[wizkid057]

Greetings,

So, I've been tinkering with the ELM327 and some custom software for a bit, and I decided I should try to calculate out how much power some of the things in the car actually use by watching the battery SOC precisely over time.

This is one of the things I've been interested in since I purchased the car so it would be possible to estimate how many miles per charge certain items "cost" in range.

This is using CAN ID 206 assuming a value of 0.25 Wh per unit. I calculated that the car uses about 375 watts while "idle" on average.


Here is a list I came up with, subtracting the idle usage of 375 watts from all values and rounding to the nearest 5 watts:

• Rear Defrost - 475 watts
• Climate, Fan Only, Full (Climate Power: 17%) - 270 Watts
• A/C, Eco, LO, Auto (Climate Power: 34%) - 1,535 Watts
• A/C, Comfort, LO, Auto (Climate Power: 59%) - 1,875 Watts
• Front Defrost, 72F, Auto, Comfort - 1,425 Watts
• Heat, HI, Full Fan, Comfort (Climate Power 75%) - 6,250 Watts
• Sitting in D with foot on brake (0 MPH) - 0 Watts
• Creeping forward in L - 400 Watts
• Pedal to Floor in Sport Mode for a moment while stopped - 49,405 watts

(Its interesting to note that the "Climate Power" percentage seems to have no real meaning...)

I also attempted tests for the radio, heated seats, and headlights, although these items seemed to have a negligible effect on the HV battery and I was not getting values outside of my idle power margin for error.

Now, lets assume you get 40 miles out of a charge which uses 10.5 kWh (which is about my average so far after 10,000 miles). This gives us a nice value of 262.5 Wh per mile for travel. Now, this is taking into account the idle usage of about 375 watts already.

Lets do some more estimating and say you'll be traveling at 55 MPH constantly for this charge. We'll just write off starting and stopping for sake of simplicity.

That gives you 43.64 minutes of EV driving. 272 Wh will be used by the idle components, the rest for momentum, 10.23 kWh. That's 256 Wh per mile for momentum.

Taking this further, lets say you were to run the rear defroster the entire trip @ 475 watts. for 43.64 minutes that would consume about 345 Wh, or approximately 1.35 miles of momentum.

Lets apply this logic to the list above and see what we come up with. I know the numbers aren't perfect because I'm assuming the additional load will be running in addition to the 40 miles for 43.64 minutes, but, it should be pretty close either way.

(Edit: This list has been updated based on the formula in this post.)

• Rear Defrost - 475 watts - 0.95 miles miles (335 Wh)
• Climate, Fan Only, Full (Climate Power: 17%) - 270 Watts - 0.4 miles (190 Wh)
• A/C, Eco, LO, Auto (Climate Power: 34%) - 1,535 Watts - 3.55 miles (1,020 Wh)
• A/C, Comfort, LO, Auto (Climate Power: 59%) - 1,875 Watts - 4.3 miles (1,220 Wh)
• Front Defrost, 72F, Auto, Comfort - 1425 Watts - 3.29 miles (950 Wh)
• Heat, HI, Full Fan, Comfort (Climate Power 75%) - 6,250 Watts - 11.85 miles (3,200 Wh)

So, as you can see, different items certainly have an impact on EV range, some quite substantial. Especially that dang high voltage heater.

Now, granted, this assumes a constant "on" of these items, which isn't always the case, but, it should give a good idea one what uses the most juice.

----

I also did some readings for various methods of charging. These are pretty interesting.

• Charging @ 240V - +2,990 Watts
• Mountain Mode Engaged while below threshold - +17,440 Watts
• Mountain Mode Engaged with accelerator depressed - +18,235 Watts
• Hood Popped, Idle - +6,850 Watts
• Hood Popped, accelerator depressed - +17,440 watts

As you can see, the ICE can dump a lot of power into the battery. I'm not sure what kind of impact these types of charge events have on the life of the battery, considering they charge it at over 5x the rate of a 240V 3.2kW charger, but, it is interesting. Theoretically, you should be able to have a dead HV battery, pop the hood, press the accelerator, and have a fully charged battery in 37 minutes. (I'd be afraid to know how much gasoline that would use, but, that's not the point...) This can't be good for the battery, but, proves that faster charging is, in theory, possible with the existing setup.

I also find it interesting that the owners manual says that the battery is neither charged nor discharged while the hood is open and the engine is running... this is obviously false.

I hope someone finds this information useful. I'll update as I tweak my software and get either more or more accurate readings for different things. Any suggestions on what to measure next and I'll do my best to try it.

-wk

EDIT: I'm sure this information would be more easily obtained with the DashDaq, however I don't own one.

---

Edit Log

• 07/10/2012 - Update list of range impacts with new formula.

