Capacity vs Hx (Health)

(bah! Top of the page blues. See my last post first: http://www.mynissanleaf.com/viewtopic.php?f=44&t=13247&p=319070#p319070" onclick="window.open(this.href);return false

Voltages on the battery page also seem to coincide with this health figure - knowing the battery is only "feeling" 70% of its former perkiness, I observe the graph dipping under load pretty clearly. (BTW, a pack voltage readout and histogram, made from a sum of the bar graph values, would be very useful. ) The drop in voltage also represents a higher amp load to extract the same number of watts from the pack, versus a healthier battery that has less of a voltage drop at the same load.

That means health also influences drive efficiency (miles/kWh or kWh/KM). When you accelerate, the voltage drops less in a healthy battery. As long as I've got my math straight regarding watts and amps...

Health appears to be - and this is only my opinion - a formula that takes in to account, and is substantially biased toward, internal resistance of the pack...

FalconFour said:

Voltages on the battery page also seem to coincide with this health figure - knowing the battery is only "feeling" 70% of its former perkiness, I observe the graph dipping under load pretty clearly. (BTW, a pack voltage readout and histogram, made from a sum of the bar graph values, would be very useful. ) The drop in voltage also represents a higher amp load to extract the same number of watts from the pack, versus a healthier battery that has less of a voltage drop at the same load.

Click to expand...

We just don't know.. And I guess that was my point in starting this thread. When we figured out how to actively ping the car's lithium battery controller for additional information we found different "groups" of data. Initially we found 5 groups but then Sam found a few more including the 6th which had the cell "shunt" balancing info. Group 2 was clearly the cell voltages and Group 4 was clearly the 4 (or 3 for 2013) battery temperatures. But what were all the numbers in Groups 1, 3, and 5? Jim, Luke and I started poking around to see if we could figure any of it out. Group 1 has almost 40 bytes of data. Using an early version of the ELM app as well as Luke's CANary and some old EDrive DDs loaned to folks in SoCal I looked for differences in the data from a half dozen different cars. We soon discovered the high precision (20 bit) SOC which matched the lower precision number on the regular EV-CAN passive data. Another 20 bit value seemed to be a number ranging from 600000 and 670000 which correlated with max Gids on a full charge and seemed to be a good capacity number given the Leaf's supposed 66Ah battery capacity. This is the Ah Capacity number reported on ELM and LeafDD. Another 16 bit number also seemed to track capacity but was in the (initially observed) range of 7000 to 10200.. Taken as a percentage with two digits of precision I decided to call this "health" for lack of a better term. On the old software and especially for cars with only moderate capacity loss (still 12 bars) it seemed to track the Ah Capacity fairly close. Now that health is reported on LeafDD and the ELM battery app we've been able to track it over a much larger number of cars and seen some strange stuff.. like REALLY low numbers for cars with more capacity loss.. or 105-111% for some 2013 packs with Ah Capacity pegged at 67.3620Ah. What does it all mean? I don't know.

Here's a snapshot of the raw Group 1 data from the 2013 rental I had a few weeks back.. taken from the old 8 line EDrive DD.

The SOC is on the 5th line 0x0B05F1 (722417) and the Ah Capacity is on the 6th (last) line 0x0a4754 (673620) and the Health is earlier on the 5th line 0x2b96 (11158) 111.58%!! This was the new 2013 rental I had with barely 1000 miles on it. Over the course of the 3 weeks I had it the Health dropped to about 108% but Ah Capacity never budged from 67.3620Ah.
My trusty 2012 which just passed 10,000 miles has 60.4897Ah capacity and 90.45% Health (old software).. Back in early April while it was still relatively cool it was at 66.0198Ah and 100.79% Health!

TomT said:

Health appears to be - and this is only my opinion - a formula that takes in to account, and is substantially biased toward, internal resistance of the pack...

Click to expand...

If this is true, why did my health number drop 1% after the 3227 firmware update? Did the internal resistance of the battery increase that much in the one hour it took to update the software? I took readings with the LBA immediately before and immediately after the update. Something else must be involved.

TT

ttweed said:

TomT said:

Health appears to be - and this is only my opinion - a formula that takes in to account, and is substantially biased toward, internal resistance of the pack...

Click to expand...

If this is true, why did my health number drop 1% after the 3227 firmware update? Did the internal resistance of the battery increase that much in the one hour it took to update the software? I took readings with the LBA immediately before and immediately after the update. Something else must be involved.

TT

Click to expand...

Because the formula used to calculate the health value changed with the firmware update.

ttweed said:

TomT said:

Health appears to be - and this is only my opinion - a formula that takes in to account, and is substantially biased toward, internal resistance of the pack...

Click to expand...

If this is true, why did my health number drop 1% after the 3227 firmware update? Did the internal resistance of the battery increase that much in the one hour it took to update the software? I took readings with the LBA immediately before and immediately after the update. Something else must be involved.

TT

Click to expand...

In a few weeks disconnect your battery and see if anything changes.

ttweed said:

TomT said:

Health appears to be - and this is only my opinion - a formula that takes in to account, and is substantially biased toward, internal resistance of the pack...

Click to expand...

If this is true, why did my health number drop 1% after the 3227 firmware update? Did the internal resistance of the battery increase that much in the one hour it took to update the software? I took readings with the LBA immediately before and immediately after the update. Something else must be involved.

Click to expand...

I don't think you can read anything into such small changes in the value, you need to look at long-term trends...

