Injector Duty Cycle spike ??

[oztracktuning]

I just noticed that my duty cycle seems to spike at high rpm. Does this indicate anything is not right??

Here is a snapshot of the log. The car has SVO 30 injectors.


http://www.oztrack.com/clubsport/dutycyclepeak.jpg

One frame earlier the injector pulse width was 80% and 14.954m/s

[joecar]

INJDC is a calculated pid using IBPW and RPM (right click on INJDC pid, More Info)...

IBPW looks flat, RPM looks like smooth ramp up, INJDC up swing (at cursor) does not make sense.

INJDC before gear shift is wobbly, does not match RPM ramp up.
INJDC after gear shift matches RPM ramp up.

Or so it seems to me (I only slept 3 hours last night...)...?

[oztracktuning]

I am wondering if fuel pressure is an issue or if i need to set the offset table to suit the injectors.

It would seem to be more likely that the fuel pump is not up to the task?? If that happened does the pcm quickly compensate for lack of pressure by ramping up pulse width??

[joecar]

PCM does not measure the fuel pressure and/or the flow rate;
PCM uses the IFR table to get flow rate.
Your IBPW is constant at 15.6ms (15.595ms).
For constant IBPW, the INJDC goes up as RPM goes up (the cyclic period decreases, making duty cycle a larger percentage).
INJDC is calculated by Scan Tool.

Regards
Joe
:beer:

[Blacky]

The IBPW scale is too large to see anything useful, try changing it to range from 0 to 20 - instead of 0 to 500. You will see the IBPW jump up, matching the duty cycle.

Regardless, based on the PLX output, the mixture gets richer without the commanded AFR decreasing. Maybe it's Cat overtemp protection kicking in.
I can't remember if the extra fuel for COT shows up in the commanded AFR or not.

Regards
Paul

[joecar]

The IBPW scale is too large to see anything useful Ah, I missed that, I just saw flat...

Ha, here's my problem...

...(I only slept 3 hours last night...)...?

[Delco]

The IBPW scale is too large to see anything useful, try changing it to range from 0 to 20 - instead of 0 to 500. You will see the IBPW jump up, matching the duty cycle.

Regardless, based on the PLX output, the mixture gets richer without the commanded AFR decreasing. Maybe it's Cat overtemp protection kicking in.
I can't remember if the extra fuel for COT shows up in the commanded AFR or not.

Regards
Paul

May also just be the VE-s are too high at that point , as the efficiency of the motor falls off the VE's need to also fall.

Change your scaling of the parameters to also reflect the expected min max values .

Also when using a injector such as the 36lb ones there is no right or wrong inj flow table , take the 36lb injector as a example , the SVO ones have a miles different flow charteristic than the Holden Supercharged ones , at 100% duty cycle they both flow the same but at every other % duty cycle the Holden ones flow different to the SVO ones , We can always get good control with the supercharged disc injector at low openings , the SVO pintle injector has problems flowing at low % openings - not really related to this thread but I noticed you asked in another thread.

[oztracktuning]

Thanks Delco
I am not sure if the 6400rpm cell in the VE table represents
1. starting at 6400rpm ending at 6799rpm and needing to cover the whole range or
2. Blending with the 6000 at 6200 and the 6800 cell at 6600?

Which one is it.

I just got another log of the same thing today.
Here are the numbers rather than the pic.

http://www.oztrack.com/clubsport/dutycyclepeak2.jpg

[oztracktuning]

My battery voltage is doing something funny??

[Blacky]

I am not sure if the 6400rpm cell in the VE table represents
1. starting at 6400rpm ending at 6799rpm and needing to cover the whole range or
2. Blending with the 6000 at 6200 and the 6800 cell at 6600?

Which one is it.


Do a search on this forum for "interpolation". That explains how the PCM figures out the values "in between cells".
Regards
Paul

[joecar]

The images are gone... this can't be me...?

[Blacky]

The images are gone... this can't be me...?
I can still see them - maybe you can't reach the server on which they are hosted - they are not hosted on EFILive's forum as attachments.
Regards
Paul

[Delco]

Thanks Delco
I am not sure if the 6400rpm cell in the VE table represents
1. starting at 6400rpm ending at 6799rpm and needing to cover the whole range or
2. Blending with the 6000 at 6200 and the 6800 cell at 6600?

Which one is it.

I just got another log of the same thing today.
Here are the numbers rather than the pic.

http://www.oztrack.com/clubsport/dutycyclepeak2.jpg

Have a look at your 6400 VE , might find its a bit higher that the rest , at 6400 rpm it uses the 6400 ve point , at 6200 it interpolates between the 6400 and 6000 rpm points . post up your ve numbers for 95, 100 and 105 kpa ranges.

Also remember the AFR measured lags behind the data logged as it happens after the event due to the placement of the o2 sensor.

[joecar]

The images are gone... this can't be me...?
And now I see the images...? :nixweiss:

[oztracktuning]

Thanks everyone
The VE table had
5600rpm 0.954030
6000rpm 0.935237
6400rpm 0.886864
6800rpm 0.867144

which in g/second is
5600 356.171297
6000 374.094740
6400 378.395166
6800 393.105201

[oztracktuning]

I richened up the 6000rpm cell and it fixed the lean area before the spike but made no difference to the rapid increase at 6400rpm

Today it was
6368rpm 80% DC 15.015 pw 100map 0.85g/cylinder 13.1V IAT 33 4.57 flow
then it went in one frame to
6428rpm 85% DC 15.778 pw 98 map 0.83g/cylinder 13.1V IAT 33 4.57 flow

The afrs were 12.7 but went to 12.1 one frame after this happened.

Interesting today is that there was no knock retard during the whole area of WOT at the same timing.

[oztracktuning]

Leaned out the 6800rpm cell in the VE and tested it today. It still gave a sudden surge in INJDC at just over 6400rpm

At 6395rpm it was at 77% DC pw 14.420ms and 'happy' with 12.6 measured afr from des afr 12.75 then in one frame it went up to 6445rpm at 84% pw 15.565ms and then a frame later afr was at 11.9 - it stayed like a few frames but by 5 frames later at 6719rpm it was at 77% pw 13.672ms and 12.6 afr again.

[oztracktuning]

Still getting this spike that last 2 frames just after 6400rpm. I just calculated what the pulse width should be and its about 1ms too big??

Can LS1s compensate or something if fuel pressure drops or could it be air making it to the injector?? Or something weird. Other tables we dont have etc.

I have the stock late 99/ early 00 pump and tank - just using SV0 30s with the stock voltage offsets (so far) and the IFR table that has 4.5547 at the top.

[joecar]

The PCM doesn't know what the fuel rail pressure is, it doesn't have a sensor for this.

[oztracktuning]

Something weird is happening my VE numbers etc. Dont seem to be able to support such a quick change in pulse width ??

[oztracktuning]

I made my VE table from 6000, 6400, 6800 rpm - all the same in grams/second and ........it still spiked upward by 0.43 milliseconds against the trend at just over 6400rpm. ??