Using clone ELM327 Bluetooth OBDII adapter with Leaf

[surfingslovak]

Percent Reserve is a percent of what, perhaps SOC?

A percent of SOC is a variable amount of energy, depending
upon the capacity of the (possibly shrunken) fuel tank
(a reduced Battery Pack capacity).

Judging by Jim's comment above, it's a percentage of the total capacity. Since the e-fuel tank size will decrease over time, the reserve will represent less range in the future and won't coincide with battery events, such as VLB or turtle.

DTE caluation uses remaining kWhrs minus either the % selected or a fixed kWhr as a reserve. Then multiplies it by the Miles/kWhr specified.

 

[TonyWilliams]

The LBW and the VLBW are not at a fixed SOC value.
They are at a constant GID value.

With an original capacity tank,the VLBW might be at about 8% SOC,
but with a tank shrunk to half capacity, the VLBW will probably
be at around 16%. In both cases, it would be at about 8.5 %GIDs.

That's a good point, I forgot about that. Would it make sense to define the cutoff in terms of kWh then? VLB at 2 kWh, turtle at 0.5 kWh and the car is dead at 0.3 or 0.4 kWh.

Since we don't yet have Gids, having absolute reference to LBW and VLB won't line up. Plus, Turtle is a function of low cell pair voltage and seemingly kW draw (or voltage sag under high kW?). The higher the power draw, the sooner Turtle comes on in relation to Gids.

I suspect all these things can be guesstimated with SOC% (that would become more and more inaccurate with degradation of the pack), but I'm confident that Gids will somehow be imported. Finding the Turtle trigger would be fun (again, it's definitely not Gid).

 

[Turbo3]

Just placed an updated version 0.016t for testing on DropBox.

Includes
* separate visible reset button to reset Whr used
* title changed on percent to "% SOC Reserve"
* Added support for Max/Avg/Min battery temp selection

 

[surfingslovak]

Since we don't yet have Gids, having absolute reference to LBW and VLB won't line up. Plus, Turtle is a function of low cell pair voltage and seemingly kW draw (or voltage sag under high kW?). The higher the power draw, the sooner Turtle comes on in relation to Gids.

I suspect all these things can be guesstimated with SOC% (that would become more and more inaccurate with degradation of the pack), but I'm confident that Gids will somehow be imported. Finding the Turtle trigger would be fun (again, it's definitely not Gid).

Yes, agreed, turtle is not indexed to stored energy and is likely triggered by the lowest cell-pair crossing a certain voltage threshold. This would align well with the observation that turtle mode can kick in sooner than anticipated under heavy load. I have definitely seen it as well.

Personally, I believe that the calculated stored energy (kWh) will align nicely with Gids, which could eliminate the need for them in the near term, and perhaps even in the longer term. I too find the usefulness of % SOC reserve somewhat debatable, but this app is at the beginning of its development cycle and some of the concepts will likely be validated (or rejected) by its future users.

Is anyone able to connect a Gid meter and an ELM327 dongle at the same time to help validate the kWh readouts? We would have to use 80 Wh per Gid, and not adjust for battery losses to get comparable results.

 

[LEAFfan]

Gary, the EVS icon came on again, so I doubt it's a seatbelt click thing. Are you still using your EV CAN one? I haven't tried switching to the CC one to see if it will disappear. I've left it plugged in now for two straight days/nights and there's been no change with the 12V.
Jim, after each update, the third BT is not there. I use the Kyocera Event.
Also, if you want to clean up the App a bit, you may want to drop the 'r' after kWh and Wh.

 

[garygid]

Certainly the 80 Wh per GID is not the Usable energy.

If the kWh displayed is not usable energy, what good
is it, if it is just another created estimate, of what?

 

[garygid]

Gary, the EVS icon came on again, so I doubt it's a seatbelt click thing.
Are you still using your EV CAN one?
I haven't tried switching to the CC one to see if it will disappear.
I've left it plugged in now for two straight days/nights and there's been no change with the 12V.
... snip...

No, I have not used the EV-CAN (ECD) dongle very much, but I
will try again. I guess that we do not know if it is "proper"
to Request the Group data on the EV bus.

