GM Volt Forum banner
61 - 80 of 102 Posts

·
Registered
Joined
·
4,728 Posts
That's very cool if they are doing it with GM's blessing.
I mean if they were able to combine the generic OBD2 PIDs with 125 Volt specific Volt specific PIDs that would make a very useful and cost effective device for Volt owners. ( Iwas under the impression it was just the very short list of 8 PIDs you listed in your post!)
Kudos fror your efforts!
WOT
 

·
Registered
Joined
·
2,632 Posts
I mean if they were able to combine the generic OBD2 PIDs with 125 Volt specific Volt specific PIDs that would make a very useful and cost effective device for Volt owners.
Well, and that's the sticky wicket I'm hitting up against them at the moment. The Generic OBD2 PID driver finds 44 fairly useful PIDs on the J1962 bus. I can even install the Ford Specific (Base) PIDs driver with the Generic PIDs and it finds a few more. But when I try to install the Volt driver it complains that some other driver is already using the J1962 bus and bails.

So I can either install the "It's an electric car - these PIDs are interesting to me" driver -XOR- the "It's a gasoline powered car -these PIDs are interesting to me" driver. But I can't install them both at the same time.

Which is just phenomenally stupid. That's the whole point!

Once I finish sorting it out with them I'll either send back the unit and develop a recommendation against it, or I'll post the combined PID list and heartily recommend the unit.

P.S. My wife would debate you the term "cost effective" :- )
 

·
Registered
Joined
·
2,632 Posts
I also have it displaying what I *think* is instantaneous MPkWh, but I need to dump the logs and the raw data in to verify my factors. For all I know it's really KPkWh!
Initial data verification does in fact lead me to believe I've got it calculating instantaneous MPkWh, which'll make steady state CD performance analysis a *lot* more repeatable. Until I can measure gas flow I won't be able to measure CS performance or any benefits of P&G. But at least it's a start. In nothing else I'll have an SOC meter during CS :- )

Presuming I keep the unit, I'm also going to switch from keeping track of speed to using the transmission output speed. It's more precise. During my driving around tonight while measuring I never hit freeway speeds, so the transmission ratio stayed nailed up at 3.984. As the second motor comes on-line I expect the transmission ratio will vary as well, and then I'll find out if the transmission output speed is measured before the planetary or after.

I've also added recording each motor's speed, torque, current, and volts (the latter two making watts). So that'll be interesting in terms of power splits and the effects on the transmission.

All fun stuff. Now if they'd just add back in things like MAF and fuel %...
 

·
Registered
Joined
·
4,728 Posts
It appears CAN is a pretty low voltage differential standard, so I actually could run it straight into the analyzer (I just bought another boat anchor: an Agilent 16700B with two 16717 cards for the connaisseurs) but I'm not going to ask for trouble by connecting an expensive piece of grounded gear to a multi-kW load and create the mother of all ground loops (and vaporize every GND wire between the LA power plug and the J1772 inlet).
I've never built anything around CAN but I guess there is an easy way to optically isolate it from the LA.
Makes sense. But I think I'll hold off on that one for a little while, to make sure I'm not issuing unknowingly "clear all" commands of sorts. First I'll see if the wiretapping goes anywhere.
Maybe you try that and tell me what happens :)
Thinking about it, I could cart over the LA to the Chevy dealer next thursday, they are going to do the software update! That would make for an interesting dump. :)
Thanks for that one, I got to skim those threads and perhaps borrow some ideas.
Yes, receiving calls from lawyers based in Munchen or Stuttgart is certain to ruin your day. :)
For those of you that are attempting to "backwards" engineer the GMLAN data streams (I generally applaud your ambition and efforts) here's a pic of the DLC #1 pin callouts (there's actually another non-OBD2 compliant DLC under the RH side of the IP as well)

Primary Data Link Connector- (DLC) X84 *Under drivers side of IP
Pin 1 is Low Speed GMLAN (SWCAN ~33.3kbps) - Body Electrical Services Only
Pin 2-3 Unused
Pin 4 chassis ground (for scan tool)
Pin 5 data reference ground
Pins 6(+) & 14 (-) High Speed GMLAN- Primary Powertrain Bus (DWCAN ~500 kbps)
Pins 7-11 Unused
Pins 12(+) & 13(-) High Speed GMLAN- Chassis Expansion Bus (DWCAN ~500 kbps)
Pin 15 Unused
Pin 16 B+ (for scan tool power)

