Injector Offset vs Fuel Rail Temp

[CCS86]

Is anyone successfully logging this PID on a Coyote?

When I add it to my channels list, almost all other channels show no data. I have an open support ticket for that, but am wondering if it is just my strategy.

As I get closer to tuning out transient fuel swings, I have noticed a low pulse width richness during my morning drives. It seems to continue until well after coolant and IATs have reached full temp. It goes away completely once the car gets nice and heat soaked.

It seems like this multiplier table could be the ticket to tuning this out. Right now I am shooting blind though, since I can't log the calculated rail temp.

[Jn2]

Fuel rail temp is inferred since there is no temp sensor for the fuel, same as pressure. I usually set this to all 1's when working with return style systems, or go off injector data sheet from injector company.

[CCS86]

Mine has been set to 1's up until now.

My stock GT tune was making use of that table, in spite of not having an actual sensor.

Not sure if the GT500 has a rail temp sensor or not, but it also uses this table (different values than the GT). I'm using the GT500 injectors, so that should be relevant data.

If I could just log the calculated temp, I could look at my trims and see if a correction to that table is what I need.

[C_Murph12]

Have you seen the worksheet that Paul Yaw released for this? It applies to any injector not just IDs

[CCS86]

Have you seen the worksheet that Paul Yaw released for this? It applies to any injector not just IDs

I haven't. Do you have a link?

[C_Murph12]

Let me see if I can find it. My tuner sent it to me.

[C_Murph12]

What's your email?

[CCS86]

What's your email?

Hey c_murph, I shot you a PM with my email.

[CCS86]

Anyone?

Still unable to log this PID in the latest beta and haven't gotten a reply from support yet.

This table has a very large effect on low PW fueling, since it can modify the injector offset by -20% / +25% using OEM values. Ford certainly would put in this level of offset modification unless it was a real phenomena, so disabling it seems like a bad idea.

[bbrooks98]

I've been meaning to try something with that table. Currently I have all 1s on my car with copied ID850 data. Car runs good but hesitates sometimes in a hot start condition which makes me think i need to do something with this table, that or ID's data doesn't work for my application too well. I'm curious how you'd Infer fuel rail temp?

[CCS86]

I've been meaning to try something with that table. Currently I have all 1s on my car with copied ID850 data. Car runs good but hesitates sometimes in a hot start condition which makes me think i need to do something with this table, that or ID's data doesn't work for my application too well. I'm curious how you'd Infer fuel rail temp?

Yeah, I'm curious too. Maybe some equation that looks at ambient temp as a baseline fuel temp, then adds heat gain based on ECT/CHT and fuel flow rate?

I have a single column graph setup for "pulse width correction", derived from STFT avg and current PW, with rows for 0.4 - 10ms. This lets me see if a uniform pw offset correction is in order. Having access to the rail temp PID would let me add multiple columns and pull out a huge source of variability/error from this approach.

[C_Murph12]

There's no PID, it's inferred. That's what Paul's sheet does for you. Gives you the correct data obtained by testing.

[HextallS550]

I've been meaning to try something with that table. Currently I have all 1s on my car with copied ID850 data. Car runs good but hesitates sometimes in a hot start condition which makes me think i need to do something with this table, that or ID's data doesn't work for my application too well. I'm curious how you'd Infer fuel rail temp?

Put the engine a climate controlled test cell and take temps of the rail and fuel at various operating conditions and then base the inferred values off of that. That's how we do it.

[CCS86]

There's no PID, it's inferred. That's what Paul's sheet does for you. Gives you the correct data obtained by testing.

Inferred means that there is no sensor; not that there is no PID. There is a PID for rail temp, take a look for yourself. It's just broken, at least on my strategy.

Coyote's use inferred MAP too, and there is a PID for that.

That spreadsheet you sent over is for calculating the density of fuel with respect to temperature. That is a completely different phenomenon than this [ECM] 1559 - Injector Offset vs. Fuel Rail Temp table is correcting for. To correct for fuel density changes, you would apply a scalar to flow rate., which would be the [ECM] 3199 - Injector Slope vs. Fuel Rail Temp table.

The [ECM] 1559 - Injector Offset vs. Fuel Rail Temp table is a multiplier on injector offset. This means it has a very large effect on short PW fueling, and very little effect on large PWs.

