Staged Injection table - not working like it should?

[Speedy_G]

Hello,

i've got 8 Injectors in use with the fuel outputs on MS3x on a four cylinder engine.
Each connector "single wired", so from INJ A-H
Firmware is latest development 1.5.1 beta 1.

I've set "fully transistion to secondaries" to ON.
According to the manual in this case both injectors should have different pulsewidth according to the table.
So 100% should be only secondaries and i would think 40% should be 60% primaries and 40% secondaries.

But now i've made some tests. While idling i changed the value i the staged injection table.
So ive tried 20%, 40%, 100% and so on.
(I changed all values in the table)

But when i use for example 20%. The secondaries come active but the Pulsewith for primary and secondary is nearly the same??
And the bigger problem, the pulsewith for PRI and SEC is the same like only using the primaries.
So this results in a much too rich air/fuel mixture. As nearly the double mass is injected.

When using 100%, everything seems to be correct. So only SEC is used.
But for 90% again PRI and SEC have nearly the same pulsewidth.

So is this a bug in the firmware. Or something else?

MSQ and Datalog is in the attachement.
In the Datalog i varied the table. Cant remember the exact values. I thougt its saved in the datalog, but it isnt.
But i used numbers like 0%, 40%, 50%, 70% and 100%
You can see that AFR gets much richer when the value isnt 0% or 100%!
Okay i had to delete a lot from the datalog, to make it smaller than 1MB.
But you can still see it.

[Dennis_Zx7r]

Can not reproduce, also using 1.5.1b.
The percentage set seems to correspond to the injected fuel mass.



You set the flow rate for the secondary injectors just like the flow rate of the primaries. I doubt you did this on purpose, as the secondaries usually are much larger.

[Speedy_G]

The injectors are the same.
So same flow rate for PRI and SEC. Also same DeadTime and so on.
(Idle PW is okay. In the log from yesterday around 1.7ms. Today 2.2ms when changed from semi to full sequential, see below. )

Maybe same injector size is the problem for the Megasquirt i dont know. Just took a short look into the source code and somehow its calculating the PW by the flow difference of the PRI and SEC injectors.
But i didnt really understand the whole calculation yet.

Had a different mistake in my config. Somehow i had 1 squirt per cycle but only semi-sequential. Dont know how this was working at all.
But i've changed to fully sequential today. But this didnt help either.
The difference now between PRI and SEC PW is a little bit bigger, and you can see how it varies with changing staging factor.
But its still nearly the same and results in way too much injected fuel.

But for example when i run only 1% Staging factor. The PW of SEC Injector is still way above the dead-time of the injector.
So still around double fuel mass is injected. Because the PW of both Injectors is similiar to the PW of only PRI Injectors working.
Similar for 99%, just the other way around. So there PRI Injector is way above its Deadtime in Idle. And still double fuel mass.
Then when you just add 1% (99 to 100%) PRI INJ is off, and my AFR gets whole 4 numbers leaner!! For example from 12 (too rich with staging) to 16 (too lean with only secondaries in idle).

And another curious thing i've seen. When using staged injection the calculated fuel flow in TunerStudio standard gauge gets negative!
So its seems like i'm producing petrol there. So another bug.


But could you please try different staging percentage on your setup?
Because 0% and 100% are treated as special cases inside the firmware.
For everything between they use a different calculation.
And for 50% i get same PW for both PRI and SEC, too.
You can try for example only 1% or 99% staging.


Edit:
To get some numbers to that example.
Idle PW with only PRI INJ running: 2.22ms
Then only 1% staging
PW1=2.088ms
PW2=1.752ms
(Deadtime of PRI and SEC injectors is around 0.756ms)

AFR changes from 14.7 to 12.3
By only 1% increased staging factor - this is madness!

[muythaibxr]

50% should be "half staged" or with equally sized primary vs secondary, equal PW on both sets. 40% should be 40% coming from secondaries and 60% from primaries.

[Speedy_G]

50% should be "half staged" or with equally sized primary vs secondary, equal PW on both sets. 40% should be 40% coming from secondaries and 60% from primaries.

yes but it isn't working this way. This is what i descripted. and what my datalog shows.
We could reproduce the bad behaviour with a stim. Try my tune, you will see it. Big thanks to Dennis_Zx7r, who testet it.
Somehow it seems that it adds the DeadTime from Bank1 which isnt used at all?! Because i'm only using MS3x outputs, which is set correctly.
Will try today with DT Bank1 set to zero and see what happens on the bike.

And furthermore we discovered that DT for seq_PW 1 to 4 and 5 to 8 is different too, though everywhere the same DT is entered in the tune.
Maybe we or he found a solution. But i will have to test it on the bike today, to see what happens real.

And dont forget the negative fuel flow. I guess thats not like its like it should be.
And why arent the threshold and hysteresis not greyed out when using table? (see attachement)

[jsmcortina]

I've tested this. Staging looks to be working correctly.

The problem is that you've entered your Deadtime Curve 1 backwards. The voltage needs to increase not decrease.

James

[Dennis_Zx7r]

The problem is that you've entered your Deadtime Curve 1 backwards. The voltage needs to increase not decrease.

Yes, there seems to have been a misunderstanding yesterday. After reentering all DT data it worked on my stim as it was supposed to, and there was no problem with my MSQ with similiar settings and also 1.5.1b firmware.

As with the greyed out fields, this also resolved after settings the staging the RPM-based and reverting to table. Difference seems to be that I use the .18 (release) Tunerstudio and he uses .19beta TS.

[Speedy_G]

I've tested this. Staging looks to be working correctly.

The problem is that you've entered your Deadtime Curve 1 backwards. The voltage needs to increase not decrease.

James

okay perfect thanks!!
Now its working.

I flipped the DT curve and set DT Bank1=0.
But i had to double the req-fuel to get same AFR with VE=30 at idle as before. And of course had to retune ASE and WUP..
But doesnt matter, now the staging is working as it should.

Negative Injector Flow is still shown in TunerStudio. But not in the datalog.
So seems to be a problem with TunerStudio.

And concerning the greyed out fields. For me it didnt help to set to rpm and then switch back to table.
Those three field are somehow always active. So maybe only a problem in TS 3.19, as its working in TS 3.18 with same firmware.