Is the EVS icon the yellow car with the "!" in it?

I used the CAR-CAN dongle (CCD) for about 10 minutes driving,
and I did not see any problems.

 

[Turbo3]

Jim, after each update, the third BT is not there. I use the Kyocera Event.

Please explain what that means?

 

[garygid]

Nice flexibility with being able to see the Max, Min, or Average
battery temperature.

The people in hot country can choose the Max value, the
people in cold areas might be more concerned about
the Min value, and the selection is simple, and remembred.

Indeed, more good work. Thanks from us all.

Now, to find the tire pressures...

 

[garygid]

I think he means:

When one selects the system menu icon, and the "Select OBDII" option,
the "Choose device" list is difficult to use on a small screen, if there are
more than two options.

 

[surfingslovak]

Certainly the 80 Wh per GID is not the Usable energy.

If the kWh displayed is not usable energy, what good
is it, if it is just another created estimate, of what?

Gary, I think the immediate objective is to find a suitable substitute for Gids, at least in the short term.

I believe we had this discussion last year already, in a thread Phil created. It was then that we learned what Gids represent: it's a measure of stored energy, or in other words, it's the amount that physically went into the battery. Why would anyone do that? Because determining the actual usable energy in a battery at any given moment is a non-trivial problem. Nissan has addressed this by indexing dash instruments to stored energy, or Gids.

Adjustments are periodically made to the Gid count based on the SOC and on coulomb counting. We have the SOC and might be able to do a coulomb count based on the data available. In either case, I think we can emulate the behavior fairly accurately. If you recall, Phil's LeafScan prototype displayed stored energy in kWh as well. Although he did not say how he calculated this figure, I believe there is enough data available from the CAN bus, and we are getting close.

To corroborate further that Nissan made a concious decision to use stored energy instead of usable energy is available online. A representative stated clearly that the low battery warning will come up at 4 kWh remaining and the very low battery warning at 2 kWh remaining. If you multiply the Gid count at each marker by 80 Wh, you will see that this all aligns nicely. However, this decision also highlights why it might be difficult to predict the actual range of the vehicle, even if you disregard the question about the length of the integration interval.

If we cared about actual usable battery capacity, the first approximation there would be to assume battery efficiency of 97%, which was quoted in the NREL teardown report. This would yield usable capacity of 21,417 Wh for a new battery with a Gid reading of 281, and assuming 5 Gid or abouth 400 Wh of residual energy left in the battery pack after the end of turtle mode. That figure is pretty close to what has been observed in new LEAFs. Battery efficiency could and likely will decline over time, courtesy of rising internal resistance.

 

[garygid]

Is there any control over the Group Request Polling rate, or On/Off?

If one is just logging during a 6-hour charge, and only logging
every 60 seconds, there seems to be little reason to do the
Group Requests every second or so.

Even while driving, the bar data is changing so fast that
it is somewhat difficult to get any meaning at a glance.

So, a polling interval, similar to the Logging interval might be nice,
especially if we might be having trouble using the EV-CAN bus.

 

[Turbo3]

Please post a screenshot. I don't see a problem.

Here is one with four items. If you had more just use your finger to move the items up to see ones off the bottom. I must be missing something here?

In fact, if you are upgrading then the preivous mac is saved so why are you even in the Select OBDII menu. It should auto-connect when you get in range.

Just updated version 0.16t so it sets the radio button of the remembered mac address if it is found in the Select OBDII selection list. It does not speed things up since you still need to tap on the item. But if you have two or more OBDII devices it will now mark the last one used. I think Gary asked for this a long time ago.

Here is a screenshot with the green selection mark showing.

 

[sorphin]

Now, to find the tire pressures...

Agreed! My 350z displayed all 4 live tire pressures... every nissan since then has only had the 'idiot light'.. i hate it.

 

[LEAFfan]

Gary, the EVS icon came on again, so I doubt it's a seatbelt click thing.
Are you still using your EV CAN one?
I haven't tried switching to the CC one to see if it will disappear.
I've left it plugged in now for two straight days/nights and there's been no change with the 12V.
... snip...

