Rotary Engine VE tuning

[Talkwrench]

OK, Faulty ECU replaced, new one installed and all CAN channels are now working, the car is running   I've started calibrating the idle and low throttle angle areas of the VE table but its needing some really large % numbers to achieve the required lambda targets, almost like its twice what you usually expect to see, I have used the auto tune with fill right & down hence the crazy numbers in the table below. Have I missed a setting in the configuration? I can foresee that I will not have enough in the table to tune the higher load areas. The fuel pressure sensor has been calibrated and I'm happy this is accurate, fuel injectors are fully characterized injector dynamics X series. What have I missed?

[rrob]

Your map sensor is reading 52kpa at 0 rpm. Fix that first. 

[ClintBHP]

What is your master Fuel set at ?

[Talkwrench]

Thanks for the quick replies, im still learning this system and any suggestions are good ones  

Your map sensor is reading 52kpa at 0 rpm. Fix that first. 

Good spot, I hadn't seen that. I'll look into that this morning. I have followed all the pre-start info i.e. tps, map calibration as well as timing offset with the light 

What is your master Fuel set at ?

Morning Clint, good to see you on here (its Andrew by the way) . Is there a master fuel number in modeled fuel equation?

Edited July 9, 2017 by Talkwrench

[Adamw]

Please post a log and your map:  https://www.youtube.com/watch?v=_P1LRANeO4A

[Talkwrench]

OK i'll get that done this morning when I take the car out again, i don't really have any logs that a worth posting

Quick question, when I input the primary injector flow rates, is total primary injector flow rate or single injector? Im guessing this could be the reason why the ve table looks wrong, I have input a single injector flow rate

[Talkwrench]

Log file & pclr attached. 

Log file is for a 2Krpm misfire, thought it could be the TPS causing fuel cut, it wasn't. Still haven't resolved the issue yet, probably needs more fuel.

Any help would be greatly appreciated.

 

As for the map sensor values at 0 rpm I can't recreate this and it looks normal to me. Is this something PC Link does when it goes offline? Sets the rpm to 0 but retains all the other current values?

Log file.llg

FD3S TDX51R V1R2.1.pclr

[ClintBHP]

Yes sorry was a quick reply to a common problem, didn't spot the VE part, you're at 70% but only 4ms Injector Duration, what size are your primary and secondary injectors, is your pclr file the same as before ?

[Talkwrench]

1065cc primary & 1763 secondary, there Injector dynamics x series injectors. PCLR is attached above, it's only changed very slightly.

[ClintBHP]

I didnt have much time tonight but here are a few ideas:

Modeled fuel does not have a master fuel setting because it calculates it from information you enter and it inputs from sensors.

A few things I would like you to check:

Have you set your base ignition timing with a light? if not you can be injecting at the wrong time as well as sparking at the wrong angle.

Your issue is common when the injectors are not as big as you think they are, You can test this by reducing the injector size and % Fuel.

I would suggest testing the injectors before going any further, you can check the fuel flow and dead times by using the advanced test on injector 1 and a container with measurement scale (Jug).

With the injector pulsewidth being so low 2.3 nearly half of that can be dead time, get it wrong and it make a lot of difference at idle.

Your Fuel Pressure is low you log shows 285kPa 

[ClintBHP]

Another thought is to switch to traditional mode and see where the Master Fuel is when you get 50 along the NASP line, this should be a quick check.

[Ducie54]

I would have a look into why you differential fuel pressure is not consistent. 

[Adamw]

I had a quick look over your files.  There are a few things that dont look quite right...

1. Charge temp table is set at 100% all over.  This will mean the ECU thinks air is much hotter than it really is (less dense) and remove fuel.   I suspect a rotary doesnt have as much coolant temp effect as your typical "water heated" piston engine manifold and cylinder head so you will probably find you need significantly smaller numbers than typical.  Try something like this as a starting point:
2.  
3. Sec/Primary ratio looks wrong assuming you have 2 primary and 2 secondary - it should be about 1.655
4. Staging table numbers are too big, I would expect biggest number to be about 62.3% for equal PW on both injectors.

With Modelled mode the injector data, MAP and fuel pressure metrics are important, I would trust the ID data to be pretty good if thats what you have used for your injector setup.  However if you can it would be a good idea to verify the FP and MAP sensors against some known good reference like a miteevac or similar.  I'm a little suspicious of your MAP under some conditions.

Edited July 10, 2017 by Adamw

[Talkwrench]

I didnt have much time tonight but here are a few ideas:

Modeled fuel does not have a master fuel setting because it calculates it from information you enter and it inputs from sensors.

A few things I would like you to check:

Have you set your base ignition timing with a light? if not you can be injecting at the wrong time as well as sparking at the wrong angle.

Your issue is common when the injectors are not as big as you think they are, You can test this by reducing the injector size and % Fuel.

I would suggest testing the injectors before going any further, you can check the fuel flow and dead times by using the advanced test on injector 1 and a container with measurement scale (Jug).

With the injector pulsewidth being so low 2.3 nearly half of that can be dead time, get it wrong and it make a lot of difference at idle.

Your Fuel Pressure is low you log shows 285kPa 

OK no worries, many thnaks for finding the time on a Sunday evening, im sure you had better things to do. Yes I have calibrated the triggers with a light using the instructions in the help file for rotary trigger calibration. My injectors are brand new injector dynamics and supplied from a reputable source so i've no doubts about there authenticity. I bought these injectors purely because they are trusted, matched and fully characterised and assumed this would make the tuning part a little easier, i do hope i don't have to qualify the injectors by measuring their output with a jug. Fuel pressure was set at 300kpa with no manifold reference connected and therefore shows lower due to the 1:1 reference.

