[subdiesel]

...

• Under normal conditions, the fuel consumption readout will read 2l/100km (lowers it will show) coasting down to <20km/h
• Under regen, I can only make it say that at approx 100km/h and over. Anything less shows a progressive increase in readout l/100km even coasting under zero throttle, whether engine braking or not (i.e. clutch in at idle revs).
• It doesn't make sense that the EGR could keep the DPF/exhaust temp up alone, especially 600+C. It'll never do 100% EGR, even if it did there would be some decay or increase in temp. Under normal running (no regen) I have had, on a steep long mountain pass descent, the exhaust temp decay to ambient (20C) and stay there. But under regen conditions, because the exhasut gas will slowly be replaced by fresh air it must continuouosly be remade, even whilst coasting- by introduction of fuel.

I suppose the EURO4 open system may not have this issue especially if it's 'openness' means it doesn't worry if the temp drops for a while.

Hi!
IIRC Euro 4 consumption display always showed its minimum during coasting (0.2 L/100km ?), regen active or not. Combination meter computer is responsible for driving that little centre display. This displayed minimum value is NOT accurate, as the ECU actually outputs zero fuel flow CAN value [mm³/s]. Somehow they don't want that display to show zero. On my car, coasting always meant zero fuel flow, except when engine speed approached idle speed where ECU gradually resumed to inject fuel in order to retain its minimum target revs.
In addition, I had been logging very internal injection variables, each individual injection part, state machine etc. Results and correlations make perfectly sense to me. Also plots show EGTs dropping, EGR opening etc. during coast. Not sure about descents longer than half a minute or so, could not test this easily. Did not sniff injector drive wiring via oscilloscope – that would be the ultimate proof I guess.
You actually reminded me I had made some detailed graphs, will fine-tune them and publish within a few days. Of course, this is all about EU4 car with closed-type DPF, can't say if everything applies to EU5 as well.
Btw., open-type DPF models are Euro 4 Legacy/Outback only, = MY 2008/2009. Everything else, = MY2009+, has closed-type DPF which comes with all those disadvantages – additional sensors, post-injections, reduced responsiveness during regen and so on. AFAIK Euro 4 Boxer Diesel models were not sold in Australia.

 

[pitrack_1]

Hi Subdiesel,

If you want any particular parameters tested let me know. Might set up your injector params to see but I am disinclined to go on test drives in a car I don't drive anymore for extended testing on the half-chance the vehicle might comply with a regen though...

Ours shows a min of 2l/100km and yes I know that's not accurate. However under regen this increases, roughly in inverse proprotion to speed and/or engine revs. I know the EGT/DPF temp holds up under coasting due to fuel injection, it's not going to maintain 400-600C or so no matter how much EGR...

I'm sure there's proof buried in my copious graphs and data logging at least of temps vs speed and time..

Yes Oz started with the EURO5 diesel.

Might try a few other of your parameters in the meantime plus another readout of 221155 , will post results.

Cheers,

 

[pitrack_1]

PID 221155 Test cont'd 3

Of course I've hit the perfectly divide by 15 again, current output is "69" with a calculated oil dilution of 3.0%. Will try gain one more time later.

 

[pitrack_1]

Regens: EGR and injection whilst coasting

Well, well, well!

A bit of a test tonight and some interesting results. Took the car out to pick up dinner with it sitting on the edge of a regen. Plugged some of Subdiesel's/Subaru Diesel Crew's parameters in and logged some data coming back during a regen. I'll download/process the data and post some pics/images/results later, but the gist of it is:

1) EGR and 2 injection parameters (final injection period and final injection amount) work.

2) Injection DOES take place coasting under a closed throttle during regen whilst it does not without a regen- shown by injection amount, and if things went to plan I've both logged and screen-shot it.

3) I pulled over to adjust some of the parameters/displays and I think I had the regen end early...don't know why. I don't beleive it advanced the counter and I don't think I stopped the engine...I'll need to check this. Certainly when it had stopped I was still at 50% soot. Evil vehicle, I was working on my EGR dial setup which wasn't (isn't) quite right...

As the soot level is still relatively high I might take it for another drive and log data to see what happened. If so, I may not post until tomorrow/later, I do need to go to work... In the meantime, I need to save my setups and program (or load) some more parameters.

