BimmerFest BMW Forum banner

335d Torque/BT data collection

38K views 86 replies 12 participants last post by  TDIwyse 
#1 ·
Seems like there was some interest is this type of thing from the other thread. So if there's interest, and you like data, please contribute.

I've just started playing with some additional fields in the Torque app (not all of the fields are able to get data) that I hadn't looked at previously. There's several others I'd like to view and record as well in following posts (like comparing how much fuel is used at idle when in D or N, or with the AC on or off, etc.)

But here's some of the data from this mornings commute. It was freezing fog conditions so no wind and the drive was slower than normal (~49-54 mph most of the way) and I hit 3 of the 5 stop lights. Better than normal average mpg for the trip (due to lower speeds and no wind). Found the Air/Fuel ratio and Volumetric Efficiency calculation interesting.

Need to read more about the HP/TQ readings as they only show up sporadically (not sure if it is triggered on G force, so at steady state it doesn't record data). Anyway, I zoomed in on a section where I accelerated from a stop light (at restrained levels due to the ice). These two numbers look like fun things to experiment with . . .
 

Attachments

See less See more
1
#3 ·
Interesting that VE never went above 100%. Does it do so when standing on it? I would think that it would for a turbo-charged engine.
You are correct. The data does have points where it went above 100%, but I clipped the Y axis at 100%. Looking at the data now I see that it hit a max of 146 at the place I'm showing the torque/hp data.
 

Attachments

#4 ·
It appears Torque over estimates the mpg's by a slight amount, at least on my car. Here's the last two days morning commutes showing engine temps and trip mpg's. Interesting to me to see how horrible fuel efficiency is on a cold engine and how directly related it is to engine temp.

Both these days had similar winds (very light) and road conditions with just the ambient temps being different.

With my current vehicle configuration and fuel filling place my vehicle trip mpg has averaged 2% optimistic from hand calculated fill ups for the last month. Torque seems to be ~5% optimistic from the vehicle trip mpg. So, the Torque data seems to skewing the mpg's about 7% optimistic.
 

Attachments

#5 · (Edited)
Here's the last two days morning commutes showing engine temps and trip mpg's. Interesting to me to see how horrible fuel efficiency is on a cold engine and how directly related it is to engine temp.
Very interesting indeed. The graphs show what I've been suspecting for a while. That MPG really is in the toilet at cold engine temperatures. Seems MUCH more so than on gasoline engines. How just driving for one mile with a cold started engine can take more than 1 MPG off the displayed tank MPG average, with 450 miles on it.

We have 2 cars, the 335d and a '00 540iT. We're driving the d almost all the time except when we need the cargo room or weather is real bad (the iT has real snow tires). My wife has a very short commute, about 5 miles. It may make more sense to have her take the 540 to work as it looks like it may actually get better MPG for these short trips.

Not that it matters my average MPG display runs about 4% pessimistic compared to manually calculated values.
 
#7 ·
I should mention the data is being recorded in 1 second intervals (so the x axis is the # of seconds). The graphs on the right are "zoomed" into the last part of the whole trip data on the left side. The whole trip is ~18 mins depending on traffic and how I hit the lights.
 
#8 ·
Caught another DPF regen event on the morning commute. It initiated as soon as engine coolant hit ~140F and lasted the whole trip. Think it was finished based on the EGR behavior at the very end of the trip, but I'll re-check on the way home to see if it has to "finish" anything else.

Environmental conditions were similar to the plots above for the coolant/mpg plots so it's obvious the negative impact the regen event has on the overall fuel economy during its operation. The cold temps seem to require more post injection than during summer months based on my observations of the instantaneous mpg gauges behavior. I'm including plots of the logged instant mpg's that show the radical dips when post injection occurs.

EGR and throttle behavior change dramatically during regen. Engine management seems to want full oxygen for the burning in the DPF.
 