 

[wizkid057]

I would be interested in knowing what the max regeneration is. shifter in "L" coasting down a steep incline.

Several folks have indicated that the battery system is more than capable of being fully charged in 30 minutes without damage, that would be about 20KW, not too much more than your observation of 18KW in Mountain Mode with accelerator depressed.

This was actually on my list of things to test next, but, I need to automate my software bit more for safety, or draft my girlfriend as button pressing monkey...

If I were to bet I'd say that the regenerative power could break the 20kW barrier, but, I'm not 100% sure yet. I may work on that code and test later tonight or tomorrow.

On an off note, I would love an 18kW grid charger... 75A @ 240V (I drive ~2000 miles per month)

 

[saghost]

I would be interested in knowing what the max regeneration is. shifter in "L" coasting down a steep incline.

Regen power in L is speed dependent. The most I've really seen is around 40kW at freeway speeds.

The numbers in the original post are within error of what I've seen from DashDAQ, with a couple caveats:

The HVAC numbers (including front defroster) are highly variable depending on outside temp and set temp - I have seen over 5 kW from just the A/C side when starting the car after heat soak, and even in comfort it can draw almost nothing (fan power at low fan speeds) if the ambient and cabin temps both match the set temp.

When floored and stopped (foot still on brake,) I only saw about 3 kW draw. It ramps very quickly once you're off the brake and the car starts moving, of course...

 

[RScott]

I hope someone finds this information useful.

Definitely! That's one of those things I wanted to do, but have never gotten around to (and probably wouldn't have for a while, either). Very useful information.

 

[tboult]

Very cool/informative wizkid..


@john
I do PID pooling, which is clearly not as detailed as the measurements that wizkid057 is getting. My measurements on the HV are quantized in reporting to increments of about .062 kWh.. (guess its time to give up on PIDs and start on monitoring.. wizkid you doing PC, iphone, android or mac?)


But with respect to the I just did a measurement down a 305ft drop in L at about 25mph and the regen 0.314 kWh and 0.377 in just over 3min.

 

[wizkid057]

The numbers in the original post are within error of what I've seen from DashDAQ, with a couple caveats:

The HVAC numbers (including front defroster) are highly variable depending on outside temp and set temp*snip*

Yeah, I understand the HVAC numbers vary, which is why I posted the "Climate Power" percentage as well, which I've seen climb as high as 90% using A/C last week in 103F heat. To note, my tests were done with the dash reading 95F.

Very cool/informative wizkid..
*snip* wizkid you doing PC, iphone, android or mac?) *snip*

Using the Raspberry Pi embedded Linux PC, actually. Runs off of 5VDC and pulls barely a couple watts. I then SSH into it from my laptop via WiFi. This way I can work on the software and such without having the fiddle with cables in the car all the time, even work inside with the car outside idling.

-wk

 

[tboult]

Cool use of a raspberrypi. did not know they had one with wifi.. that would make the a lot more useful. Presuming you are than talking usb OBDII.

I'm looking at doing it as an phoneapp over blutetooth.. get the display and such even if its a tad slower on the I/O. doing MA is not very good on the phone.. but the ELM327 does support some type of monitoring of particular device IDs which I'll have to learn to do what I want.

 

[wizkid057]

Cool use of a raspberrypi. did not know they had one with wifi.. that would make the a lot more useful. Presuming you are than talking usb OBDII.

I'm looking at doing it as an phoneapp over blutetooth.. get the display and such even if its a tad slower on the I/O. doing MA is not very good on the phone.. but the ELM327 does support some type of monitoring of particular device IDs which I'll have to learn to do what I want.

The Raspberry PI doesnt come with WiFi, but, you can easily plug a cheap (<$5) eBay Hong Kong Linux Compatible USB WiFi dongle into it.

For the ELM327, you can set it to monitor ONE CAN bus ID at a time instead of all of them in monitor all mode. This is done by sending "AT CRA xxx" where xxx is the ID of the item you want to monitor. Then do "AT MA" as usual and you'll only get that ID. I wish it would allow more than one at a time, but, aside from masks (which aren't all that useful in this case) I don't believe you can.

-wk

 

[tboult]

yeah I knew about the monitor one.. I wanted to figure if the masks would let you monitor many. Doe the rasbury keep up with the OBD in MA mode?

 

[MrEnergyCzar]

The heated seats use 15W, 30W and 45W (on high) supposedly.... big difference versus running the actual heater...

MrEnergyCzar

 

[Hi-Voltage]

The heated seats use 15W, 30W and 45W (on high) supposedly.... big difference versus running the actual heater...