Perhaps the 1% change - which is really not that statistically significant - came about for the same reason that many saw a temporary increase in range and/or GIDs after the reprogram: The BMS was relearning the battery. It may also be that the algorithm was updated...

ttweed said:

TomT said:

Health appears to be - and this is only my opinion - a formula that takes in to account, and is substantially biased toward, internal resistance of the pack...

Click to expand...

If this is true, why did my health number drop 1% after the 3227 firmware update? Did the internal resistance of the battery increase that much in the one hour it took to update the software? I took readings with the LBA immediately before and immediately after the update. Something else must be involved.

Click to expand...

dm33 said:

Because the formula used to calculate the health value changed with the firmware update.

Click to expand...

This. Mine is down more than 8% since P3227. I doubt that it is a real change in the battery: the Ah capacity has dropped only a small amount, similar to the decline before the update.

It is unfortunate we are calling this "health" at this point. We really don't have a clue what it is and I fear calling it health will cause heart-burn with the folk who are not paying full attention to every post on the forum (like me ). All we know is it starts near 100% and reduces with age but at a much faster rate than capacity. Series conductance ratio is a valid theory, but only a theory right now. My leaf is reporting 51.7% for this parameter and I am confident there is nothing I can detect that has degraded 50%. Can I propose until we know more that we call this something else. Perhaps Xfactor, or paramA, or, if we feel we need continuity, maybe "health???" instead.

I like XFactor!

TickTock said:

Perhaps Xfactor, or paramA, or, if we feel we need continuity, maybe "health???" instead.

Click to expand...

drees said:

I don't think you can read anything into such small changes in the value, you need to look at long-term trends...

Click to expand...

I was more interested in why Ahrs went up yet Health/Xfactor went down with the update. GregH was saying "we don't know" what it is but that the Hlth/X number seemed to track with capacity. TomT thinks the number is "biased towards internal resistance of the pack." Certainly, long term trends might be more indicative of what this number really is, but it seems noteworthy that they moved in opposite directions with the update. Is this typical post update? Will my Ahrs drop over the next 3 weeks while Hlth/X rises, or will they both drop, as some have reported?

QueenBee said:

In a few weeks disconnect your battery and see if anything changes.

Click to expand...

You mean the 12V aux./starting battery? What will this accomplish?

TT

ttweed said:

I was more interested in why Ahrs went up yet Health/Xfactor went down with the update. GregH was saying "we don't know" what it is but that the Hlth/X number seemed to track with capacity. TomT thinks the number is "biased towards internal resistance of the pack." Certainly, long term trends might be more indicative of what this number really is, but it seems noteworthy that they moved in opposite directions with the update. Is this typical post update? Will my Ahrs drop over the next 3 weeks while Hlth/X rises, or will they both drop, as some have reported?
TT

Click to expand...

Yeah, I saw the same thing. After everything settled in a few weeks, Ah has increased from 45.6 to 47.3 and Xfactor decreased from 62.4 to 51.2. While we know the series resistance didn't magically change as a result of the update, the algorithm may have. I was told that the main change the update provided was more accuracy at higher temperatures - quite possible this would apply to a resistance test, too (or whatever it is that this field represents).

That's why I wrote up that long couple posts on the last page. :/

Since I discovered "health", which does make sense as monitoring the internal resistance of the cells, I've been changing my driving habits quite a bit - doing less things that I know put strain on the battery. Since I've been gentle to the car, health has improved at a much more responsive rate than capacity, which is also slowly ticking upwards as well.

A histogram would be extremely useful here, using a persistent database of samples to draw a line graph, filling in whenever the app is open (and automatically interpolating missed samples, like the hours the app's not open, just by recording the time with each sample). Constant logging is impractical, and information in broken chronological order would be misleading. I picture this being implemented by maintaining a small Android database that would easily be maintained (binary-wise) by the OS. The logic would be "if previous sample was over an hour ago, take a new sample and record it in the database with the time, then trim log entries older than [1 month; 3 months; 6 months; etc]". The data could be displayed on a line graph with the time on the X axis, health and capacity on the Y axis.

Then we can start comparing trends, not just values

FalconFour said:

That's why I wrote up that long couple posts on the last page. :/

Since I discovered "health", which does make sense as monitoring the internal resistance of the cells, I've been changing my driving habits quite a bit - doing less things that I know put strain on the battery. Since I've been gentle to the car, health has improved at a much more responsive rate than capacity, which is also slowly ticking upwards as well...

Click to expand...

By "improved" do you mean the number has gone up or do you mean that it is dropping at a reduced rate?

TickTock said:

It is unfortunate we are calling this "health" at this point. We really don't have a clue what it is and I fear calling it health will cause heart-burn with the folk who are not paying full attention to every post on the forum (like me ). All we know is it starts near 100% and reduces with age but at a much faster rate than capacity. Series conductance ratio is a valid theory, but only a theory right now. My leaf is reporting 51.7% for this parameter and I am confident there is nothing I can detect that has degraded 50%. Can I propose until we know more that we call this something else. Perhaps Xfactor, or paramA, or, if we feel we need continuity, maybe "health???" instead.

Click to expand...

I agree. Xfactor until it is determined what it really represents.

I'm not sure the Health is directly related to impedance. I will continue to call it health or at least H as that is what it's labeled on the LeafDD. If you need another name, please find something that starts with "H".
On my 2012 with about 90%h the LeafDD shows about 80-90mOhm.
On the 2013 rental with 110%h the LeafDD showed about 50-60mOhm.

Your LEAF is telling you to exercise more and drink less soda. :lol:

GregH said:

I will continue to call it health or at least H as that is what it's labeled on the LeafDD. If you need another name, please find something that starts with "H".

Click to expand...

Hunknown. the "h" is silent.

Hanssens or Hanss.