Attachments

#11 ·
Think it was finished based on the EGR behavior at the very end of the trip, but I'll re-check on the way home to see if it has to "finish" anything else.
It was finished. Looking back it had been 2 weeks and ~350 miles since the last regen event. Based on memory and observations from last summer, it seems warmer weather months gives me longer miles between regen events.
 
#10 ·
I might have been too hasty in my drawing of the DPF dips in the instant mpg above. Some of those dips are due to elevantion changes and stop lights. These plots do a better job of showing the impacts (the instant mpg overlay is a zoomed in section of non-stop light driving, same "winter cover" for the front grill openings). I'm comparing this mornings data with last weeks for a very similar condition (same trip, temps within 1 degree, winds within ~4-5 mph). OK, now I need to focus on work . . .
 

Attachments

#12 ·
I usually shift into N when stopped. Wanted to look at differences between this technique and leaving it in D. Can't get MPG data, but I did find this plot of the "Engine Load" interesting. It's pretty obvious where I shift from N to D . . . I would expect this to mean more fuel is used to idle in D than N.
 

Attachments

#13 ·
TDIwyse, that is excatly what my Garmin Ecoroute shows me. Also shows higher engine load with A/C on.

Over past 2+ years my average MPG has gone up in the winter months and down in the summer months, even though traffic is theoretically lighter in summer with the kids out of school.

It's the A/C. I'm sure of that.
 
#16 · (Edited)
Cmon guys. It hit a low of about 62 Deg F this morning. I had to wear a long sleeve T-Shirt to do my morning stretch!:p I'm considering turning my pool heater on as well!

My wife had to turn the A/C on midday, still gets into the 80's during the day.

This has been a rather warm winter for us in South Florida. Two years ago it hit 27 Deg F and all the iguanas, fish and snakes froze to death!:eek:
 
#19 ·
Sorry for the late response but I've been on vacation. Put almost 1k miles on the D and it's just a pleasure to drive.

I'm using a cheap OBD blue tooth adapter I bought off eBay for ~$18. Some people seem to have issues with them, but others don't. I haven't had any issues.

On the trip I did some experimenting with mpg data, especially during DPF regens. During regens (last one for the mostly highway trip happened at 450 mls from the previous) the Torque app's "average mpg" recorded numbers ~30% higher than the cars recorded mpg's. The Torque algorithm doesn't seem to be able to measure the post injection mpg impacts correctly.

Also, for non regen conditions the Torque app's standard mpg algorithm is much more accurate than the newer "enhanced mpg" algorithm. At least on my car with the settings I'm using. The enhanced algorithm was 30-40% low. The standard algorithm was ~5% high. Some tweaking is apparently allowed by various methods within the app that I haven't tried yet.
 
#18 ·
My only recommendation would be to unplug the adapter when not in use. It's convenient to just leave it in place all the time, but it draws a small amount of power with the key off and due to the BT, it may also interfere with your X going nite nite. Other than that it is just sending data. I have only used the el cheapo on Amazon, so I don't have anything to compare it to, but it works well. I think they all come from HK/China anyway.
 
#20 ·
Been playing around the BT. It's very powerful compared to the Torque app. Much more data.

Here's what I had it set to log on today's drive:

Time
Coolant temperature
Exhaust Gas Recirculation duty cycle command
Actual boost pressure Actual rail pressure
Exhaust gas pressure before particle filter
Traveled distance with this oil
Average engine speed
Fuel temperature
Engine oil temperature communicated on Low CAN
Command time for main injection 1
Command time for pre-injection 1
Command time for pre-injection 2
Command time for post injection 1
Command time for post injection 2
Traveled distance since last regeneration (16bit)
Actual ash quantity
Soot mass in particle filter
Differential pressure of the particle filter
Absolute pressure in particle filter
Corrected pressure before particle filter
Fuel quantity consumed since last succesful regeneration
Temperature before oxydation catalyst
Temperature before particle filter
Ratio between actual and maximal torque output
Current engine torque output
Desired drive torque
Actualsliptorque


You can adjust the time step on when data is recorded, but it's a slide bar without units. I had it set in the middle which, for all these parameters, resulted in ~ 1.2 sec steps. This isn't fast enough to capture rapidly changing results like what's happening during rapid accelerations. But I included some plots of a section of the drive where I entered a freeway on ramp (coasting at first) and was able to fuel it hard for a bit (but there was a slow vehicle in front of me so I had to let up).

