GM Volt Forum banner

1 - 20 of 34 Posts

·
Registered
Joined
·
1,018 Posts
Discussion Starter · #1 · (Edited)
I was wondering whether (and why) our Volt is faster when the gasoline engine is running. In either case, the Volt is being propelled by electric motor MGB, but the source of the current is different. Would that affect the speed of the car?

I decided to check it out using my DashDaq data logger. I did a few sprints from 0 to almost 60 MPH in 'normal' (CD) mode, in 'hold' (CS) mode and in 'mountain' (also CS) mode and recorded data for MGB currents, voltages, speed, time, and torque as reported deep from the dungeons of the Volt's brain using DashDaq. The electric power into the drive motor MGB is calculated using the signals 'MtrB Volt' * 'MtrB Cnrt'. Here I plotted that power against the speed for CS and CD mode, when flooring the pedal:



There are a few interesting conclusions:

  • For speeds above 45 km/h (28 MPH) there will be more power available in CS mode. In CD mode, peak power is about 105kW, while in in CS mode the power hovers around 114kW. That's ~8% difference,or about 0.3s faster to go from 0-90km/h. So for any racing from a traffic light: put it in mountain mode or hold mode! Do not put in Sport mode!!
  • For speeds below 45 km/h there is no difference between any of the modes.
  • In Hold mode the ICE turned on between 1 and 2 seconds after the begin of the sprint. In Mountain mode the engine was running from the beginning. There is no significant difference between these two mode in terms of power output, so I only drew hold mode.
  • Reported torque at low speeds is the same for all modes: around 2550 N-m (1880 lb-ft) for speeds between 2 and 45 km/h. The Volt is rated at 368 N-m, so that is totally off. Likely a bug in DashDaq, but the relative value seems to be correct.
  • Torque softening seems to apply only for the first 0.1s of the sprint. After that it has already reached it maximum value, so removing torque softening is unlikely to affect 0-60 times by much.

The data suggest that the battery internal resistance is to blame for the difference. The battery voltage drops by 25V from 344V down to 319V during the sprint in in CS, while in CD mode its less (344V down to 327V). It also shows that some 8% (25V out of 319V) of the energy is wasted in the battery under maximum load.

In CS mode the generator MGA produces some 40 kW of juice which unloads the battery, so the latter only needs to deliver the other ~75kW. That can explain the better power. My charge while testing was 10%, so it might be that the CD numbers are better when the battery is full. I will try that.

At speeds below 45 km/h it doesn't matter because MGB cannot run at maximum rated power output yet until it turns 2600RPM.

It is likely that GM limits the battery power output to reduce stress and wear. Running the 55kW generator MGA gives significant extra power without overstressing the battery system. That is true if we assume that the battery is the limiting factor, and not MGB or the power electronics. The difference in power output and voltage drops seems to suggest that.

The Cadillac ELR's funny power rating of '117-135 kW' likely means that its 117kW in CD-mode and 135kW in CS-mode. And its 0-60 value is likely for CS modes as well. With a little software tuning that should be easily achievable from a stock Voltec.

There have been long threads on such tuning of a Volt, unfortunately without much proof whether the tune actually makes it any faster. The tuner BNRracing wrote me that its all a 'trade secret', so they don't tell what they did. It is my wild guess that BNRacing did a conventional tune of the Volt's ICE, as that is a stock motor. But a slightly faster running ICE would give insignificant extra power at MGB without some additional tuning of the current limiters. So it will be interesting whether any of such tunes materialize.
 

·
Moderator
Joined
·
8,680 Posts
Your conclusions mirror the results Motortrend got in their initial testing:

http://www.motortrend.com/roadtests/alternative/1010_2011_chevrolet_volt_test/viewall.html

Since you're hitting the correct (promised) power levels at the correct speeds, I'm going to assume the reported torque is confused. I've come across a couple fields through DashDAQ that I think may be in bad units - the math says that they are metric when the labels claim english or vice versa. In particular, the engine reported torque doesn't seem to jive - I tried building engine power output and it was far higher than it should be (unless MG A losses were >40%, which doesn't match the performance charts from the SAE.) I tried to make a road power number with reported axle torque and it didn't really seem to work right, either.

Still interesting insight. :)
 

·
Super Moderator
Joined
·
6,380 Posts
In Hold mode the ICE turned on between 1 and 2 seconds after the begin of the sprint. In Mountain mode the engine was running from the beginning.
FYI on what Mike said about keeping the ICE on.
http://gm-volt.com/forum/showthread...mance-tuning-for-the-Volt&p=900977#post900977
FastMike:
Car is quickest with a zero mile battery while running in Sport mode using the ICE.
"power brake" the car to keep the ICE running then go with trac off.
I think SPORT mode because that is how the tune is triggered since we know it is normally just pedal remapping for the first 1/4 or so travel.
 

