[rusEfi] Micro Rusefi for a Renault K4M engine #58

[JPh]

My project is to tune a MRE for a K4M engine. The idea is to have a stand alone ECU for trackdays with its own loom and easyly go back to OEM for street use. I will share the steps I follow so perhaps readers will not have to search on the forum, wiki, github, slack channels to find some informations.

[JPh]

First step : connect the MRE to tunerstudio.
Some problems with the drivers (see micro rusefi connection to tunerstudio https://rusefi.com/forum/viewtopic.php?f=5&t=1671).
Now everything is working and some hours will be necessary to use all the features.

[JPh]

Next step : cut an OEM loom from a spare engine to make the mre one. Need it to calibrate the sensors and for the bench tests. With the mre connector it's easy to make.

[AndreyB]

I will share the steps I follow so perhaps readers will not have to search on the forum, wiki, github, slack channels to find some information.

We need more threads like this, thank you!

[JPh]

You help me a lot making such a product available. Open source and co-desing are great!

[JPh]

Today construction of the flywheel drive.
Why bother with hardware when we can simulate it?
Because real world signals learn you more about functioning, gap limits, voltage increase with rpm, sync loss...
I have all I need handy, so I jumped in.
À VR simulator is also on the list, I don't want to spin my heavy OEM wheel 10000 rpm...
When I will go for an aluminum flywheel, the different trigger wheel will also be worth a spin.

[JPh]

It's running!

15848068282586754662431216785393.jpg (2.28 MiB) Viewed 27394 times

60 to 5000 rpm

[AndreyB]

60 to 5000 rpm

Can you safely go above 5000rpm?

[JPh]

I'm limited by motor power, and fear getting the flywheel flying...
High rpm test will be done with function generator.
I will capture the signal at 5000 and feed the wave form in the generator and then go up in revs.
But I have to find an amplifier to get the right voltage level.
I will post a 60 and 5000 rpm sample for reference purpose.

[AndreyB]

Thank you for amazing content!

[JPh]

I hope there be more to come!

[JPh]

Here are the waveforms for low and high speed :

slow speed.png (7.78 KiB) Viewed 27379 times

high speed.png (7.95 KiB) Viewed 27379 times

If there are some data scientists in the community : attached the samplings corresponding.
Sensor gap 0,5mm.

[mck1117]

Here are the waveforms for low and high speed :
...
If there are some data scientists in the community : attached the samplings corresponding.
Sensor gap 0,5mm.

Signal looks great! Should do just fine with MRE.

[JPh]

Some data calculated from the above test :
Rpm 71 peak to peak 1,1v
Rpm 5555 peak to peak 88v

[dexter_lab]

Looks VR by the book the TLE8888 will handle it. It will be interesting to see how it look like when ignition is attached. Do you plan to simulate with hardware as well?

[JPh]

Yes my MRE will be tested with hardware up to 5500 rpm.
From that point to max possible it will be tested with a signal generator.

[JPh]

Now that injectors and coils are running, it will be time to test sync with spinner

[JPh]

Not much time to test today.
First run the vr sensor was wired inverted : there were two big spikes on the fast logger with the shorter first. Mre worked with 60/2 trigger but timing was not on mark.
Inverted the vr wires spikes are in order after but 60/2 trigger didn't work anymore.
Switched to 60/2 vw had sync losses at low speed.
Need more research and work on the system.

[mck1117]

Not much time to test today.
First run the vr sensor was wired inverted : there were two big spikes on the fast logger with the shorter first. Mre worked with 60/2 trigger but timing was not on mark.
Inverted the vr wires spikes are in order after but 60/2 trigger didn't work anymore.
Switched to 60/2 vw had sync losses at low speed.
Need more research and work on the system.

What do you mean timing wasn't on mark? The trigger logger page in TS should show 58 short teeth, then one tooth that's 3 times the height. You also don't need to swap the wires - there's an option in the trigger input page to invert the trigger polarity.

[JPh]

To test I ran on fixed advance, setting it to 0.

I don't know what is 0 advance for rusefi.
On other systems it's during the large tooth and the next normal tooth is number 1.
During the different tests I made, the timing light showed a point in the large tooth, this point changed dured the different test I made.

But this was not the problem for me : the problem was with inverted vr 60/2 setting for trigger worked but shifted with rpm.

With normal vr wiring 60/2 stopped to work and 60/2 vw worked but only at high rpm. At slow rpm the mre lost sync randomly for a short moment an re synced and so on.

[JPh]

After some more work and with Gwendal's help, mre is working now.

My problem was an operator's error. I mis interpreted TS fast logger : I thought that the mid long tooth followed by the long tooth was the sign of an inverted vr polarity. This was wrong, if inverted both 60/2 and 60/2 vw were no more working.

With a working sync I could check ignition accuracy.
First with an old timing light. I saw some stability problems that needed more checking. With the scopes, the problems disappeared so my timing light has triggering problems.

I decided to make a new test setup to have real world results. I know you cannot pick up the vr signal to sync the scope. There are common ground problems which make the ecu malfunction.
So I added an other vr sensor only for scope sync :

IMG_20200326_182209.jpg (2.71 MiB) Viewed 27069 times

There is an angle between the two sensors, So I set the fixed ignition advance to 2 degrees which compensated this angle.
Here the résultats for different rpm :

750 rpm

Image750.png (6.73 KiB) Viewed 27068 times

2500 rpm

Image2500.png (5.96 KiB) Viewed 27068 times

4000 rpm

Image4000.png (7.04 KiB) Viewed 27068 times

5700 rpm

Image5700.png (7.25 KiB) Viewed 27068 times

For me these are very good results.

[mck1117]

Looks great, that's a good test!

[JPh]

This (sunny) afternoon, MRE found its place under the hood of the Secma F16.

It's just over the oem ecu, attached to the 4 ignition drivers who are clamp bolted to the frame tube.
The loom will be easy to route over the oem one and the ecu is on the cool side of the engine.

[AndreyB]

Oh it's behind the driver?! https://en.wikipedia.org/wiki/Secma_F16

[JPh]

Yes, so the USB cable won't be too long to log for tuning.

[mck1117]

Hey, that thing is just an 80% scale Lotus 340R!

[kb1gtt]

Nice.

[JPh]

Some more pictures for those who need inspiration about mechanical setup :

IMG_20200405_094311.jpg (2.35 MiB) Viewed 26588 times

IMG_20200405_094345.jpg (2.29 MiB) Viewed 26588 times

IMG_20200405_094324.jpg (2.24 MiB) Viewed 26588 times

IMG_20200405_094409.jpg (2.13 MiB) Viewed 26588 times

[AndreyB]

I am not sure how reliable this would be without rusEfi stickers.

[kb1gtt]

I understand the rusEFI sticker gives you an extra 10 HP.