injector voltage compensation

[fphil]

This is my injector settings



This is what I get when the battery voltage drops by about 0.8 because of the ventils (I got a small battery)


(plots are RPM, IacDC, IntSTEPS, PW, lambda, dwell, battery)

Obviously the compensation of the injector timing for that voltage drop is not enough.
This is a close look



where we can see a small increase of injector PW  command but I guess not enough to compensate for a shorter actual opening because the lambda value get higher.

Do I understand right?. What would be your recommendation?

[VEMS]

If you need more PW increase when battery voltage goes lower, increase "Injector Voltage Compensation". Entered value is value of max extension at lowest VBAT 6v (railed to entered value under that). More details in F1 help for the Injector settings menu.

Best regards, Dave

[fphil]

Yes, but is the lambda computed value stable versus voltage variation of the sensor polarity?
Was this constancy checked ? because the PW compensation I have set is already sensible 0.1 ms

[VEMS]

Do you mean to ask if the lambda reading is dependent on VBAT voltage ? On this i can be conclusive, lambda reading calculation is not VBAT dependent. If you mean something else, please elaborate.

Regards, Dave

[fphil]

If the voltage of the power source for the Vems and the one for the wbo2 sensor happens to not be the same, could there be at some event, per ex. the ventils running, a variation of the lambda value?
On my car I still got an experimental set up of the power lines since I have managed to run the OEM calculator with Vems.
Anyway thanks Dave for your above answer, I shall check this issue on the car next week
Regards, Philippe

[VEMS]

Hello Philippe,

The measuring side from the LSU probe is isolated from the heater side, only heater side is connected to VBAT and temperature of heater (or actually heater resistance) is closed loop monitored and regulated. Output of the measurement cell is not very temperature dependent unless much over or under target, tiny VBAT variations therefore have negligible effect (little to none).

Best regards, Dave

[fphil]

Good enough, so my concern about inj. volt. comp. does not relate to the wbo2.

What about the value of the "Battery compensation" added time which is used for the PW computation.
See the shot here after:



Voltage is 13.65V good, however I get 0.5ms for "battery compensation".
What is that?.
VT help says that compensation is  linearly added down to 7V and subtracted  up to 19V which should makes 0ms @ 13V.

Another shot from the bench
.
Here you see I need to compensate for 0.3V, my injector voltage compensation is 1.2 ms , this gives (1.2/6)*0.3 = 0.06 mV.

Do you get similar results? Have you an explanation? Is it a bug?

Thanks

[VEMS]

Hello Philippe,

Injector opening battery compensation is calculated as follows:

Battery compensation = linear interpolate from "6V, ( Injector open time @ 13.2v + Injector voltage compensation (max 4080usec together) )" to "19V, ( Injector open time @ 13.2v - Injector voltage compensation (min 0 together) )" by current VBATT

Therefore depending on setting of Injector open time @ 13.2, you might or might not get deduction but total battery compensation will never go sub zero. Also for obvious reasons the required fuel part is never altered.

Best regards,


Dave

[fphil]

Hello Dave,

Indeed this is what I have understood from the VT help.

Since I have put Injector Open Time = 0, why do I get Battery compensation = 0.5ms @ 13.53V?
Is it because of the "Injector Effective RampUp" value which should be added to the "Injector Open Time"? so that the "Injector Open Time" never taken less than the "Injector Effective Rampup"

Now, probably there is a Voltage compensation for this "Injector Effective Rampup" which is named "Injector Rampup Voltage Compensation" but I do not know how this is computed.

Also when you say
@6V, battery compensation is : "Injector voltage compensation + Injector open time @ 13.2v (max 4080usec together)"
why "max 4080usec  together"?

Do you mind to explain the calculation for the examples from my pictures?

Thanks

Philippe

[VEMS]

Hello Phillipe,


I see i made an error in semantics in my elaboration of calculation procedure, corrected. But as requested:

Battery compensation image #1(injCompActual2) = { (19v - 13.53v) * 1104usec / (19v - 6v) } = 464,5 usec (rounded to 0.5 in VemsTune variable display)

Battery compensation image#2(injComp) = { (19v - 12.59v) * 1216usec / (19v - 6v) } = 599,6 usec (rounded to 0.6 in VemsTune variable display)

The limit of max 4080usec battery compensated opening time comes from 16usec configuration variable resolution * byte_size = 16 * 255 = 4080 usec. Internal calculation resolution is much higher, but before interpolation we limit variables to min/max byte value (0/255) in this module.

Best regards,


Dave

[fphil]

Alright, thanks.

This not what is said by the VT doc: ".. 540..600usec .... .With 540usec/6V = effective slope is 90us/volt". Equally The doc speaks about 7V not 6V as the low limit.

It follows that "Battery compensation" = 0 only when the battery gives 19V ??
There are some reasons, but this is hard to get.


Regards

[VEMS]

Hello Philippe,

The docs need an update, correct. When i have a bit more time i will update that section.

Perhaps i need to elaborate a bit more what all these variables are for, in VEMS firmware we created a dynamic model for the injector opening/closing times, using VBATT and Pulswidth als input values. All these values together will calculate how much the PW needs to be extended to get the desired amount of fuel from injector (which is in the end what we are after) in your examples you see that to provide 2.5ms of true fuel flow the PW needs to be extended by 740/806 usec.

Best regards, Dave

[VEMS]

Hi Phil,

Nice wire-up on your wiki page: http://www.vems.hu/wiki/index.php?page=MembersPage%2FFPhil%2FInjection

Indeed simplified allows manual entering of curve. As a side note, Traditional Battery compensation calculation as stated is only for the battery compensation part (in calc model) more ramp-up specific battery compensation is in Injector rampup part, these together account for opening time/closing time.

Regards, Dave

[fphil]

"more ramp-up specific battery compensation is in Injector rampup part"
Yes indeed I have noticeD

About injector calibration, do you thing this could done by some new VT tool?
Think that the EGO correction is done by only one injector, eventually for different lambda tagets, one after the other, for a steady run, and put some stat on the data.
For the least we get a comparison for the injector and a way to fill the "Injector trim" menu.

BTW what this "Additive trim" made for?

Philippe

[GintsK]

Hi Phil,

Nice wire-up on your wiki page: http://www.vems.hu/wiki/index.php?page=MembersPage%2FFPhil%2FInjection

Indeed simplified allows manual entering of curve. As a side note, Traditional Battery compensation calculation as stated is only for the battery compensation part (in calc model) more ramp-up specific battery compensation is in Injector rampup part, these together account for opening time/closing time.

Regards, Dave

Long ago I performed tests of traditional method. And find out: no matter what is those four values - output [msec vs voltage] is straight line.