C. Ludwig

iTrader: (27)

STFT is simply closed loop O2 control. It's the ECU looking at the current Lambda reading and comparing that to the target Lambda table and making and instantaneous correction to try to achieve the target. STFT is not held in memory to be applied to the base map. It's simply a trim that is applied in that moment and will change constantly.

LTFT looks at a rolling average of the STFT and saves that value in the LTFT map. If you're consistently lean or rich in a given area of the map, the LTFT makes that note and stores that info in the LTFT map. This map is directly applied to the base map.

Proportional, Integral, and Derivative control is a deep subject. Motec has a couple great webinars on the subject. I'd recommend watching those before attempting to make adjustments.

Delay is a compensation for the time it takes for a particular combustion event to move from the combustion chamber to the O2 sensor. It's a compensation for the mounting position of the O2 sensor and charge speed.

Your LTFT map will only be populated by relevant STFT values and only within the parameters of the LTFT Gain map. If STFT settings are jacked up and you're not employing it properly, you may not be seeing results in the LTFT map. The LTFT map is also dependent on a steady TPS signal. If you're city driving and making constant throttle changes, you're not going to see LTFT results. Steady highway cruising is the place where LTFT really comes into play.

In short, you still go back to the age old method of proper fuel tuning. Turn the closed loop control off, put the car on the dyno, properly calibrate the base fuel map to within a couple % of the target and reduce the amount of work the ECU trims need to do to hit the target. The Haltech system of STFT and LTFT works pretty darn well but it's still not a replacement for proper dyno calibration.

stickmantijuana

iTrader: (1)

my understanding is that STFT never gets applied to the base map. Only LTFT does. and even the LTFT, you need to do it manually to get it applied permanently to the base map. It was under main menu I think.

Like c-ludwig mentioned, PID is a complicated concept. I watched the motec videos and read their PDF. It makes some sense after spending about an hour reading them, but the application is not really striaght forward. For example, the motec's PID selection numbers are limited to 0-5. Ours is 0-3000, just all arbitrary. don't know what to type in when motec people say go "low" or go "little more aggressive than proportional". anyway, I played with PID using my DBW since the response is quite immediate. when I increase P just "10" out of the available 3000, it has noticeable change. it'll act like aggressively cam'd v8 with wobbling rpm. anyway, I gave up on it. I think the consensus is that you wouldn't need it so much if you have a properly tuned car on a dyno.

Havoc

Thanks guys. I guess thats what I was looking for - more reading so I get a better understanding of the subject.

In regards to the LTFT chris, was about 20-30 mins of freeway, back streets and having fun. I kinda expected it to pick up more. But maybe the VE map is so well refined its just not finding that much to trim.

At some stage I want to change the injectors settings (which is one of the errors in my map) and the redo the VE map to suit. But I need to learn more before I try it.

Havoc

Skeese, I need you to show how to use this megalog viewer
(how do I change the units?)



But I got the car back today and got 2 small cuises in. Need to fix the lappy as keep wanting to power down while cruising.

Im still really surprised the LTFT isnt making more adjustment. And yes the injector settings ares till out, Im just trying to learn more before I start from scratch again.

Havoc

Oh one of the logs FYI

Skeese

iTrader: (11)

Are you talking about the MAP scale that is in kPA in your image? Trying to go to inHg - PSI? Or change the spacing?

Or are you talking about the overlay on the map...

Is there maybe "Lambda" and a "Target Lambda" you can overlay? It looks like you have the Target Lambda overlaid with the O2 Wideband, and I'm not sure how that outputs or reads into the logs for your specific case.

If you put target in the Z direction and the logged value in the delta Z direction, then positive values will be rich from the target and negative will be lean, in the case of fueling whether it be in lambda or AFR.

I've always used AFR instead of lambda, even for E85. Given that the number is just a visual reference for your brain to interpret and rich/lean I don't see the difference in using one or the other so long as you know where the car is mixture wise, its just a reference point.