Interesting to see how the pre-injection pulses work. Under hard acceleration it seems it only uses 1 pre injection pulse (pulse # units are in usec). Also interesting to see the pressure across the DPF and the common rail fuel pressure as a function of engine load.

Maybe next weekend I can try to get faster steps while doing some hard runs.
 

Attachments

#21 ·
Finally caught the illusive DPF regen event today. 420 miles since the last one. Had my laptop and BT cable in place from the get go so I had the full pre and post event data . . . Here's the 5 fuel injection pulses from the startup until a bit after the regen was completed.
 

Attachments

#25 ·
Thanks.

Saw some comments in another thread regarding DPF back pressure and since I had data available . . .

Here is some measured data which includes the backpressure on my d from the BT tool during full fuel acceleration runs. The March 10th one was directly after a completed DPF regen, so is likely the lowest possible back pressure condition possible regarding the DPF. The March 4th one would've been as it was getting loaded up previous to the regen.

It's neat that using the recorded rpm and "torque" data you can make a dyno plot. This could be fun :) I'll need to increase the recording speed of the data next time (the "dyno plot" has a shift occuring in there and time is too coarse, but it shows some good potential).

Interestingly the pressure across the DPF during full fuel isn't as bad as I would've expected. 140 hPa is ~2 psi.

Oh, and here are the reported "units" for the data I've been taken.

Time sec
Coolant temperature C
Exhaust Gas Recirculation duty cycle command %
Actual boost pressure hPa
Actual rail pressure bar
Exhaust gas pressure before particle filter hPa
Traveled distance with this oil km
Average engine speed rpm
Fuel temperature C
Engine oil temperature communicated on Low CAN C
Command time for main injection 1 usec
Command time for pre-injection 1 usec
Command time for pre-injection 2 usec
Command time for post injection 1 usec
Command time for post injection 2 usec
Traveled distance since last regeneration (16bit) km
Actual ash quantity g
Soot mass in particle filter g
Differential pressure of the particle filter hPa
Absolute pressure in particle filter hPa
Corrected pressure before particle filter hPa
Fuel quantity consumed since last succesful regeneration l
Temperature befor oxydation catalyst C
Temperature before particle filter C
Ratio between actual and maximal torque output %
Current engine torque output Nm
Desired drive torque Nm
Actual slip torque Nm
 

Attachments

#27 ·
So I'm not happy with the update rate on the Torque app for measuring HP/Torque numbers. And the Bavarian Technic ECM reported torque #'s appear to be based on calculations of other engine parameters (and it's update rate is pretty slow as well).

So, I broke out my old GTech Pro. However I was unable to lock-on to an rpm signal when connected to the power outlets due to BMW doing such a good job filtering the noise. So I cludged together a way to power the GTech from the battery and got a crystal clear rpm signal.

Then I found a local car whose owner has a BT and a JBD that is modified for in-cab adjustability.

http://www.bimmerfest.com/forums/showthread.php?t=573553

Last night was near perfect conditions. 70F and no wind. Said vehicle had just completed a DPF regen and had a full tank of fuel. The D is supposed to weight ~3804 lbs according to insideline (http://www.insideline.com/bmw/3-series/2009/2009-bmw-335d-full-test.html) and with the driver and some items in the trunk it would be total ~4000 lbs.

The runs below are with the JBD at 0, 1 switch thrown (~65%), both switches thrown (~85%).

Procedure was over the same stretch of flat road (interstate), gently accelerating in 1st to 2nd then getting into 3rd gear before giving it the fuel, then letting up at ~80mph so I didn't run the risk of loosing my license if . . .

