RTT for injectors

[truckmann]

I was just thinking with all the fun people seem to be haveing trying to properly tune for larger injectors that it would be nice to have RTT available for at least the offset table.

[5_Liter_Eater]

I'd say thats next to impossible. All RTT tuning does is allow you to make changes on the fly based on things like AFR Error or STFT or STIT. What would you log/tweak for injector offset?

[truckmann]

Well just from what I have been reading I was thinking something like AFR Error, Ignition Voltage and either MAP or VAC, depending on what Offset table the OS uses?

[5_Liter_Eater]

The thing is there is not a good scientific way to tune the offset table. It's pretty much a guessing game or going off of what others have found to work by guessing or if you took the injectors out of another vehicle then you can copy the injector tables from it's PCM. There are so many things that could affect the AFR error you can't pin it down to offset.

[truckmann]

I guess my thinking was if you had a good stable AFR tune before an injector swap and didn't have any offset data on the new injectors, that one should be able to tune the settings for the new injectors to get everything back inline to where they were with the old injectors. It would just be easier to do with RTT for the offset table instead of having to stop and upload a new tune everytime changes are made. So in effect to help speed up the guessing....lol

[Chris@HPTuners]

There are good scientific ways to tune the injector offset table.

Injector offset can only be calibrated by measuring flow at a known voltage, pressure and injector pulse width. Basically you take an injector and apply a certain pulse width at a given voltage and pressure across the injector, then measure how much fuel comes thru. You keep increasing the offset until you get the amount of fuel you are expecting. You need lab conditions to do this and it requires that you have already calibrated the injector flow rate at each pressure point. One way to calibrate the IFR is just hold the injector open at a certain pressure and measure how much fuel comes thru in a certain time. Even if you just do two points the IFR cal should be linear. Obviously, injector manufacturers should know all these values but good luck trying to get them!

Injector offset is the time it takes to physically open and close the injector from when the pulse is applied and removed. ie. the extra pulse width you need to command to get the amount of fuel you actually want.

To think of it as an analogy, imagine this:

You have a friend who wants you to exactly fill a glass of water using a garden hose. You cannot see the glass itself and have to rely on him signalling you to turn the valve on and off. The tap on the hose is a special spring loaded valve that you need to turn anticlockwise to open and then when you release the handle the spring quickly snaps it closed again. all you need to do is watch for the signal and open the valve then let it snap shut when the signal is gone.

So, your friend is no idiot, he has already done his homework ahead of time. He has measured your water pressure and also estimated how strong you are by trying an arm wrestle the night before after a few beers. He has also secretly worked out the flow rate of your garden hose by turning the tap on full then placing a 10 gallon bucket under it and seeing how long it takes to fill. He is ready for action. The only unknown is how strong you are feeling to twist the handle on the valve... did you get a good sleep last night after all those beers?

Okay so he has worked out the ideal case to fill the glass is 3 seconds. He decides to give this a shot. Not surprisingly, it was close but not quite. The glass wasn't quite full because of the actual time it took you to twist the valve open so the flow was not full for the inital valve opening.

The next time he signals you for 3.5 seconds and amazingly the glass was perfectly full. So he knows that for todays water pressure and your current strength level he needs to add 0.5 sec to his signalling to you.

The links here are:
strength = voltage
water pressure = pressure difference across the injector
hose flow rate = injector flow rate

Your friend then tries something, he goes out to your water meter and closes the main valve a lot to reduce the water pressure to half what it was before. Goes and measures the hose flow rate at this new pressure (it should be half) and works out he needs to signal you 6.5 seconds this time (6 seconds to account for the new water pressure and .5 seconds to account for your non-ideal opening of the valve). This time the glass was overfilled a small amount. His first reaction is to blame you, so you try the test another 10 times each time with the same result. He then decides to try 6.3 secs and voila! the glass is perfectly filled. What happened? Well it turns out that this particular valve gets easier to open and close when there is less pressure difference across it, so the lower water pressure meant you could open it faster than before.

You should also, be able to see that the final test would be to make you do 100 pushups and then try the test with reduced strength. The result would be he needs to signal you for longer becuase you are so weak now it takes you a lot longer to open that valve. You can see these effects in a regular IFR table, lower pressure difference needs less offset, lower voltage needs more offset.

So this is the theory of tuning the "garden hose operator offset"

i need a drink...

[5_Liter_Eater]

Alright, I guess I meant no scientific way to use HPT histos to automagically RTT your injector tables.

In the above example are those Foster's size beers or standard US size beers?

[truckmann]

Ha nice story/explanation.
So with HPT I would think that if you can trust your MAF and/or VE table and trust the wideband and you are confident that you have the correct IFR data then you would be able to tune the offset values around that data for proper AFR instead of skewing say the VE table around the offset data that may not be correct.

[Chris@HPTuners]

Ha nice story/explanation.
So with HPT I would think that if you can trust your MAF and/or VE table and trust the wideband and you are confident that you have the correct IFR data then you would be able to tune the offset values around that data for proper AFR instead of skewing say the VE table around the offset data that may not be correct.

That's a lot of trust! In theory it is possible but in practice the accumulated errors you have made to get that far would swamp the magnitude of the parameter you are trying to adjust. Remember, this is only one part of the puzzle, the PCM then has a model for the fuel to actually get into the cylinder (transient fuelling) then the thing has to burn and the wideband has to measure it correctly.

[alanderson1978]

Excellent explanation Chris. Finally, injectors explained in a way I can understand.
Now we've got 'Flow Rate Modifier vs. Volts', set to '1' across the table. This would be if your injector doesn't open all the way at reduced voltage?

I think I understand what Short pulse adder, minimum pulse and default pulse do, I'm just not sure why we need them.

[Chris@HPTuners]

Now we've got 'Flow Rate Modifier vs. Volts', set to '1' across the table. This would be if your injector doesn't open all the way at reduced voltage?

At reduced voltage there are a couple of effects, firstly the injectors have less "strength" to function with and compensaiton for this is in the injector offset table as discussed above.

The other effect is the fuel pump has reduced voltage and doesn't deliver the same fuel pressure. The Flow Modifer vs Volts table is used to correct for that. Usually its not calibrated by GM though for whatever reason.

[4.8t]

so what about RTT for short pulse adders. say you have a given offset table. but the injector missfires at lower pulse widths. if you could RTT the adders you could simply raise or lower the adder until the missfire goes away and the afr comes in line. assuming you had a know offset table and your ve/maf was tune prior to injector swap.