[sharagan]

Because it is programmed to show 0% at a certain voltage , but the accuracy of that figure is beside the point. My concern is that at this rate the battery looses around 30% of its charge in a month parked in a heated garage.

[TheRan]

The SOC is based on the voltage (as shown in the Unofficial Manual). The only filtering done on it will be to take an average reading over time so it's not constantly jumping up and down as the voltage drops under load and then recovers. The reason the bike can still ride at 0% is because, as Sharagan says, that just correlates to a certain voltage. That doesn't mean there's absolutely no energy left in the battery, it's just the lower safe voltage limit. For there to be any voltage at all there must be usable energy left.

[Richard230]

Even if there are any claimed algorithms and filters in play to show an approximate or absolute number, it does not change anything about the fact that the SOC is decreasing more rapidly than it is normal for this type of battery.

Well, that's exactly the opposite of what I'm saying... all you're seeing on the dash is a number that the software is coming up with. Think about it: how can the bike carry on going for a little while even when this number reaches 0%, indicating that it is totally empty?

Cas

I once rode my 2018 S with PT down to "00" SOC and then it ran another six miles and got me home. I don't know how much further it would have gone before I was pushing. On my 2014 S with PT I did the same thing and that bike stopped dead when the display showed "0" SOC. By turning off the ignition and turning it back on again I was able to travel another 100 feet before the power shut off again. I did that four times before arriving home. Personally, I prefer the 2018's programming that allows a little bit of SOC cushion. 

[princec]

Hm yes my point is the number on the dash is a calculation not an absolute reading. Just because the number on the dash drops by 5 does not mean you have lost 5% of the actual state of charge in the battery. I do believe there is probably more going on than a simple interpolation over a voltage LUT. For all we know it's hardcoded to reduce the number by 1% a day whatever voltage it reads. I speak from the experience of conking out on an Energica that claimed to have 60% charge after just 2 miles of riding: the voltage clearly goes up slightly if it's left standing for a while, making the system assume 60% charge when in fact the bike had done 108 miles just a couple of weeks prior and hadn't been charged since.

Cas

[sharagan]

The claim of 1% a day being hard coded is just an assumption. Your further observations of the Energica are probably not relevant as it is a completely different bike. I have experience with electric motorcycles from KTM and these do not have a drop in SOC even after 6 months of storage, so it is something completely different aswell. Before starting this thread I called Zero and asked if the drop is normal, I was told that a drop of 2-3% per month are normal and that I should contact the dealer for further analysis. If it would be hardcoded as you assume, they would tell me that I am fine and that I am observing a completely normally functioning bike. Furthermore if your claim of an of the SOC percentage being just a number was true, we would observe random numbers not a steady decrease, also the BMS would not be able to stop charging when full and also the mileage prediction of the bike would not be accurate. So far when riding the percentage of the SOC was in perfect correlation with the miles I did on that charge. Therefore I can not agree with you on your theories.

[princec]

I make no claims whatsoever. I am proposing theories which might fit, with corroborating evidence.

Cas

[Crissa]

Yeah, if 0% was really 0 then the battery couldn't be recharged.

But that the SoC is a calculation is irrelevant to observations made using the same calculation.

-Crissa

[sharagan]

I see far fetched theories with no evidence at all

[princec]

I write firmware for electronics and we do this sort of shit all the time... and I see plenty of evidence for it, which is what I outline in the theories. I mean, I am just as likely to be wrong, but so far I don't see any alternative theories that explain it.

Cas