The numbers include not only loss of the tranny and tires, like a rolling dyno, but also the loss due to wind resistance. Looks like the Cd and A for the e90 is 0.26 and 2.17m^2. Using this (http://www.gtechprosupport.com/support/AeroDragCalc.php) says that at 80mph there's roughly 21 hp lost due to wind.

I'll post some of the BT data later of some runs that has some interesting data. Unfortunately it was trying to record too much data and the time steps between readings is larger than I'd like. But things like EGT's, DPF pressures, injection timing, etc are there.
 

Attachments

#30 ·
Forgot to mention the ECM calculated torque values. They seem to be based on something other than what is actually being developed by the vehicle in the modified state since they report nearly the same values in the 3 separate runs. I'm guessing it's using things like the rail pressure, injector opening times, boost, etc. to calculate those values. But that's a guess on my part.
 
#39 ·
The data seems to show similar delta's as these runs: http://www.burgertuning.com/images/JBD_dyno.jpg

And I see it wasn't stated specifically, but the plotted engine parameters correspond to the runs with the JBD at 0, ~65% and ~85%. This is why the parameters like the DPF differential pressure and EGT's for the three peaks are progressively increasing.
 
#40 ·
Got some more data.

Differences from last data: 67F with 6mph headwinds gusting to 12. The engine cover was off this time so the JBD could be installed after the stock runs. There was no DPF regen previous to runs, and less total time driving previous to runs. Less heat soak?...

Due to time constraints the runs were: 1) stock 2) JBD 0 3) JBD ~65% (1 switch)

The interface for the GTech to a PC is serial to serial and I don't have a serial port on my home PC's so I'll have to wait 'til Monday to use my work PC to get the GTech graphs. But here's the peak #'s.

Stock ~210 hp 353 lb-ft
JBD 0 ~ 222 hp 367 lb-ft
JBD 65% ~ 240 hp 387 lb-ft

The fact that I'm using 3rd gear pulls and the runs happen so fast is likely why the low rpm #'s are lower than others who are reporting higher tq #'s using 4th or 5th gear pulls on a dyno.

Here's some of the DPF and EGT data for the runs.
 

Attachments

#41 ·
Here's the GTech plots of stock vs JBD 0% vs JBD ~65%. The JBD runs were a bit higher hp than the previous ones for the same settings even with a bit of head wind. Think this is due to less heat soak from the DPF regen previous to the other runs.

Also found this plot of the "actual" vs "calculated" rail pressure interesting. The previous data didn't show the the ECM recording less "actual" pressure at JBD ~65% but seemed like it was at the 85% level. But the time steps were more coarse on those runs so it's hard to see for sure. Wonder if since the fuel temps were also lower on these runs if the HPFP was having a more difficult time flowing enough fuel? And this might also be the reason some people throw codes at the higher settings as the difference between these two readings might exceed an acceptable margin.

P.S. I forgot to set the date/time on the GTech on all these data sets (each time you plug it in it comes up at its "default" date/time which is Jan 17, 2017 . . .).
 

Attachments

#47 ·
Also found this plot of the "actual" vs "calculated" rail pressure interesting. The previous data didn't show the the ECM recording less "actual" pressure at JBD ~65% but seemed like it was at the 85% level. But the time steps were more coarse on those runs so it's hard to see for sure.
Well, was I wrong! I hadn't zoomed into the rail pressure data on the previous runs like I did on these recent runs. Here's the "actual" vs "calculated" rail pressure from the previous runs at 0, 65%, 85%. Even with the more coarse time intervals it's obvious that there's an increasing error going on. It seems the pressure relief valve is opening up and keeping the rail pressure in "safe" region even though the ECM thinks it's not making correct pressure.

A lot of the truck guys have in the past plugged the relief valve but this caused a lot of issues with cracked injector bodies. It's good to see this system has some built in protections.
 

Attachments

This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top