[bbrooks98]

Put the engine a climate controlled test cell and take temps of the rail and fuel at various operating conditions and then base the inferred values off of that. That's how we do it.

Ok, I get how you'd calibrate it to real world conditions, but how does the ECM get its calculations to look up? How is inferred temp calculated? Time off since last start, maybe a cylinder head temp, ambient and IAT temp relationship?


I went back and looked at the car that stumbles on a hot start with ID850s. It does have the ID fuel temp data. Now those injectors are actually ID's GM style for a LS3 that i had laying around and just shortened the fuel rail spacers,changed the o rings and used the 10mm fuel hat extensions to use them. I don't think that has anything to do with it though? The data should be the same.

My other ID1000 car is actually the one with the 1's and it starts every time without hesitation. Go figure

[HextallS550]

Ok, I get how you'd calibrate it to real world conditions, but how does the ECM get its calculations to look up? How is inferred temp calculated? Time off since last start, maybe a cylinder head temp, ambient and IAT temp relationship?

That's exactly how it's calculated. During development the engine is instrumented with pressure transducers, surface temp sensors, thermocouples, pressure differential sensors, etc. We know that at 22*C ambient, 93*C ECT, 0.1 load and 1250 RPM the rail temp is measuring X. This is then worked into another formula that compensates the injector slope or offset. I don't know the specific parameters that a TC-1791 will use, however, the state of the art in the industry is similar.

[CCS86]

That's exactly how it's calculated. During development the engine is instrumented with pressure transducers, surface temp sensors, thermocouples, pressure differential sensors, etc. We know that at 22*C ambient, 93*C ECT, 0.1 load and 1250 RPM the rail temp is measuring X. This is then worked into another formula that compensates the injector slope or offset. I don't know the specific parameters that a TC-1791 will use, however, the state of the art in the industry is similar.

It's interesting to know how they might calibrate these things in house, but it's not helpful in us trying to achieve a good state of tune, because it's not practical (or necessary) to recreate.

I don't see any tables to control how the ECU calculates/infers rail temp. So, the best we can do is track the value and use it to sort fueling data to tune offset multiplier vs rail temp.

[HextallS550]

It's interesting to know how they might calibrate these things in house, but it's not helpful in us trying to achieve a good state of tune, because it's not practical (or necessary) to recreate.

I don't see any tables to control how the ECU calculates/infers rail temp. So, the best we can do is track the value and use it to sort fueling data to tune offset multiplier vs rail temp.

It's not how they might do it. It's how it's done. I work in OE engine research for one of Ford's competitors. I've never seen a Ford strategy that didn't have this logic. The table is right there under Fuel>Injector Control>Rail Temp. The ECU uses these values and applies them to the offset and slope as modifiers.
You guys in the aftermarket can do most of what we do, you just have to spend money in different areas than most are used to. Also, be ready to read patents.

[bbrooks98]

That's exactly how it's calculated. During development the engine is instrumented with pressure transducers, surface temp sensors, thermocouples, pressure differential sensors, etc. We know that at 22*C ambient, 93*C ECT, 0.1 load and 1250 RPM the rail temp is measuring X. This is then worked into another formula that compensates the injector slope or offset. I don't know the specific parameters that a TC-1791 will use, however, the state of the art in the industry is similar.

Bingo the other formula is what I'm curious about. It's verified as you said with all kinds of sensors, but the math behind it is probably an industry standard? If we know what it spits out via a logged value and have the formula to work backward with, in theory we should be able to get something in the ballpark.

Cracks me up that they infer values off of other values that are inferred and yet somehow this stuff stays within a window of accuracy. Good job to you guys.

[CCS86]

It's not how they might do it. It's how it's done. I work in OE engine research for one of Ford's competitors.

Until you state your profession, you are just a guy on the internet saying how something is done; hence the might. I'm an engineer myself, and protocols like this do change; so it's not just a given that it is done a certain way. All I was saying is that it is interesting to know, but has no impact on this discussion.

The table is right there under Fuel>Injector Control>Rail Temp. The ECU uses these values and applies them to the offset and slope as modifiers.

You do realize that the title of this thread is one of these tables. Why do you have the impression that I don't know about them. You are missing the point. We have access to these modifier tables, so we need access to a functioning rail temp channel, in order to use them. Rail temp drives the multipliers, we don't seem to have control of the rail temp derivation, so we need to log it and use to sort our data.