Skeese

Havoc

sorry should have clarified.

I thought when you logged with the lappy in the car yu basically get every output.

But when I bring the Log into MLV.

the MAP is in Kpa and id prefer psi/inhg
I have no AFR's only the wideband voltages
I have no kph, just speed pulses

Obviously with is via the .cvs file. If I had of exported it as a .dl (haltech data log) if that would have worked better.

Now a new error is that when you play it back through the haltech the timing screen doesnt work

ah fun and games

Havoc

hey guys I did a quick video to try and show why I dont think ym ltft are applying. Maybe Im underestimating how much I should get but to only get 1 cell with a 0.1 change after a 45 min drive Im really surprised.

I must have a setting wrong!

stickmantijuana

iTrader: (1)

turn your LTFT correction conditions off--the 2 checkboxes that say do not do LTFT during start-up or temperature correction or something to that regard. or anything else you may have working to turn it off. another thing is to pick the same axis as your fuel base map. just keep the coolant temperature threshold condition on.

Havoc

Both are off already.

stickmantijuana

iTrader: (1)

They are both checked. Try unchecking both boxes.

Havoc

Ive done another video as I got a quick cruise in today.

Im wondering if its fuel temp correct tick box that's stopping it?

I updated my P, I and D stayed the to the default settings after I read that if you run the stock Haltech wideband thats the settings you should be using. (considerably different to those of thebase map with narrow band)

But, its still not working.

I also turned on my closed loop boost control as well. Again 0 changes tot he LTBM.

stickmantijuana

iTrader: (1)

you still have those 2 boxes checked in the video. I would start by clicking them off. If that doesn't work, try getting rid of the TPS-D and MAP axis and just set the gain as single number like 40. Start simple then add complexity. This ECU has so much integration between various settings, so it's easy to overlook a few once you start playing with it.

Havoc

yeah I know.

I actually dont know where I got the %pedal from. must have been a very early base map/ but Ill tick them off and see.

I wish my head lights were working, would be easier at night

Havoc

hmm not sure why the sound didnt work?

Havoc

so I unticked both and that's where the error must be finally started doing LT trim again!



funny now that the short term trim is set up I'm getting a few stumbles at about 6K rpm in 3gear at WOT. I might start a separate thread but keen to see what AFR map everyone is using.

Additionally close loop boost. Didnt pick up anything, stable as for me at 40% duty cycle.

Close loop idle is still one fo the best functions to the 1500 Friggin love the Drive by wire TB.

Soo ill have enough confidence to update the injector setting setting and start from scratch!

Havoc

So to cut a long story short, when my car got tuned on the dyno tunes one little feature that got missed was the injectors size data. The original base map FD settings were unchanged.

(I'm really just posting so others don't make the same mistake )

So the catch is, the VE fuel map is fine, runs great. got 408 hp or whatever it was.
But on the E85 tune we started to get some strange results as basically the injectors were out of flow. Although the VE ran alright it apparently was bloody hard to tune the staging as it was getting way more fuel than what the map thought it was getting.

this was my finished 550cc and 850cc VE map (even though I'm running 1000's and 2000's ID's)




Wrong injector settings completed VE map


So as of last night, I updated the ID injector sizes and dead times.
Did a few free runs and played with the map a bit. Got it going, small cruise and without a passenger was really just letting the LTFT find the new VE's. Now I've only done a small part of the map. But I thought it was intersting to see the vast change in the VE map with the right settings now been used.

Also small tip for new players. I never had the Auto VE air temp selected. turn on that tick box!

Jarik

Love the thread mate. Great to see another Aussie DIY tuning on an Elite.

So if you're using Auto VE temp correction, do you have your air temp correction map zeroed out?

I'm curious how Haltech do their auto VE correction model, as their air temp correction base map doesn't really reflect the ideal gas law (roughly 1% per 10 degrees is what the base map has set up).

