Dead time injector testing - "for hire"?

[Dozer2FEYota]

I simply don't have the time, and other life/work issues have kept me from setting up a bench test for injector dead time. Since this is out of an vehicle uncommon to Megasquirt installs (Toyota Landcruiser) I haven't been able to find it published anywhere. I have my motor running reasonably well, but the need to make changes to the VE table for steady idle, on a fairly regular basis have me thinking dead time is off some.

I'm hoping someone on here has a bench test setup I could "hire" you to test an injector for me. No idea what a fair cost would be, but any takers?

[whittlebeast]

Do you have a big data log off the motor driving around on the street? If you do, post the data here.

[ChevelleFan]

Do you have a big data log off the motor driving around on the street? If you do, post the data here.

I've had this topic bookmarked and was hoping to see what the methodology was. My car runs pretty well too, but I never measured the dead time. But if it can be gleaned from a street-driven datalog, I'll gladly record/upload a log. How big of a log is needed?

-Dave

[whittlebeast]

A half hour drive is fine.

[Dozer2FEYota]

Somehow I missed this offer for help. Here's a link to a pretty long log... thanks!


https://drive.google.com/open?id=0B8Lym ... U9HVkx6aXc

[whittlebeast]

Dozer

To view your data, the first thing I do is select the optional field MAPxRPM in a registered version MLV and reload. Next, go to Scatter Plots View and open two view boxes.

Next, I select on the left pane RPM, MAP and AFR. On the right pane, I select MAPxRPM, Duty Cycle 1 and AFR. In the filters setup, I like to create TransientMA and define that as [Engine]>1

On a motor that is a true Speed Density Motor, the right pane will have a straight line. Following along so far?

Andy

[whittlebeast]

Here is what my hyper 300 HP 1800 cc blower motor right pane looks like.

http://www.nbs-stl.com/MLVDemo/026%20Sc ... gMotor.png

I happen to have VE as my color in that screen shot.

Here is the same motor with RpmPerSec as the color.

http://www.nbs-stl.com/MLVDemo/800%20Mo ... d%20PW.png

Notice the two distinct lines. A motor with injector dead times issues will have a wide line on the bottom left. One other of the dead giveaways that the Injector dead time is messed up is the motor does not have a steady idle when you turn on PWM EGO control.

Andy

[Dozer2FEYota]

I follow.

[Dozer2FEYota]

At first this plot looked pretty fat at the bottom left, but it cleaned up pretty well when I filtered for deceleration. Thoughts?

[whittlebeast]

Yours is not that bad. Remember that if you mess with opening time, it messes with the tune across the board. On your tune I would address two things. I like to run closer to about 14.2 AFR most of the time below about 75 KPA. The other thing I am suspect of is it appears that you could use way less AE. On normal trace screen, on the top trace, turn on MAP, RPM and TPS.

On the second trace, turn on MAPxRPM (essentially air flow or MAF) and Duty Cycle 1 (essentially fuell flow)

I want to see about a 10% jump up in Fuel Flow for a short period of time and a clean increase in RPM (never a drop) as you stab the throttle. Any more added fuel is not really necessary.

Turning MAP and PW on a third trace also helps on a reality check.

Hope this helps

Andy

[whittlebeast]

see http://www.msextra.com/forums/viewtopic ... 34&t=65200 for more on this AE and tuning stuff.

Andy

[Dozer2FEYota]

Thanks Andy, guess I'm on to other things to try and address my somewhat unstable idle. Since this is a fresh rebuild, I'm really praying it's not something mechanical. I did run an older head with a carb for initial startup and breakin, and it ran quite smooth there at idle.

Cheers!

[whittlebeast]

Dozer see http://www.msextra.com/forums/viewtopic ... 34&t=65200

That is your data.

Andy

[Dozer2FEYota]

Dozer see http://www.msextra.com/forums/viewtopic ... 34&t=65200

That is your data.

Andy

Very cool Andy. Much appreciated and nice of you to commit your time to help the collective community on tuning as well!

Cheers

[dll67]

Andy,
I envy your ability to understand MLV histograms and scatter plots.
For me I have been trying to nail my own tune down, off-n-on, for 3 or so years, in between breakage and the winter months. For me I dont have the MLV vision you do and I am left with asking questions in the forums and reading MS manuals to try to nail down answers, as in this instance, injector dead times.
The information in this quote is an example I found on injector dead time ...
" Please Note:

If you have lo-impedance injectors (less than 4 Ohms) , after getting a good idle you need to adjust the PWM duty cycle on the Constants Page in TunerStudio down in 1% increments until you notice a change in idle quality (be sure to hit the "send to ECU" button each time you change the value). This is the point where the current limit is too much and the injectors are not being held fully open. Then move the value back up 3 - 5% (for example, if the idle falters at 45%, then put in a number of 48% to 50%). Move on to adjusting the time threshold. Lower the time threshold by 0.1 milliseconds at a time until the idle quality deteriorates. Then increase it 0.3 ms. "

I recently rediscovered this quote here ..... http://www.msextra.com/doc/ms1extra/MS_ ... Manual.htm about a third of the way down the article.
When I first read this "Note" I am positive I had no idea what it was trying to describe or what these directions were going to do for me. In fact I had no clue the second time either. I have no idea how many times ive read this article but I turn to it and several others on the basics when I need to take a step back and look at the problem with fresh eyes.
When I read this paragraph this last time I recognized that it was describing a way to fine tune injector dead time, despite it being where it is in a article about the "basics to getting started". I'm not sure how it got there in a manual for people new to the MS world or without a meaningful description of what it is supposed to accomplish. I'm happy I found it though, and it led me here, after some searching, to share and maybe help someone else who is scratching their head.

Anyway thanks Andy, I look forward to trying your method with scatter plots to home in on my injector dead times.

[dll67]

Andy,
Heres a screen shot of MLVHD scatter plot. i'm not sure but is the lower left of the plot on the right showing a dead time problem?
Among other problems

[size=150attachment=0]inj dead time.jpg[/attachment/size]

[whittlebeast]

You do not have a serious issue with dead time. Where you have that hook up on the right end, you can also see the color turn rich so everything makes sense. Try turning on EGO or GEGO as the color next.

Andy

[whittlebeast]

This is what a plot looks like that has opening time issues. In this case, the injectors have to be opened way longer than you would normally expect at high RPM and in a downshift, so low MAP. It is not really causing me any issues that concerns me as this is a pure race car. I included some of the data as the motor was warming up for the race to get some low power data in the plot.

See the area in the red circle. It is far from a single relatively straight line.

http://www.nbs-stl.com/tuning/Plot%20wi ... issues.png

Andy

[dll67]

You do not have a serious issue with dead time. Where you have that hook up on the right end, you can also see the color turn rich so everything makes sense. Try turning on EGO or GEGO as the color next.

Andy

I'll have to do some datalogging with egocor on and report back.
Thank You

[dll67]

egocor didnt give the expected results

egocor.jpg