1st Gen Haltech ECU Discussion

The issue is that the pre-configuration using a static value and referencing anything but atmosphere can create inconsistent conditions to adjust injector output to, resulting in oscillating LTFTs. Particularly in my situation at boost reference, I can have varying boost at the turbo output vs manifold pressure depending on throttle position, and therefore varying fuel pressures at the same cell in the fuel table, and therefore different flow rates. I would assume manifold reference would worsen this. An alternate solution is to have it atmosphere referenced, where at least you're correcting to a more static condition.

My pre-configuration from Gabe was 650cc @ 3bar, set to reference manifold pressure, which on it's own does nothing to calculations since it assumes 650cc static. However, most of us are not referencing atmosphere alone (and even if we are, atmospheric pressure changes), so the fuel pressure at the rail changes, and thus so does the fuel injector flow. My proposal is to at least fill the "cc" referenced to pressure differential so at least some compensation is happening. Where one actually references pressure from there is up to them.

Like so:

Note, the x-axis values may not be what HT is expecting; I haven't tested this out yet or observed the variable in real time.

At vacuum for instance, your injectors would be closer to 522cc than 635cc, almost an error of 20%.
With boost reference we like the aspect of richer fuel mixtures, so this is more ok to overlook.

 

All it's doing is matching to a common trendline for an individual injector so engine doesn't matter. We see that as pressure rises, the change/increase in flow decreases, and vice versa. A simple linear approximation would work nearly as well.

 

I see what you're doing and it would be valid if it was a measured value, some injectors certainly have major variations at lower or higher pressures and usually higher quality injectors come with this data. The Haltech approximates this already with its fuel pressure estimation variable though and this axis you've added is only good if you actually have the real observed flow data at these greater or lesser pressures. You could certainly estimate these for no real gain and make your tune really convoluted. But you're only shooting yourself in the foot if you ever changed injectors or something since nothing would re-scale properly. I'd be more curious as to why you're having large LTFT issues currently, mine barely change by more than +/- 2% ever... is it possible you still haven't fully scaled and tuned your base fuel map properly? We have the same injectors and I have no major variations.

 

Can you point me to some documentation on this if it's different from what I noted? Wouldn't want to double up on an adjustment

 

It should be noted that the fuel pressure sensor if not a physically presented one is simulated instead by using the static fuel pressure you provided in setup, along with the slider control for whether or not you have a 1:1 rising rate referenced pressure regulator and then uses manifold pressure to estimate the effective pressure.

The data you're populating isn't a required dataset because the computer can estimate this already, it's a basic feature of Haltech's speed density modeling. If you populate this extra injector data (and it's not actual valid measured data) you're just needlessly complicating things and making it harder to adjust later.

 

Haltech has said that the IC7 update and its integration into NSP is coming later this year, around August or so. So I guess along with the UC10 release. Anyone with programming/app development knowledge want to relocate to Aussieland and help them? They're hiring software devs

 

That's quite vague tbh. I guess we would have to assume its a simple linear approximation from 630cc from the static fuel pressure rail variable. So in that case ya, I wouldn't want to double up.

Where is this slider? I don't see it in mine.

 

Sorry, it was a slider in ESP, it's a drop down in NSP

 

Oh ok, ya i have that, but still no FPR options. But that was just a curiosity.

See though, this is almost counter to what you posted:

 

Fuel output vs pulse width is different though. The ECU is changing pulsewidth based on the total flow potential of the injector because of the MAP reference. The output (injector pressure differential) would be so it can reference a different flow rate because of the injector characteristics. Again though, those settings are only useful for refinement if you have real information, not random guesswork.

 

Flip this option though and you could have some fun issues if you don't unplug your regulator... well that and you also lessen the total amount your injectors are good for.

 

Some guesswork would be more accurate than 0 change though.... we know for a fact that injector cc/min will increase with fuel pressure, and at a decaying rate is logical, which is what I'm seeing. But I agree that two injectors may vary quite a bit and having the real data for these injectors would be nice.

I'm basically thinking right now that does nothing since it's creating the "Injector Pressure Differential" virtual sensor and no table is mapping with it. Effectively it's operating as 'Constant' internally (since there's nothing telling the ECU what the pressure change means for injection) whereas the rail is seeing a pressure increase which the ECU is compensating for by pulling fuel trims. That's the inconsistency I'm tackling.

 

It shouldn't have to pull anything there because you should have mapped it properly. Sounds like you really just don't have a properly mapped base fuel table.

 

Ideally the fuel table would be hit in the same cells at a given fuel rail pressure, but in reality that's just not going to happen. It's a small detail, but getting it right will give greater control over your maps

 

I think having that data would be nice yes, but I don't think it's your issue. I, nor the others I've mapped share this problem

 

You're assuming this is causing me tangible problems - it's not. This is just in pursuit of scientific refinement

 

There's a ton of that to do then haha. If I tear mine down for something soon I'll send an injector out to Mr. injector to have the characteristics mapped

 

It's worth pointing out too that even manufacturers don't have 100% of it right for every situation and climate, that is in fact what the trims are for.

 

The IC7 dash is on sale for a week. 10% off. Just got an email notification about this.

https://www.haltech.com/product/ht-067010-haltech-ic-7-colour-display-dash/

 

You should probably show them a few pictures of your setup as encouragement