More to come.

 

[dpm]

it'll end after 12mins regardless of soot level.

 

[pitrack_1]

it'll end after 12mins regardless of soot level.

Thanks, I know dpm, but this one is different...


Hmmm, about 1 hr of driving tonight and the car steadfastly refused to budge from the 50-60% soot level during the whole time. This car *is* evil! :evilatyou:

Had a quick look at the data:

The previous regen did end early after ~7:15. The DPF temps had decayed to ~220C whilst parked, I suspect this 'detriggered' the regen. WHy I don't know, I can't recall it happening before whilst idling, done plenty of that at lights, even had the regen trigger whilst idling at lights. Confirmed by non-iteration of the DPF regen (completed) count.

It's rather late- actually rather early, ~2:30am, so I'll do the data and screen shots later.

Yawn!

 

[pitrack_1]

Screen shots and logging of injection params

Hi all,

I need to apologise, can't deliver on my promises as unfortunately something doesn't gel from my work last night. The data saved in the CSV file doesn't not seem to match what I was seeing on the Torque screen for what should be the same parameters.

However one upshot is that you may be right subdiesel, that (at least the main ) injection goes to zero under coasting under DPF regen. Certainly does under non-regen- both on-screen and in log. However I'm sure I'm seeing discrepancies under regen between the displayed values and the logged ones...or I'm hallucinating .

Screen shots were difficult when the car seems to want to pull slightly left, I'm coasting uphill from 60km/h, there's oncoming traffic (even at 1:30am now in Canberra) and you have to press two buttons on the tablet simultaneously. They haven't worked as well as I hoped.

I wanted to do a more extended downhill coast from 80km/h (or above) but the cancellation got the better of me and in ~90KM (>1hr) of driving the system refused to regenerate, even starting from 50+%.

Furthermore I had an EGR reading of "1" the whole first log time even though I though I'd implemented a correct formula, even if I hadn't it should have changes somehow...and the second, with no formual (just reporting the value) gave between 50 and 80. Subdiesel says the formula should be A-50 (probably with a multiplier, e.g. 10/3 to give a 0-100% range) makes sense when lowest reported value is 50, but that's what seem to generate the "1" before...

It could be I am reporting or displaying the wrong values. It could be hysteresis in the screen values. I'll need to do some more testing and see what I'm logging vs screen display are consistent and correct.

I'll need to try again but am unsure when I'll get a chance.

For what it's worth, here's a image of my latest dashboard setup, bit of a mess admittedly.

Cheers,

 

[pitrack_1]

EGR's response and data issues

I've been getting some rather strange readbacks for my EGR values run-to-run.

I ran the car for ~90 mins, the max and min were (respectively) 80 and 50. No problem, let's use that as the basis for a 0-100% scale: subtract 50 and multiply the result by (100/(80-50)) = 10/3, so:

(A-50)*10/3 (does this make sense?)

Next run, what do I get? A gauge min/max scale from 0 (good) to 140%... repeatably.

Run after... 0-120%.

Those max values seem repeatable, intra-run.

Now I don't know if there's a:
- 'gauge' response/scaling issue (and the saved data are different) giving it an overshoot on a sharp change,
- an issue in the data (need to check logged values)
- an issue with differing reporting from the ECU
- the EGR value is a commanded value, measured (readback) value, measured (independent) value or commanded 'drive strength' value (with 140% being to drive the EGR open quickly)

However I can confirm that under normal (non-regen) conditions the EGR's response is:
- rather binary under rapid/heavy/sharp throttle,
- variable under steady/feathered throttle or cruise control.

Need to see under regen again if I can ever organise it.

And I missed any sort of coasting under regen last time as I was climbing a mountain whilst it decided to regen...grrr!