·
Registered
Joined
·
1,018 Posts
Discussion Starter · #4 ·
Hi Scott,

I'm highly skeptical of fastmike's observations and even more so of BNRacing's tuning claims, as you might have guessed ;) With empty battery the easiest way to keep the ICE cranking at high RPM is using mountain mode (must have < 50% SOC). I don't understand what he means by 'power braking', as I've not been able to get the ICE on when hitting both brake and accelerator pedals.

My data proves that it doesn't matter, because I saw no difference between a running ICE or not (in CS-mode), which is understandable because ICE output is only needed after ~4 seconds when the speed is about 45 km/h. The ICE comes on after 1 second, so its ready in time to start working. At 45 km/h its already cranking seriously at 3500 RPM, that should be more than enough unload the battery. I wonder how often he gets about 45 km/h in a parcour race.

I expect that is better to drag-race on a full battery, since there is a much bigger battery voltage to start with. It might be that that reduces the difference between CS and CD above 45km/h. Fastmike needs to get the battery empty since he wants to run in 'sport' mode which is CS only on an empty battery. In a stock Volt sport-mode does not give any faster times or more output.

I do suspect that BNRacing's tune is just a conventional tuning of the ICE's tables and nothing more. By itself, that should have a totally insignificant effect on CS-performance and none in CD. I would be happy to be proven wrong on this.
 

·
Registered
Joined
·
1,018 Posts
Discussion Starter · #5 ·
I tried building engine power output and it was far higher than it should be (unless MG A losses were >40%, which doesn't match the performance charts from the SAE.) I tried to make a road power number with reported axle torque and it didn't really seem to work right, either.
I calculate engine power by multiplying the signals 'MtrB Volt' by 'MtrB Cnrt' for each MG. That seems to be correct indeed. When done in the DashDaq it lags by about a second, so I did that in the spreadsheet. There is another 'HV Volts' signal that is slightly higher than 'MtrB Volt', likely because its measured in the battery unit and not in the motor control. that all adds up, but torque doesn't I suspect that its not measured by derived from rpm and power, with a but in it.
 

·
Registered
Joined
·
1,650 Posts
[*]Reported torque at low speeds is the same for all modes: around 2550 N-m (1880 lb-ft) for speeds between 2 and 45 km/h. The Volt is rated at 368 N-m, so that is a little off. It could be a bug in DashDaq, but the relative value seems to be correct.
Just a wild guess, because I'm not an expert, is it possible the DashDaq is reporting the torque at the wheels, instead of at the motor shaft? The gear ratio is about 7:1, I think.
 

·
Moderator
Joined
·
8,680 Posts
I calculate engine power by multiplying the signals 'MtrB Volt' by 'MtrB Cnrt' for each MG. That seems to be correct indeed. When done in the DashDaq it lags by about a second, so I did that in the spreadsheet. There is another 'HV Volts' signal that is slightly higher than 'MtrB Volt', likely because its measured in the battery unit and not in the motor control. that all adds up, but torque doesn't I suspect that its not measured by derived from rpm and power, with a but in it.
Yup. I've been doing that for motor power since right after I got the unit. But I said engine power, and that's what I meant. The ICE reported torque number is what I was struggling with (torque*rpm/<a constant based on units> equals power.)
 

·
Super Moderator
Joined
·
6,380 Posts
I'm highly skeptical of fastmike's observations and even more so of BNRacing's tuning claims, as you might have guessed ;) With empty battery the easiest way to keep the ICE cranking at high RPM is using mountain mode (must have < 50% SOC). I don't understand what he means by 'power braking', as I've not been able to get the ICE on when hitting both brake and accelerator pedals.
No worries. Love your data gathering, analysis, and chart. Checks and balances are always good. Hopefull there will be some more test cases and comparisons. That is what I assumed he meant by power braking as well and what I did in highschool.

http://www.dsmtuners.com/forums/1194607-post2.html
powerbraking is when you use the brakes with an automatic car to build engine rpm before launching. Basically the car is in gear and you have a foot on the brake and the other on the gas. The brakes hold the car in place so you can launch at a higher rpm than idle. Mostly this is done in automatic RWD cars to do a burnout without the car moving much. I would highly suggest you not try this on a stock automatic tranny. If you have a built racing tranny with a high stall torque converter, shift kit, etc then go for it and tear the dragstrip up.
 