No, I have not used the EV-CAN (ECD) dongle very much, but I
will try again. I guess that we do not know if it is "proper"
to Request the Group data on the EV bus.

Is the EVS icon the yellow car with the "!" in it?

I used the CAR-CAN dongle (CCD) for about 10 minutes driving,
and I did not see any problems.

Yes, and the yellow triangle also shows in the brow.

 

[TonyWilliams]

Certainly the 80 Wh per GID is not the Usable energy.

If the kWh displayed is not usable energy, what good
is it, if it is just another created estimate, of what?

I'm with Gary on this one. Certainly, we want end user data, and the best (easy) guess so far is 75 or 76 Wh per Gid.

On a human factors note, having the three digit temperature and three digit miles together in the same size and font might cause a few confusions.

 

[LEAFfan]

Jim, after each update, the third BT is not there. I use the Kyocera Event.

Please explain what that means?

I still can't see the correct 3rd battery temp (BT) sensor. It shows a Huge - (minus) number there (-11699.8) instead which is probably why the fourth reading goes off the screen. When I tried 'Average', it was a - four digit number (-2855.). I must be the
only one showing this glitch.

 

[surfingslovak]

Certainly the 80 Wh per GID is not the Usable energy.

If the kWh displayed is not usable energy, what good
is it, if it is just another created estimate, of what?

I'm with Gary on this one. Certainly, we want end user data, and the best (easy) guess so far is 75 or 76 Wh per Gid.

On a human factors note, having the three digit temperature and three digit miles together in the same size and font might cause a few confusions.

Yes, agreed, having usable energy on the display will be better for the end-user. No question about it. The only problem with this approach is that battery efficiency won't be the same for everyone, and it will likely change as the pack ages. What I suggested above was to start with estimated stored energy first. Once we have something that matches Gids relatively closely, then we could move onto something better, if everyone agreed that this was the best course of action.

I remember seeing a comment from couple of days ago about usable energy being a fuzzy number. This was another reason to focus on stored energy, and make that work first. It won't be too difficult to get a usable energy display out of that by applying an adjustable efficiency factor. Phil displayed stored energy in kWh on his LeafScan prototype, and although I didn't like it then, I think I understand where he was coming from.

Having a smartphone app instead of a custom-built device has several advantages. One of them is that it's easier to let everyone personalize the default settings, and make other adjustments to match their car and driving style, as Jim has already demonstrated. It really takes the edge out of these discussions. Would it make sense to define a battery efficiency factor in the preference and set the default to "1" or "100%" for now?

If this factor is set to 95% later, the usable capacity of the pack will be about 21.4 kWh for a new LEAF. This is a bit more than the figure we agreed to use earlier, but it would correspond to what NREL quoted in their teardown report. To get 21 kWh even, the efficiency factor would have to be set to 93%. It's also possible that we are missing something in the equation used to calculate the stored energy estimate.

 

[jackal]

I just want to confirm that the Virgin Mobile prepaid phone LG optimus V ($39.90 from Walmart, Android v2.2) and a cheap 'stubby' ELM that I bought from AliExpress ($8) works.

Thanks Turbo!

Now I need to figure out how to read all those numbers!

By the way, does everyone leave the ELM or remove it from the CAN bus when not in use? Does it causes significant phantom drain of the 12V battery?

 

[LEAFfan]

I just want to confirm that the Virgin Mobile prepaid phone LG optimus V ($39.90 from Walmart, Android v2.2) and a cheap 'stubby' ELM that I bought from AliExpress ($8) works.

Thanks Turbo!

Now I need to figure out how to read all those numbers!

By the way, does everyone leave the ELM or remove it from the CAN bus when not in use? Does it causes significant phantom drain of the 12V battery?

You must have skipped some recent posts because I posted that mine's been plugged in with phone for two days/nights without any voltage drop in the 12V. There's o guarantee your Elm will work. I ordered mine from CA and had it converted to read the EV CAN. TickTock told me that the car's OBD Connector could become damaged with daily plug ins and removals. So I'm going to try and have a switch put in so it can read both busses without removing one and plugging in the other. The one with the four screws can be modified. Garygid posted how to do it.