MrEnergyCzar

In other words I will have to limit my wife 's driving in the cold weather or she is gonna kill my stats

BTW thanks for all those numbers and work

 

[briveland]

This was actually on my list of things to test next, but, I need to automate my software bit more for safety, or draft my girlfriend as button pressing monkey...

hmm... not sure I would would lt her see this post!

 

[wizkid057]

yeah I knew about the monitor one.. I wanted to figure if the masks would let you monitor many. Doe the rasbury keep up with the OBD in MA mode?

It seems to do fine with the MA, but the ELM327 device I have has some issues occasionally with data errors and such. I've peaked the RaspberryPi at about 8 megabit through a USB ethernet adapter, so, 500kbps should be fine for it.

hmm... not sure I would would lt her see this post!

Too late!

 

[VeridianJoule]

Does it make a difference which battery the specific accessory is powered by to get accurate calculations or is the system all working together? I'm guessing there is no traditional alternator that re-charges the 12 volt battery in our Volts is there?

 

[saghost]

Does it make a difference which battery the specific accessory is powered by to get accurate calculations or is the system all working together? I'm guessing there is no traditional alternator that re-charges the 12 volt battery in our Volts is there?

Items on the 12V bus are powered through the APM when the car is on - a DC-DC converter that can pull up to 2kW from the main battery. In theory, you lose a little precision from the presence of the 12V battery and the efficiency of the APM; in practice the numbers we're working with aren't that precise to begin with.

 

[RScott]

I think one of your posts said you were dropping data on the Bluetooth ELM327 devices. Are they BT 1.1 (1 Mbps) or 2.0 (3 Mbps)?

http://en.wikipedia.org/wiki/List_of_device_bandwidths#Wireless_personal_area_networks

From what I have heard, the Bluetooth adapters are too slow to keep up with the Volt's 500kbps. The ELM327 isn't fast enough to capture all the packets all the time, either, but at least gets most of the data (enough to be useful). I haven't tried a Bluetooth adapter to see what kind of rates it really gets.

Part of the issue is that the 500kbps is raw data, whereas it gets converted to ASCII, so you need more than a 500kbps connection to be able to read the data (unless the adapter sends it in its raw form).

 

[PatsVolt]

Climate Power, relates more to total BTU capacity rather then the raw electrical power and is an estimate of percentage of BTU's needed to keep the cabin at what ever temp is selected. A good examlple of how Climate Power numbers can change for a given setting is when in Very cold weather, when below 27 degrees the ICE will start, in this process as the engine heats and begins to cotribute to heating the cabin, you will see the Climate Power drop because the engine can produce a very large number of BTU's. thus the 100 % point gets readjusted upward and the since the cabin does not demand more Climate Power then the Climate Power number decreases. When the Ice turns off you will see the number increase. You may also see this in CS mode in the winter if the ICE turns off periodicly.

Pat

 

[wizkid057]

From what I have heard, the Bluetooth adapters are too slow to keep up with the Volt's 500kbps. The ELM327 isn't fast enough to capture all the packets all the time, either, but at least gets most of the data (enough to be useful). I haven't tried a Bluetooth adapter to see what kind of rates it really gets.

Part of the issue is that the 500kbps is raw data, whereas it gets converted to ASCII, so you need more than a 500kbps connection to be able to read the data (unless the adapter sends it in its raw form).

Exactly. The issue is that the ELM protocol sends the data in ASCII based hexidecimal, so, 500kbps of binary CAN data ends up being well over 1 megabit, probably closer to 1.5 megabit. If you really really REALLY wanted to make something like this work over Bluetooth PAN or similar, you'd have to compress the data on the fly on the sending side and decompress on the receiving side. Not really worth it IMO.

-wk

 

[Fulgerite]

This is a really an excellent and informative post. (Thank you wizkid.) I did some tests of the fan blower current consumption which I measured by removing the fan fuse and measuring across it with a digital amp meter. (VOM) I got about 20-21 amps with the fan at full speed... (Which I think is pretty close to what you calculated... Within a few watts.)

 

[wizkid057]

This is a really an excellent and informative post. (Thank you wizkid.) I did some tests of the fan blower current consumption which I measured by removing the fan fuse and measuring across it with a digital amp meter. (VOM) I got about 20-21 amps with the fan at full speed... (Which I think is pretty close to what you calculated... Within a few watts.)

Yeah, I had actually done some similar tests for some 12V accessories and it turns out the for anything over 25 watts or so measuring the APM's draw on the HV battery is pretty accurate. And, as it turns out, the APM is insanely efficient (> 90%).

Current draw/gain on the actual 12V battery appears to be near non-existent regardless of what accessories are powered while the vehicle is powered on, once the 12V battery is fully charged after a power up. So, the 12V battery should have little effect on any readings. This also seems to thwart my idea of adding a minimal amount of range by adding supplemental 12V power to offset accessory power usage...

-wk