·
Registered
Joined
·
1,409 Posts
...
  • For speeds above 45 km/h (28 MPH) there will be more power available in CD mode. In CS mode, peak power is about 105kW, while in in CD mode the power hovers around 114kW. That's ~8% difference,or about 0.3s faster to go from 0-90km/h. So for any racing from a traffic light: put it in mountain mode or hold mode! Do not put in Sport mode!!
...
Either this statement is backwards or the blue/red lines are backwards. Cause above 45kph clearly the blue line shows higher power.

BNRacing modified the software to allow the system to add the the ICE and MGA battery draw together to produce 111kW for MGA and allow the 111kW for MGB. The 165kW draw on the battery made the power management system very unhappy which is why the PPR showed up at the end of each run.

I'm sure lots of us are waiting to see what Cadillac actually did.
 

·
Registered
Joined
·
1,018 Posts
Discussion Starter · #10 ·
Either this statement is backwards or the blue/red lines are backwards. Cause above 45kph clearly the blue line shows higher power.
Thanks for noticing. My bad accidentally swapped CS and CD in the text. Just fixed that.

I do not entirely understand your interpretation of BNRacing's mod. When in CS and at maximum acceleration, the total available electric power is ~105kW (from the battery) + ~50kW (from MGA that is cranked by the ICE) = about 155kW (330V, 470 Amps). Now the question is how much overdrive can the power electronics and MGB handle before they melt down into a smoking mess. There are likely multiple security mechanisms. None of this extra power from MGA is useful at low speeds (below 45 km/h). Looks like the ELR overdrives MGA until 135kW.
 

·
Registered
Joined
·
808 Posts
With units of HP and ft-lbs: Power = (Torque x rpm) / 5252

Power is what accelerates, torque is not fundamental to the process.

Batteries all have an internal resistance: the higher the current flow, the greater the voltage output drop. In CS mode, the generator is giving current to the traction motor free of this resistance. It's the same effect as when one is trying to start a car with a weak battery; if even a rather small battery charger is connected to the battery, the car will turn over faster, even though it's not appreciably charging the battery nor can it produce enough current to turn the car engine over by itself.
 

·
Registered
Joined
·
1,409 Posts
...
I do not entirely understand your interpretation of BNRacing's mod. When in CS and at maximum acceleration, the total available electric power is ~105kW (from the battery) + ~50kW (from MGA that is cranked by the ICE) = about 155kW (330V, 470 Amps). Now the question is how much overdrive can the power electronics and MGB handle before they melt down into a smoking mess. There are likely multiple security mechanisms. None of this extra power from MGA is useful at low speeds (below 45 km/h). Looks like the ELR overdrives MGA until 135kW.
BNRacing is not in CS mode, he's in CD mode. The difference is the ICE is also running at a red line of about 5500 RPMs. I'm not sure of the precise sequence of clutching and electrical power application but it all happens in less than a second. The ring gear is clutched to MGA and the ICE. So the ICE and MGA as a motor are spinning the ring gear. At the same time MGB is adding its max power to the sun gear. If you remove battery draw limiters MGA and B can both draw as much as they can from the beginning.

Does that make any better sense?
 

·
Moderator
Joined
·
8,680 Posts
BNRacing is not in CS mode, he's in CD mode. The difference is the ICE is also running at a red line of about 5500 RPMs. I'm not sure of the precise sequence of clutching and electrical power application but it all happens in less than a second. The ring gear is clutched to MGA and the ICE. So the ICE and MGA as a motor are spinning the ring gear. At the same time MGB is adding its max power to the sun gear. If you remove battery draw limiters MGA and B can both draw as much as they can from the beginning.

Does that make any better sense?
Not really. with the engine's ~100 ft-pounds and MG A's max of 200 Nm (~147 ft-pounds,) fighting against 2.3:1 adverse gearing, you have to reduce MG B's torque to ~108 foot pounds (~40%) or else it'll spin the engine backwards.

There's a reason GM designed the low speeds in series, people. With the Voltec architecture and the components we have, the car is slower with the engine linked in until you get to higher speeds, somewhere well beyond the critical speed of MG B.
 

·
Registered
Joined
·
1,415 Posts
For whatever its worth, I get two distinct surges while accelerating hard in CS mode at highway speeds. My guess is that simply applying max torque numbers is misleading. The electrics both have torque reductions as speed increases while the ICE is relatively flat. I will go with the first surge being B with help from A charging and the second being direct engagement of the ICE as it speeds up. I seem to recall the axle torque bouncing around on the DD, not smooth.
 

·
Registered
Joined
·
1,018 Posts
Discussion Starter · #15 ·
Not really. with the engine's ~100 ft-pounds and MG A's max of 200 Nm (~147 ft-pounds,) fighting against 2.3:1 adverse gearing, you have to reduce MG B's torque to ~108 foot pounds (~40%) or else it'll spin the engine backwards.
Yup, there is no other way to speed up a Voltec car than to crank more current into MGB.
 