Auxillary Data Connector - X84B * Under RH side of IP
Pins 3(+) and 11(-) High Speed GMLAN High Voltage Energy Management Bus (DWCAN ~500 kbps)
High Voltage Powertrain Pin 4 chassis ground (for scan tool)
Pin 5 data reference ground
Pin 8 ground for auxiliary data connector identification
Pins 12(+) & 13(-) High Speed GMLAN- High Voltage Powertrain Expansion Bus GMLAN (DWCAN ~500 kbps)
Pin 16 B+ (for scan tool power)

NOW that being said, it is very important that your efforts DO NOT result in any damage to the Data Link Connector terminals as this could make for expensive problems with dealership equipment.
PLEASE DONT PUSH PAPERCLIPS OR OTHER POINTED PROBES IN THESE CONNECTORS!

So in a perfect work your testing would only be performed with a J35616-2A test probe. (The dealer tool for testing on these circuits)
However you could also create test probes using the correct 150 series Metripack male terminals harvested from some aftermarket heated oxygen sensor harness repair kits.
If you visit a local AC Delco parts distributor and purchase PN# PT919 you will get a connector 4 pigtail wires that use the correct male terminals for about $20. You can release the wires from the housing using a large saety pin going in from the front-side. (PM me if you have trouble)
There are other HO2S connectors connectors that will also work but they might be more expensive. Most harness connectors use FEMALE terminals so that's why the O2 sensor kits are so useful.

Finally, there IS some General Motors OBD2 MODE 6 and CAN data published into the public domain that might assist you in your efforts located here:http://service.gm.com/gmspo/mode6/

Although there's nothing there yet specific to the VOLT, there are some "tidbits" that you might find quite useful ;)
HTH
WopOnTour
 

Attachments

·
Registered
Joined
·
847 Posts
OK, it's time for me to chime in; I didn't realize this thread had continued recently.

I bought a $30 OBD2 USB CAN adapter (the ElmScan 5), having no idea what (if anything) I would be able to get out of it.

All I can say is "Wow". The key is its "Monitor All" mode, which lets you see most of the high speed CAN traffic. It needs to be reverse engineered, but I've already got a fair amount of useful data from it (accelerator position, MPH, minute/second, and think I've just found the battery state of charge data). More information (such as GPS) is likely available via SWCAN (single wire), but few adapters support SWCAN.

I'll be posting a separate thread about my experience.
 

·
Registered
Joined
·
2,632 Posts
My first results that might be of interest from the DashDAQ OBD scanner, I eyeballed the transitions of each battery bar. The DashDAQ has at least two signals for battery level in the Volt extension driver, one for battery level % and one for the battery gauge %. The gauge turns out to map directly to the battery bars on the display, with each bar ticking off every 10% of the signal range (at least it did for the transactions I happened to watch - staring at the scanner while driving being a bad idea I haven't watched to confirm the entire series).

Here's how it maps between the gauge display and the actual battery SOC (or at least I think the VICM term is SOC).

Code:
Battery Gauge  Gauge %     Battery %
10 bars        91%-100%    81%-86.5%
9 bars         81%-90%     74.4%-81%
8 bars         71%-80%     68%-74.4%
7 bars         61%-70%     61.5%-68%
6 bars         51%-60%     55.3%-61.5%
5 bars         41%-50%     48.7%-55.2%
4 bars         31%-40%     42.1%-48.7%
3 bars         21%-30%     35.6%-42.1%
2 bars         11%-20%     29.3%-35.5%
1 bar           1%-10%     22.7%-29.3%
0 bars          0%-1%      20%-22.7%
The ICE comes on around 20% SOC, and the CS SOC is around 22%. So the ICE runs a bit to bring the SOC back up from 20%ish to 22%ish. This somewhat confirms some of what we've heard about the battery and battery SOC.

I've not yet checked out MM SOC.

The unit can report kW for the battery and each motor, RPMs for each motor and the ICE, and the amount of torque each is generating (with MG-A generating negative torque in CS mode). There's WAY too much data there for me to boil down in a hurry, but suffice it to say the Volt appears to have some tricks up its sleeves as regards how it uses the different units.
 