 

[subdiesel]

Thanks for testing! Feel free to try and verify any other parameter IDs, no risks attached as mode 22 is read-only.
According to SSM-III docs and my own observations, Number of Times Injected and Final Injection Amount do not take all injections into account. However, if FIA > 0 then ECU most likely also continues post-injections during regen.
Entirely possible that regeneration and other behaviour has been changed over boxer diesel generations, that's why I hesitated to publish tons of my Euro 4 log data – the majority of diesel models out there are Euro 5 spec nowadays.
Unfortunately, highly advanced data, i.e. tracking all individual injections, cannot be gathered through diagnostic protocol items. The reason is simple - if such PIDs are not implemented in stock ROM then these are not available. A stock ROM may support ~200 pre-defined PIDs, however there are many thousands of internal variables. That's why I had to reverse-engineer firmware logic, patch my ROM to allow fast and unrestricted CAN-logging - necessary on Euro 4 at least, define and log ROM specific RAM addresses (RomRaider jargon: extended parameters). Then you also get full internal ECU accuracy, 32 bit floating point values for the most part. Disadvantage: costly, not exactly a one-weekend endeavour.

 

[pitrack_1]

Hi Subdiesel,

thanks for the comments and your work. That type of engineering work is well beyond me, however I fully understand the idea of the underlying varaible and precision complexity. I didn't realise the speed volume and complexity of communications in a CANBus car though!

I'll keep testing a parameter or two as I go in my slow ad-hoc way. I might put up a little table of the additional parameters with my testing result for EURO5-Oz-spec so i can work through your list.

If there's anything specific parameter you wish me to test (at least test read-out) or a particular way you wish me to test a parameter, let me know and I'll give it a go.

I guessed the final injection and OBD-II PIDs may not incorporate all pre- and post- components but maybe just the base 'power' component. That's OK, I'll see what happens anyway.

Given the Torque gauges behave a bit like a real gauge with a damped response time I may have been a little too quick on the screen snapshot. Besides, I guess the injection may behave like a petrol injection or carby dash-pot (remember those?) and not snap off the injection immediately, for emission (or other) control reasons.

I'll have a look at some actual data tomorrow.

 

[pitrack_1]

Couple more parameter tests

Hmm, well sorry I haven't looked at the data yet...even a couple of weeks now.

However just driven up to Syd from Canberra, no regens... now at approx 600km since last regen. How do I know this? I tested a couple more parameters from Subdiesel on the way up: both work!

Pressure Difference between DPF Inlet and Outlet (DPF Press diff), A, [kPa], 22 11 4A
Result shows a positive integer, got no feel for what it should be so I'll trust subdiesel that it's a reasonable figure.

Running Distance since last DPF Regeneration (dist since last regen), A*256+B, [km], 22 11 56
Seems to show the correct figure.

Those who have Torque Pro should note there's been a few updates recently: There is the possibility of headers and diagnostic start/stop commands for each extra parameter. These are beyond me but I expect someone 'out there' will know what to do with them?

Also I think there are some extra 'standard' parameters that have been added to the standard list: such as distance since last code reset (or maybe I didn't notice this previously) :biggrin:

 

[pitrack_1]

Injection time vs amount...

here's an interesting screen shot coasting under regen. It shows finite timed final main injection and null final injection amount (ottom-right corner of following screen image).

 

[pitrack_1]

Should've taken a screen shot...

Got to scan another 2010 Forester Diesel today. Has 230,000 km, country car but apparently only driven into/out of town ~15min now. No. of completed regens is ~370 (ours ~406).

But most interesting thing was the distance since last completed regen... 3303km..! :icon_eek:

Something funny going on with that car...maybe it's blessed with no DPF?!? :biggrin:

Only thing the owner doesn't like about the car is the dicky gear shift (hmmm, heard that somewhere before)

 

[kutavyz]

Hi!

In 2 weeks I have to do the first oil change in my MY16 diesel Forester and I'm going to put a 0W30 as the owner's manual says, probably Motul Eco-Clean 0W30 ACEA C2/API SN.
I've been researching about a lot of oil brands and I can't find a "Low SAPS" oil with ACEA C2 standard. The motul has 0,78% of sulfated ashes and is "Mid SAPS".
I know that vehicles equipped with DPF must be "Low SAPS", and I don't know if 0,78% is too high for my DPF.
I talked to Motul customer service here in Spain and they said that Subaru recommends SAPS approximately between 0.35% and 0.8%.
So... which oil do you use?

Motul specs: https://www.motul.com/system/produc...ions/technical_data_sheets/47/original/8100_Eco-clean_0W-30_(GB).pdf?1451925824

Thanks!

Enviado desde mi Nexus 6 mediante Tapatalk

 

[pitrack_1]

Hi!