Another thought is to switch to traditional mode and see where the Master Fuel is when you get 50 along the NASP line, this should be a quick check.

NASP line? 

I would have a look into why you differential fuel pressure is not consistent. 

I hadn't picked up on this and obviously a very good point, i'll check over the hoses tonight and make sure I haven't got anything loose, eveything is connected with AN fittings & hoses.

I had a quick look over your files.  There are a few things that dont look quite right...

1. Charge temp table is set at 100% all over.  This will mean the ECU thinks air is much hotter than it really is (less dense) and remove fuel.   I suspect a rotary doesnt have as much coolant temp effect as your typical "water heated" piston engine manifold and cylinder head so you will probably find you need significantly smaller numbers than typical.  Try something like this as a starting point:
2.  
3. Sec/Primary ratio looks wrong assuming you have 2 primary and 2 secondary - it should be about 1.655
4. Staging table numbers are too big, I would expect biggest number to be about 62.3% for equal PW on both injectors.

With Modelled mode the injector data, MAP and fuel pressure metrics are important, I would trust the ID data to be pretty good if thats what you have used for your injector setup.  However if you can it would be a good idea to verify the FP and MAP sensors against some known good reference like a miteevac or similar.  I'm a little suspicious of your MAP under some conditions.

I was reading up on the charge temp approximation last night and realised the table set at 100% probably wasn't helping, these must have been the default values? I will change the table and see how it affects the VE table numbers. I'll also look at the staging table numbers although looking through the log i don't think the secondary injectors have fired yet, will the staging table affect the primary pulsewidths below the staging threshold?

I have used the data provided by ID for the injector set up but must have missed the ratio numbers, when I configure the primary injector flow rate do I use the flow rate of a single injector or the total primary inejctor flow rate? I suspect if I have this wrong it would impact those numbers in the VE table.

I have verified the fuel pressue with a aeromotive fuel gauge, grated I doubt these are hugely accurate but it roughly confimed the sensors readings. the MAP sensor is also a brand new AEM sensors threaded directly into the manifold, but I haven't been able to confirm its output. I'll get a mityvac ordered.

 

Big thanks for the all input and I will check over the car tonight and see if I can get out for another log session and check for improvements and report back

[Talkwrench]

Updated PCLR file attached.

Changed all the  vac / boost hoses for hydraulic crimped JIC hoses and still no better, made all the suggested changes above to the config and it hasn't reduced the numbers in the VE table by much. I'm all out of ideas at this point, the car is hardly driveable. I'm still waiting on the mityvac to arrive to double check the sensor cals but tallies with the gauge in the line. I forgot to log the last attempt , I'll change plugs and try again tomorrow. Any more suggestions

When setting the primary injector flow rate is this per injector or total primary flow???????? I have 2x 1000cc & 2x 1700cc

FD3S TDX51R V1R2.2.pclr

[cj]

I notice from your first PCLR and log file that you've really only tuned below about 2000rpm, but your test run spends most of its time about 2000rpm where you have compltely untuned 70 values in most cells. Assuming your lambda sensor is set up correctly, then its running very lean (1.2 lambda) for 90% of your log, which explains it running badly. 

There are a few seconds in your log about 1:26 where you're speeding up in a tuned section of the fuel map, and while its not at your target lambda (0.85), you are at least getting a reasonable lambda number (0.95ish).

Try putting sensible guesses into the untuned part of the map, or simply using the mixture map function with a log file to roughly adjust the untuned numbers. If it still runs terribly with a roughly correct fuel map then go looking for other problems.

Have you had a look at the fuel maps in a couple of the modelled fuel base maps? They tend to look roughly flat at 90-110 for the majority of the map above 60-80kpa, drop off to 50-60 below 1500rpm or so, and slope down to 50-60 between 80kpa and 20kpa. There are a few bumps around peak torque, but they dont slope smoothly from 20 up to 100ish like a traditional map does. Google up volumetric efficiency table and have a look at the shape of them for any engine/ecu - even though the numbers may be different than what link uses. I've never tuned a rotary, but if that was a piston engine, I wouldnt expect the <2000rpm fuel table to be far off what you've go so far, so dont actually see a problem with the ~80-90 values you've got in there. Remember, this isnt directly a "how much fuel do I use" number in modelled, its a "how efficient is the engine running vs its theoretical maximum at this amount of engine load" number. so another part of the equation in the background already deals with the fact that you need more fuel at 200kpa at 6000rpm than you do at 110kpa at 3000rpm, these numbers may both be say 105 on a finished fuel map. 

[Talkwrench]

Hi CJ, thanks for the response, its really helped me see past the numbers and just try to tune the car. The previous log file only got up to 2000rpm as it was misfiring hence the really lean lambda numbers (wideband is a LINK CAN module so no real setup needed?), I've added a bunch of timing at the low load & RPM areas and the car is now a lot more lively! I've done some more fuel tuning this morning and I'm starting to get somewhere, even managed a cruise up and down the coast and it sounded epic  Still not pushed it into boost yet but I'll keep refining the cruise areas and creep up into boost when i feel confident. At least I can take the car out now and start logging.

Log 2017-07-16 2;02;32 pm.llg