·
Registered
Joined
·
4,728 Posts
P.S. My wife would debate you the term "cost effective" :- )
Still it's quite inexpensive as compared to an MDI and GDS2 subscription. LOL
Looks like you're having fun. Keep up the good work! But be careful.
I've been road-testing with scan tools and data collection equipment for nearly 30 years and have had my share of "close calls" and even a couple closer! lol

WOT

PS> PIDs displaying negative an positive tq values (or RPM) when it comes to electric motors* refers to the direction of rotation from the perspective of output rotation. (*from GM anyways)
 

·
Registered
Joined
·
2,632 Posts
PS> PIDs displaying negative an positive tq values (or RPM) when it comes to electric motors* refers to the direction of rotation from the perspective of output rotation. (*from GM anyways)
Yeah, I figured that part out. I'd guess wrong on amps though. Since batteries power things, I'd figured positive amps meant power was flowing from the battery, and negative meant into. Nope... The DD allows sign change operations easily enough, so I've flipped signs on the battery and motor amp readings to fit my preference. And while the signals for that from each is in volts and amps I'm converting to kW for my viewing pleasure.

The *vast* majority of the DD's time is spent un-watched, merrily recording data to a memory card. I do confess to spot checking it, though I reserve that for times where there's little to nothing around me (including curves). Right now all the data's being displayed numerically (since I'm watching some 24 different signals or composite signals at the moment). Eventually that'll blend in to several pages of dials and bar graphs to allow better "status at a glance".
 

·
Registered
Joined
·
847 Posts
Code:
Battery Gauge  Gauge %     Battery %
10 bars        91%-100%    81%-86.5%
9 bars         81%-90%     74.4%-81%
8 bars         71%-80%     68%-74.4%
7 bars         61%-70%     61.5%-68%
6 bars         51%-60%     55.3%-61.5%
5 bars         41%-50%     48.7%-55.2%
4 bars         31%-40%     42.1%-48.7%
3 bars         21%-30%     35.6%-42.1%
2 bars         11%-20%     29.3%-35.5%
1 bar           1%-10%     22.7%-29.3%
0 bars          0%-1%      20%-22.7%
This pretty much matches what I saw. I found a number that I thought was battery SOC (in .25wH increments), that I tested over the weekend, and saw numbers very close to what you are seeing for Battery %. The interesting thing is that the '10 bars' (full charge) seems to vary a fair bit (with 4 samples of a full charge, I saw 87.6%, 88.8%, 88.9%, and 89.2%).


The ICE comes on around 20% SOC, and the CS SOC is around 22%. So the ICE runs a bit to bring the SOC back up from 20%ish to 22%ish. This somewhat confirms some of what we've heard about the battery and battery SOC.
I have been seeing something similar. On one trip, when the gas engine first turned on, the SOC hit a low of about 21.7%, then for the next 10 minutes stayed between about 22.2% and 23.3% (or 21.7% to 22.7%, as I'm working with undocumented numbers). Looking at the hour after the gas engine turned on, it stays within that narrow 1% or so band except for the occasional quick dip below (which may be due to the time it takes for the engine turn back on and starting to get the SOC up), and a couple of times where it goes above that 1% for a while (likely due to regen).

I've not yet checked out MM SOC.
I've just glanced at the numbers I got this weekend, it got up to about 44.1%, backed down to 43.8%, then went up to 43.91% and stayed there. It looks like it adds back right around 1/3rd of the usable SOC.

The unit can report kW for the battery and each motor, RPMs for each motor and the ICE, and the amount of torque each is generating (with MG-A generating negative torque in CS mode). There's WAY too much data there for me to boil down in a hurry, but suffice it to say the Volt appears to have some tricks up its sleeves as regards how it uses the different units.
I am envious! I've got more data to work with, but a lot is going to be impossible to decipher without any documentation, and a lot may be unclear. It looks like I found the latitude/longitude information, though, which will be very useful to me.

Please keep sharing anything interesting you find.
 

·
Super Moderator
Joined
·
6,434 Posts
Very cool guys!! Coolest stuff for months. Nice job. 65% number still looks generally accurate as 20%-85%

Had to use CS mode last night when some road construction screwed up my plans!!
Electronics Technology Multimedia Electronic device Screen


Some CS testing would be novel. Volt in in yellow here from a presentation I saw recently.
Text Electronics Line Diagram Font


http://gm-volt.com/forum/showthread.php?7050-CS-mode-2200-RPM-1500-RPM-then-quot-CD-mode-quot
The experience was very much different than my first post. The RPM was pretty steady on the flat drive. We did not see it rev up and the go into "CD mode". We drove about 5 minutes at each speed.

