question on injector dead time

[Rick Finsta]

NO! They'll pop out. An Aussie company makes nice little clips that will hold them in place; I haven't been able to make myself pay what they want plus shipping for a set, though. I got one of the clamp-on bung setups from DIY so I can clamp the injector in place, connect it to my fuel pressure sensor -4AN line, and then do the test. I just need to get a replacement injector to stick in the rail while I do it...

[nathanhardy]

It is very likely, yes. Also remember that some injectors (some of the Bosch 160s come to mind) actually get more fuel delivered for lower PW in the 1.2-1.6mS range if I recall correctly, so the 2/4 squirts method won't really work too well for you. Plus small changes in dead time are a very high percentage of idle PW at 4 squirts, which is problematic in and of itself.

As to the scatter plots, if you plot the MAPxRPM versus Duty Cycle, with VE in the Z axis, you should see a linear function from the bottom left to the top right going from blue to red. For injector characteristics, since a small error in dead time makes the most difference at or near idle, if your settings are off you will see a very wide blob in the lower left corner instead of a thin line, which will thin out for the rest of the graph.

The tuning method in the manual to which I refer is actually for setting PWM constraints, but since PWM threshold and dead time are going to be very close, it will work as well for just finding dead time. Drop your dead time 0.05-0.1mS at a time until idle quality suffers, then add 0.3mS back onto it. This worked for me and I actually ended up with the literature values for my injectors and a nice smooth idle area of my scatter plot.

Think of that scatter plot this way - it should show that as you increase MAP (load) and engine Speed, you get a smooth (close to linear) increase in the delivered fuel to go along with that (duty cycle), and a smooth increase in the filling efficiency (VE) of the engine (blue to red in the Z axis). I realize you're playing with ITBs and split plenums, plus boost (I think?!), so you might find difficulty getting a linear function since actual engine load for you might not correlate well with either MAP or TPS. I'ld just figure out which gets you closer on the scatter plot and that's the method you should use (IMHO).

I'm no good at the calculations needed for the custom field map x rpm needed to do this could i please get some help with thanks

[Yves]

[MAP]*[RPM]/100

[nathanhardy]

[MAP]*[RPM]/100

Thank you! How does this look?

[Yves]

You need to do some driving and see if a shorter dead time yields any positive results. If it was during driving, i think you might go a tad thighter.

[nathanhardy]

That was over a pretty long drive I'll try a shorter dead time but when you say shorter how much I'm running .750

[Yves]

That was over a pretty long drive I'll try a shorter dead time but when you say shorter how much I'm running .750

Try .600

[nathanhardy]

Got this today at 0.575 it was a pretty good drive what do you think now?

[Yves]

A lot cleaner.

[nathanhardy]

A lot cleaner.

What would be a sign of too short of a dead time?

[Yves]

Based on the scatterplots ? I don't think you would be able to see that.

Best is to actually test them.

[Yves]

correct that last one : I think your engine might react differently depending on IAT. So to lean when cold for example.

[Yves]

Nathan,

An interesting read on the subject of scatter plots in MLV : http://www.nbs-stl.com/tuning/

The objective is to get a clear trend in the scatterplots. As for dead time, it's effect shows itself more prevalent in the small PW's opposed to the large one (big power parts of the table) as proportionally the dead time has more influence at lower numbers. So if you would want to evaluate if the dead time is correct based on the scatter plot, you need to be able to vary one aspect of the fuel equasion that has an influence on dead time and see how the reported lambda reacts. Like for instance air temp. Hotter air is less dense, so your ecu should react by injecting less fuel. If your dead time is correct AFR should stay the same.

So you could do a test with a blowing hot air from a hair dryer into the intake of your engine or making it suck air over the exhaust manifold for this test. You would not need to be driving as larger PW will show less of an effect.

[87quest-TSI]

Hi Yves,

Did you ever test to see how close the scatter plots where to your testing?
I'm just a bit curious.

David

[Yves]

David,

I'm not sure what you are asking, but I actually tested my injectors dead time and have set it like that. The line in my case is a pretty thin concentrated line.

[87quest-TSI]

I think you kinda answered my question.
Was just wondering how close the scatter plots where to your fiscal testing of the injectors.

David,

I'm not sure what you are asking, but I actually tested my injectors dead time and have set it like that. The line in my case is a pretty thin concentrated line.