In 2 weeks I have to do the first oil change in my MY16 diesel Forester and I'm going to put a 0W30 as the owner's manual says, probably Motul Eco-Clean 0W30 ACEA C2/API SN.
I've been researching about a lot of oil brands and I can't find a "Low SAPS" oil with ACEA C2 standard. The motul has 0,78% of sulfated ashes and is "Mid SAPS".
I know that vehicles equipped with DPF must be "Low SAPS", and I don't know if 0,78% is too high for my DPF.
I talked to Motul customer service here in Spain and they said that Subaru recommends SAPS approximately between 0.35% and 0.8%.
So... which oil do you use?

Motul specs: https://www.motul.com/system/produc...ions/technical_data_sheets/47/original/8100_Eco-clean_0W-30_(GB).pdf?1451925824

Thanks!

Enviado desde mi Nexus 6 mediante Tapatalk

Hi Kutavyz,

I must ay my vehicle has been dealer serviced. However they source their oil from normal manufacturers and I'm not even sure the dealer needs to go through Subaru for that.

Go by the spec (C2) rather than the description ("low SAPS"). If it meets your car's specs then it should be good enough. Just check there aren't other specs to meet (forgotten which) too.

C2 is a good common spec. Plenty in Australia even from local manufacturers/blenders. However C1 spec ('"ultra-low SAPS") in Oz (for Mazda diesels) is less commonly available, at least at retail level.

Cheers,

 

[dpm]

According the factory service manual, either C2 or C3 is good, at 0w30 or 5w30 viscosity.

If it meets those, you're good.

 

[kutavyz]

Hi Kutavyz,

I must ay my vehicle has been dealer serviced. However they source their oil from normal manufacturers and I'm not even sure the dealer needs to go through Subaru for that.

Go by the spec (C2) rather than the description ("low SAPS"). If it meets your car's specs then it should be good enough. Just check there aren't other specs to meet (forgotten which) too.

C2 is a good common spec. Plenty in Australia even from local manufacturers/blenders. However C1 spec ('"ultra-low SAPS") in Oz (for Mazda diesels) is less commonly available, at least at retail level.

Cheers,

According the factory service manual, either C2 or C3 is good, at 0w30 or 5w30 viscosity.

If it meets those, you're good.

Thank you so much, both!



Enviado desde mi Nexus 6 mediante Tapatalk

 

[pitrack_1]

Solution to Problems

Been a bit quiet lately here, that's 'cos of the Great Password Kerfuffle and I've solved most (if not all) my drivability issues.

"What's that??!" I hear you all cry.

"Problems solved???!!"

Weeeelllll, sorta- in a manner of speaking! :biggrin:

The main issue was driveability especially under regen. Well that is no more. I've managed to reduced the lag under regen to almost zero. Only those without any mechanical sympathy at all will find the merest hint of a delay under acceleration, so small you only notice when looking for it.

Regens are further apart and more consistent, approx 500km between regens. The regen engine note has changed too. It's now arguably quieter and smoother under regen than not.

I've had a small tweak in power (to 110kW ) and torque (to 360Nm). Max torque spread is better too, 1800-2600rpm and the rolling acceleration is really sharp.

Fuel economy is much better to boot (~800+km from a tank) although still goes through the roof under regen.

Gear change is far better and the clutch behaves itself.

I'm likely to get a ticket if I keep enjoying it in traffic too much! Yessirree, there's a deal of "yee-har!" to driving my new-old 2011 Mazda 3 BL2 MZR-CD !

Sorry, just had to do that! :redface: Mind you, it's a lesson to Subaru on just how to build a diesel pax vehicle engine. My confidence has been restored. Sat nav, Bluetooth and several other things are included too. It's only EURO4 but it still does active regens. Dunno open or closed loop.

 

[Tvilleguy]

Hi Pitrack1 and other users...
I couldn't find a file to import the new PIDs anywhere... I've created all the new PIDs into a CSV file for Torque Pro... I haven't tested them as yet but they load into the extended PIDs

File attached... let me know if there are any problems with them.

 

[pitrack_1]

Thanks for the effort to transcribe and post the new PIDs tvilleguy!

I'll give them a go sometime.

Cheers,