Wonder if auto correction use that table, or does it apply the ideal gas law?

I've currently got it turned off with the table adjusted for the ideal gas law. Haven't actually tuned it. Yet though.

Havoc

Um good question and something I didn't look into. I was just going to turn off the Air temp correction and leave on the Auto VE correction.

There isnt much writen about it. but Ive now seen that the default is off. So Maybe I change it back and use the basemap Temp correction chart and tick it off.
Auto V.E. Air Temp Compensation - Official Haltech Forums

Jarik

I think theoretically, at least in other engine packages, Auto VE Air Temp Compensation should follow the ideal gas law. You would then zero out the air temp compensation table to start with and make fine tuning adjustments from there. You definitely want to make sure you're not double compensating though.

The adjustments would probably be to take into account stuff like heat soak of the sensor, etc. There is also the heating of the air that occurs once it passes the air temp sensor which could potentially be accounted for either in the Air Temp Compensation Table or in the Calculated Air Temperature Coolant Temp Bias Table.

Havoc

yeah, all good points + I expect that most Haltech 1500 are still setup in cars with stock AIT sensors. I'm running the upgraded FD fast reacting sensor.

Even though most of my setup is ceramic coated, I do see quite a bit of heat soak when stationary. Still its looks like its gone up heap's but its only 3deg in variation so sweet FA really.



I think Ill try to auto and Zero out the compensation map and see how it goes.

-Anth

stickmantijuana

iTrader: (1)

I have my air temp and barometric corrections zero'd out with VE tuning. No problems going up on local mount lemmon which is over 8000ft and driving in Phoenix desert heat at 110*F and desert morning at 50*F. This is with closed loop tuning btw.

If you have these correction maps turned on with VE, you're over-correcting which is something you may or may not want to do. I tried to say this earlier in another thread, but I don't think the person that posted a map understood what I was talking about.

as for air temp, what's getting into the combustion chamber is probably much hotter than what we measure before the throttle body as it travels through the manifolds anyway... and with fuel being added and constant changes in manifold pressure, honestly, I'm not sure if fast acting air temp is even necessary. It probably was helpful in ecu without closed loop tuning.

If you want to add more fuel at hot temp, I like just going tad richer on target AFR rather than adding % fuel in temp map. It does the same thing, except under closed loop which I am a fan.

I stored my car for the summer because it's getting unbearable to drive without a/c and exhaust heat making the car into an oven after just 30min of driving. see ya guys in november

Havoc

Mate I'm from Perth we have the same issues

Havoc

So small update got some cruises in and the new tune is going well, about to start looking at some of the boost's zones of the map.

BUT: amazing what you pick up when you think about it.

Was looking at my fuel pressure gauge and thought, Im not sure i set it that low. So found out the tunes I have have been done at a base pressure of 2.2Bar (31psi). where I use to run about 43.5 (3 bar)

****.

So question
1: in the ECU there is a setting for base pressure. what does that impact? the fuel fuel map? I could set it to 31psi and the VE map won't need to be updated?
2: If I turn the fuel pressure back up, Ill have to redo the VE map right...
3: whats everyone running for base pressure. Im running upgraded pumps and injectors I can run 43 base, not issues.


I really need to get my hand on the 12 I/O extender box then can feed all my Defi's into it... (Fuel pressure, oil temp, Exhasut Temp, oil pressure etc)

RGHTBrainDesign

I'm going to run a 55psi base pressure for my build. The pump I intend on using can handle 670lph at 90psi without breaking a sweat, so that's 90 - 55 = 35psi Boost - Pumping Losses, so likely 32psi capable fuel system (without overworking anything or running a higher voltage fuel system).

Bottom line is, if you have the pump(s), it's going to be worth the atomization. Good injectors go a long way. You'll reach a point where you have too much pressure if you can't maintain the light throttle inputs with consistency (or any low injector pulse width areas).