60 mph: 1600-1800 rpm
65 mph: 1700-1900 rpm
70 mph: 1800-2000 rpm
75 mph: 1900-2100 rpm
As well SOC/RPM/etc in the four modes: CD-Low|High ; CS-Low|High.
http://gm-volt.com/forum/showthread.php?6611-Voltec-drive-unit-now-has-four-basic-modes-of-operation
(and http://www.insideline.com/chevrolet/volt/2011/how-the-2011-chevrolet-volt-works.html)
 

·
Registered
Joined
·
2,632 Posts
Some CS testing would be novel.

As well SOC/RPM/etc in the four modes: CD-Low|High ; CS-Low|High.
One of the stats reportable by the DD is pedal position. So while it'll record it, I'm not sure now well I'd be able to maintain a static 80% position.

Now 100%, that's pretty easy...

For a CD acceleration test I suspect battery and motor voltage would be more interesting than SOC. I can't easily get to a flat acceleration test location with a full SOC so I can't easily test that. But how much the voltage drops under load might be entertaining.
 

·
Registered
Joined
·
847 Posts
I've just glanced at the numbers I got this weekend, it got up to about 44.1%, backed down to 43.8%, then went up to 43.91% and stayed there. It looks like it adds back right around 1/3rd of the usable SOC.
I should add that while parked, it seems that the gas engine generates about 11kW of energy, at around 1700RPM (if I've got the right number). If there were a way to charge the battery completely after depleted, it would take just about an hour. I calculated that 1 gallon of gas would produce about 7.56kWh of energy to the batteries.
 

·
Super Moderator
Joined
·
6,434 Posts
I should add that while parked, it seems that the gas engine generates about 11kW of energy, at around 1700RPM (if I've got the right number). If there were a way to charge the battery completely after depleted, it would take just about an hour. I calculated that 1 gallon of gas would produce about 7.56kWh of energy to the batteries.
No doubt some of that juice will be used to do the TMS (ie. if even if you are in the shade of a tree and the outside temp is 72 degrees, then charging will heat up the battery so it will need to cool down when it gets to 80 degrees or something ... using juice)
  1. Does that mean parked with the hood open to force the ICE/GG on?
  2. Can you just progressively see the SOC go up?
  3. Does the ICE/GG stay on with the hood up until it runs out of gas?

I guess this would be like running a "honda" generator and plugging your car in but the Volt's ICE/GG would be more efficient and "direct"? Hey do they make "honda" generators that run on diesel... you could use biodiesel. <grin>. I recall a thread where some guy in an apartment was going to buy the car and charge it using a "honda" generator as a regular solution.
 

·
Registered
Joined
·
2,632 Posts
I found a number that I thought was battery SOC (in .25wH increments), that I tested over the weekend, and saw numbers very close to what you are seeing for Battery %. The interesting thing is that the '10 bars' (full charge) seems to vary a fair bit (with 4 samples of a full charge, I saw 87.6%, 88.8%, 88.9%, and 89.2%).
There are actually a bunch of signals that sound like their represent battery capacity in one form or another. I've not looked at all of them yet. Yes, I agree I too am seeing some fair variability on what it thinks constitutes a full charge. And yes, the monotonic step size on nearly every signal I've seen is far greater than the reporting resolution.


I am envious! I've got more data to work with, but a lot is going to be impossible to decipher without any documentation, and a lot may be unclear. It looks like I found the latitude/longitude information, though, which will be very useful to me.
LAT/LON would be *way* cool to add to my reports! The DD supports a NMEA GPS serial input, and I've been debating either bringing my old Garmin GPS III+ out of mothballs, or seeing what happens if I plug a GPS into one of the USB ports. If nothing else, the Linux kernal source is available, so I'm almost curious enough to add a generic NMEA serial driver for some of the various test phones I have access to, and see what it does with that.

The DD also supports two analog inputs (on the same connector as the two serial ports), so I'm debating hooking up a "tick switch" to allow operator click input on "interesting locations" in the log reports. The amount of data is just so vast that finding the interesting bits in the logs requires either good post-mortem programatic detection or dumb luck.

Any chance you're near the southwest corner of the country? Would an OBD Y-cable (possibly with a diode bridge to isolate directions) help to see what the DashDAQ is injecting onto the bus?
 

·
Registered
Joined
·
847 Posts
  1. Does that mean parked with the hood open to force the ICE/GG on?
  2. Can you just progressively see the SOC go up?
  3. Does the ICE/GG stay on with the hood up until it runs out of gas?
This was with the car parked (battery already depleted), turned on, and switched to "Mountain Mode". The MM setting automatically caused the engine to run, even though the car was in park. This is something I wouldn't normally do, except for testing!

