Injector driving

[clem]

Now I get it. Nice job!

[PBU]

Hi

Did the first test with the car after changing the dead time values to the real values from my test.
It completly changes the mapping !!
Not only ,as expected of course , the VE values themself but also the way the AFR reading reacts to the throttle opening or closing.
Reaction to throttle is much more linear and logical ( better ) as before

So my advice is before start mapping be sure the dead time of the injectors is really correct.
It is way more important then you think and not only a time factor to add to the PW.
I have found out now.


Grts
Patrick

[clem]

Hi Patrick,

I tried to do measurings, to determinate the deadtime of my original SR20DET Sidefeed Injectors but discovered problems.
The deadtime of the seems to be influenced by the pulsewidth itself, which theoretically should not be possible (in certain Duty Cycle regions).
Did your measurements put out similar dependencies? Please see my attached picture.

The only explaination I have for this sympthom is, that the fuel Pressure is too much inconsistent over the whole range. I use a normal RB25DET Fuelrail with OEM Fuel pressure Regulator and approx. 0,5m Fuel line with a Fuel Filter inline. Did you use some certain Fuel pressure Dampeners or similar to achieve a real CONSTANT Fuel Pressure?

Thanks

[Karel]

Clem i am not Patric, but i think that everything over 20ms is little too long

what we have:
6000RPM and 1injection per 2 revs

so 3000/60 = 50injections per second => 20ms per each injection at 100% DUTY (your 60ms means, that you want 100% DUTY at 1500RPM, and i think this will never happen)

Now what i think is, that if your pulse width starts to be too long, the coil in injector start to warm up, and the resistence of the coil start to climp up also..... result, less current and less magnetic field to move needle inside the injector and larger dead time.

Also diference 0,3ms of death time at 60ms pulse is very low, and will not make huge AFR change.

[PBU]

Hi Clem
I can't see the picture from here but Karel has right.

A realistic max PW is 20-24 msec.
I tested 2-4-6-8-10 msec.

I made a fuel supply by putting 3 bar air pressure on a little container with 1L petrol.
This way i could control the pressure very good and i didn't need a fuel pump.
You can just see the system on my youtube movie.

What is your max and hold injector current ?

Grts
Patrick

[clem]

Well Pbu , you may have to log in to see it ?

AAAh Ok I did not think very much about the cycle time as it should be not significant to the results. But it makes sence as you explain it!
Ok thanks for the Explaination.

I used High-Z Injectors with no current measuring... But as they have approx. 11Ohm Resistance, there should be about 1 A current, the whole cycle long.

I will test soon with your given pulse and pause times and report

Kind regards

[PBU]

Well Pbu , you may have to log in to see it ?

AAAh Ok I did not think very much about the cycle time as it should be not significant to the results. But it makes sence as you explain it!
Ok thanks for the Explaination.

I used High-Z Injectors with no current measuring... But as they have approx. 11Ohm Resistance, there should be about 1 A current, the whole cycle long.

I will test soon with your given pulse and pause times and report

Kind regards

Clem , If you send my your email by PM i send you my excel file so you can fil in your values..

[clem]

Hey that would be great! spaxsi(at)gmail.com is my address.

cheers

[bmwmaster]

Hi

Did the first test with the car after changing the dead time values to the real values from my test.
It completly changes the mapping !!
Not only ,as expected of course , the VE values themself but also the way the AFR reading reacts to the throttle opening or closing.
Reaction to throttle is much more linear and logical ( better ) as before

So my advice is before start mapping be sure the dead time of the injectors is really correct.
It is way more important then you think and not only a time factor to add to the PW.
I have found out now.


Grts
Patrick

Congrats to this nice result,
but now i somehow don't want to continue my map tuning until i know the real dead time of my injectors
I didn't understand your method 100%. Can you maybe make a kind of manual how it works.

I would easily buy an osziloscope to get real dead times

[PBU]

"but now i somehow don't want to continue my map tuning until i know the real dead time of my injectors "

This is very sensible and i would advise that also.

The principle is very simple.
You need to be able to send a programmable number of squirts of a programmable PW at a programmable supply voltage to a injector.
Then you measure the fuel output of every PW and voltage and fill this in my excel table et voila.

Easy said but not so easy done

I did this with a little arduino board as you can see on my youtube movie.

https://www.youtube.com/watch?v=66QZ4JVlgyI

You don't realy need a scope.
But if you want to buy one not to expensive : i just bought a Rigol 1054Z. Unbeatable for the price of 350 euro !!


Grts
Patrick

[Maciekx]

How to define injector dead times using the osciliscope?

[PBU]

How to define injector dead times using the osciliscope?

Read first page of the thread.

[clem]

Hi
I still am confused about the correctness of this measuring method. PBU your estimated deadtimes (of the two injector types you sent me) are also depending on the pulsewidth
You let excel calculate the deadtime via a statistic model, I am not sure if this really IS correct Statistics is not my favourite friend...
Could it be, that the closer the pulsewidth to the deadtime, the more exact is the deadtime calculation?
cheers

[PBU]

Hi Clem

I send you 2 examples of the calculation.
One gives a graph at a specific battery voltage with dead calculation.
The other one is the same calculation!! over a given battery voltage spread.

I can assure you this method is widley used and correct.
It has nothing to do with statistics . It's only the mathematical calculation where the graph line would toutch the horizontal axis if it would run further.
This value is the dead time value.
Meaning , during this puls time the injector gives no fuel output.(lost time )

Only issue is that the injectors in the very low PW are not linear and so with the excell calculation here whe have a little error.

Hope it makes sense.

[clem]

Thanks for the explaination! I always thought you would like to calculate other times within to the deadtime (like explained in the graph on the first page...)
Now I understand Thanks again!

[bmwmaster]

Hi PBU,

im now building an Arduino Injectors controller as you did. Can you maybe send me your excel table to make the calculations
I already sent you an pm.

Thx

[bmwmaster]

As written in last post, im now also building an injector test station.

Which injector out MOSFETS are installed in the EMU ?
I like to use the same, to be sure getting the same results.

[Jadzwin]

VNP20N07 however most of mosfets will work for you