·
Registered
Joined
·
1,269 Posts
There are a few interesting conclusions:

[*]In Hold mode the ICE turned on between 1 and 2 seconds after the begin of the sprint. In Mountain mode the engine was running from the beginning. There is no significant difference between these two mode in terms of power output, so I only drew hold mode.
Thank you for testing this and posting the result, I found it informative. I've always noticed slightly more power at higher speeds with the ICE running and also had wondered if it would make any difference to use Mountain to give the engine a head-start from idle. Good to know that there was no significant difference.
 

·
Registered
Joined
·
1,409 Posts
Not really. with the engine's ~100 ft-pounds and MG A's max of 200 Nm (~147 ft-pounds,) fighting against 2.3:1 adverse gearing, you have to reduce MG B's torque to ~108 foot pounds (~40%) or else it'll spin the engine backwards.

There's a reason GM designed the low speeds in series, people. With the Voltec architecture and the components we have, the car is slower with the engine linked in until you get to higher speeds, somewhere well beyond the critical speed of MG B.
Your assuming there is something stopping the car from moving or accelerating to force MGB to turn backwards other than of the inertia of the vehicle. It like saying if you disconnected the mass of the vehicle from the drive that one motor would drive the other backwards. All of the forces are moving in the same direction. It's the same as saying one person who can push an object at 1MPH will not be positively assisted by a second person who could only push the same object at 1/2MPH alone in the same direction. They are both overcoming the same inertia and the object would be pushed at more than 1MPH, not less, even though the second person only has half the ability to add force.
 

·
Moderator
Joined
·
8,680 Posts
Your assuming there is something stopping the car from moving or accelerating to force MGB to turn backwards other than of the inertia of the vehicle. It like saying if you disconnected the mass of the vehicle from the drive that one motor would drive the other backwards. All of the forces are moving in the same direction. It's the same as saying one person who can push an object at 1MPH will not be positively assisted by a second person who could only push the same object at 1/2MPH alone in the same direction. They are both overcoming the same inertia and the object would be pushed at more than 1MPH, not less, even though the second person only has half the ability to add force.
Am not. :) The inertia is plenty in this case, in and of itself. Like electricity, mechanical power follows the path of least resistance.

The way you're describing the situation would be correct if the engine and MG B were independently connected to the wheels. It's even valid in the Prius's Hybrid Synergy Drive, because on an HSD car, MG2 is connected to the wheels by a fixed gear ratio even though it is mechanically in the same path (it's downstream of the power split planetary set.) Your way of thinking does not describe the Volt.

Think of it like a rear wheel drive race car with an open differential on a slick road. If one wheel loses traction, it starts spinning faster, and all of the power goes to it and even though the other wheel has good traction, almost no power goes to it, because it can't carry more torque than the other wheel has - the gearing prevents it by transferring the power.

The same thing is happening in the Voltec unit - MG B doesn't have a solid foundation to push against with the engine linked in; it is pushing against the engine and MG A. If the torque it delivers exceeds their capacity, it just causes them to slow down, like the slipping wheel mentioned above - and like that wheel, any excess power doesn't contribute to moving the car.

If you want to see this in action, put the car in Hold (or drain the battery,) and drive at a nice stable ~45 mph for 30 seconds or so (to get the car into the described configuration, with the engine on and crosslinked; you should hear the engine running, then a moment when the sound fades, and then the engine will come back but sound slower/quieter.) Now progressively floor the car over a few seconds, and watch what happens.

You'll feel it accelerate, then a distinct hesitation, followed by a surge of power. The initial acceleration is the car in power split, using MG A to supplement the engine; the hesitation is the car shifting back into series because it knows it can't meet the requested acceleration while in power split because of the limits on MG B torque. That surge of power is all the acceleration the unmodified car has from MG B with the ring gear locked that it can't use without the firm foundation to push against.

It might be different if you had a high torque engine to work with (like the 2.0T,) or if MG A were substantially larger than MG B (2.3:1 gearing, remember?) In the current design, even with any typical chipping modifiers, the car is faster in series at any legal speed because you can't effectively utilize MG B's power in Power Split.
 

·
Registered
Joined
·
1,641 Posts
I have a 2011 without Hold Mode. Is this the reason my Volt feels more powerful at highway speeds when operated in Mountain Mode?
 

·
Registered
Joined
·
6,464 Posts
Do we know exactly how the electricity is routed differently when the Volt is in CS Mode? I could see an 8% difference in power coming from nothing more than the electrical losses of pulling power from the battery rather than feeding it directly from the generator.
 
1 - 20 of 34 Posts
Top