The SOC does continuously go up while this is happening. MM will then automatically turn the engine off when a certain charge is reached (in this case, right around 44% of the full 16kWh).
 

·
Registered
Joined
·
847 Posts
LAT/LON would be *way* cool to add to my reports! The DD supports a NMEA GPS serial input, and I've been debating either bringing my old Garmin GPS III+ out of mothballs, or seeing what happens if I plug a GPS into one of the USB ports. If nothing else, the Linux kernal source is available, so I'm almost curious enough to add a generic NMEA serial driver for some of the various test phones I have access to, and see what it does with that.
One of the things I want to do is figure out if there is a more efficient way to get where you want to go. Around here, there are a lot of small hills, and I figure that depending on the terrain and speed limits and such a slightly longer route may use less energy. The GPS data will make that possible. It would also be interesting to see on Google maps or something like that the average kW you're using on various roads.

The CAN ID I get for the latitude/longitude is 32A, if that helps. I get my data from a 'monitor all' function that looks at all data that goes on the CAN bus (with the drawback that I cannot make any requests, just being an observer). If you have the ability to simply monitor the traffic (rather than make requests), you can look for the 32A ID (it's reported about once a second on average).

The DD also supports two analog inputs (on the same connector as the two serial ports), so I'm debating hooking up a "tick switch" to allow operator click input on "interesting locations" in the log reports. The amount of data is just so vast that finding the interesting bits in the logs requires either good post-mortem programatic detection or dumb luck.
I had the same problem, and did something similar. In my case, I've got a laptop connected to the OBD2 port, and I have my program set so that it will add a line to the log file if a key is pressed. That way, for example, I could press it when the battery bars would go down, and didn't have to worry about looking for the data and writing it down.

Any chance you're near the southwest corner of the country? Would an OBD Y-cable (possibly with a diode bridge to isolate directions) help to see what the DashDAQ is injecting onto the bus?
Unfortunately, I'm on the other side of the country. I would love to see what DD does, but I'm guessing they have a license that would prevent that (if they have to pay GM, GM is surely going to require it). If not, I'd be very tempted to buy one, as it looks like there is a lot of data that they have access to that I do not with the 'monitor all' mode.
 

·
Registered
Joined
·
2,707 Posts
Discussion Starter · #77 ·
Excellent!!
 

·
Registered
Joined
·
2,632 Posts
It looks like I found the latitude/longitude information, though, which will be very useful to me.
Which CAN bus are you monitoring? Or are you able to monitor both? Are you looking at the standard CAN bus on pins 6/14, or the GM additional CAN bus on pins 12/13?
 

·
Registered
Joined
·
847 Posts
Which CAN bus are you monitoring? Or are you able to monitor both? Are you looking at the standard CAN bus on pins 6/14, or the GM additional CAN bus on pins 12/13?
I believe it's 6/14. It's the 500kbps CAN 11-bit ID (ISO 15765-4), connecting on the driver's side. I've tried connecting to a 125kbps CAN bus (using the same connection), but wasn't able to -- I'm guessing that may be the one on pins 12/13 you mention (which I probably wouldn't be able to access with the adapter I have). I thought there was also SWCAN, which uses pin 1 (which I also don't have access to). And I still haven't figured out what is available on the driver's side OBD2 connector (probably, again, something my adapter won't access).
 

·
Registered
Joined
·
2,632 Posts
Auto-tranny torque creep cost

The torque creep the Volt has is there so the car more mimics a typical automatic transmission car. I'd heard it doesn't hurt anything or really use up any extra power if you weren't moving, so it was a freebie...

Not so, says the DashDAQ!

I used the DashDAQ measuring axle torque, brake torque requested, and the battery kWatt draw. I believe axle torque is how much the car is trying to move forward, brake torque is basically how hard your pressing the brake peddle (at least at these speeds), and battery draw is battery draw (leave me alone, it's late :- ).

As brake torque requested approaches axle torque the drain on the battery increases by roughly 150-300 watts...

So if you're pressing the brake *just* enough to keep the car from creeping? Yup, that may be a third of a kW down the drain as you sit. What do I normally do when I'm stopped? Press the brake just enough to keep the car from rolling...

Two ways to avoid this energy drain. Press the brake harder, or put the car in neutral. Both drop transmission creep energy drain back down to zero. How hard do you have to press the brake? When they're equal, the torques are both about 200 units (haven't decided if it's nm or ft-lb yet). Raising the brake torque to 400 units kills the drain.
 
61 - 80 of 102 Posts
Top