Tuners Rejoice! Free Tuning For M4.4!

Recommended Posts

Look at your injector pulsewidth at idle, now look at the deka spec sheet. The idle pw is well into the non linear range of the injector. You'll never get the car to idle consistently well, unless your turn off fuel trims at idle and idle significantly richer than stoich 

Link to comment
Share on other sites

I'm trying to figure that out... Can someone give a little hint, what should I put for minimum injection time? With 0.4766 constant and 0.54 TEMIN, it runs like 17.7 until lambda kicks it. If I raise TEMIN to around 0.77, it idles around 10.1-11, but overall, it runs quite well. Didn't take car out this morning for a spin. Too tired for that. 

I've set the voltage correction by copying the values from sheets, and calculated the difference for injector constant, which *should* be around 0.47...

We had this running somewhat, but seems the luck of guessing values has ran out.

Any help is appreciated! Thanks in advance!



Edit: I'm running at 3bar fuel pressure, so multipliers should be quite easy to calculate (mostly 1's).

Edited by Ondalf
Mobile seems to be horrible for attaching pictures. Added one Enter to make it somewhat readable.
Link to comment
Share on other sites

Good thing that idle isn't the required feature, as the car is more likely made to drag strip or such use. So economy isn't the issue. Just the correlation between VE map and higher loads. I'll just turn the narrowband off to ensure, it doesn't try to lean out the idle, which it can't ever succeed due long minimum pulsewidth. 

Although, first tests have been kind of success with 1.5bar and it sure revs beautiful. Now I just need to revert the settings to those, which we had 2 days ago. 

Btw, is there some reason, I cannot edit the information behind the speed scalar: http://prntscr.com/bsi9in I can change the scalar, save it, but as soon as I doubleclick it again, it shows value 0F. Basically, it doesn't change, whatever I try to put there. Other bins have 5A there, but this doesn't do anything. It doesn't allow anything to be done.

AEM UEGO allows to use the meters output in "narrowband" mode, aka Nernst cell mode, P04 mode, when it outputs 0-1V inverted just like any narrowband. Seems, the input for narrowband has resistor in it, which causes anomalies with the input. Multimeter shows 1V at idle (due rich) and it runs like --- and when letting foot off from pedal, it shows negative values.Basically the scale would be -0.2something - 0.68. Multimeter shows right values, but ECU has potential off quite well. We've tried to figure the potential error by adding signal negatives to ground, but came up with nothing and wired AEM back to tank pressure and changed scale to 0-5V aka P00 mode. It might've been cool feature, if it worked, but seems it didn't.

Actually, another way to tackle the idle STFT is to change load threshold for lambda control for lower rpms to around 0.5, or even 0.1. so it works between 1770-2580rpm up until 8 load (for example) and after the when turbo spools in, i'll probably set it to 4 or 2... Seems the suggestion using Bosch is really a viable thing. These injectors somewhat works with PW of 1.6ms (when looking the idle pw from TunerPro dashes) but if anything under it, it starts to stutter and dies due no fuel comes out. TEMIN of 0.55 still causes this. I used to use 0.24, but after trial and error, I think I've started to understand, what it really does. And how its calculated/estimated in it.

Edited by Ondalf
Question about stuck scalar and the second rambling
Link to comment
Share on other sites

Just a short update, seems I was getting a lot of false knock. As I couldn't remember exactly how I tightened my knock sensors, I went ahead and had a look - they seemed way overtightened and not in the right positions either. Now loosened them, set them to 3 and 6, and tightened with locktite to 20 Nm.

Upon my first pull I could already see the knock retard values tumbling back down quite a bit. Seems this was a major part in this.

I'm not sure what overtightening does to the quality of the piezo, so next time I go in there I might opt for new sensors. Anway, seems like a rookie mistake - I already was surprised my first build and swap wouldn't yield any problems.

Link to comment
Share on other sites

1 hour ago, Avinitlarge said:

I don't want to ask this question, I'm embarrassed to ask it.
How do you use the test adx? Ive not used it for over 2 year now I can't get it working. I thought it was load the adx and press the acquire data button then that was it.

I use it at home on the bench. Very handy for testing or debugging meters attached to the k line.

Link to comment
Share on other sites

I was with a tuner today, and when discussing my situation with him, he was able to tell me that the knock detection routine in our ECU's is tuned for the standard motor. Since I built my motor with forged internals, the sound of it would be radically different making the internal knock detection system inaccurate and unusable without re-tuning (as the added noise may trigger for knock).

That being said, does anyone know how to read the numbers in the Knock Detection Map (it says "units unknown" in the description), and how this map is used beyond a load of 8.02ms, as that's where its scale ends?

Link to comment
Share on other sites

Nonsense !!... (uuhhh..in my humble opinion, of course :biggrin:)

What sound would that be which forged conrods add that would confuse the knockroutine?

The sound of knocking is a very distinctive sound, the sound conrods make is totally different from that.


I also have forged conrods and in my experience the knocking routine still is very accurate and extremely effective.

No different from what is was when I still had standard conrods.

I don't see significant differences in ignitionvalues people can run having the same setup except for forged or standard conrods either.


B.t.w. basic knockdetection isn't done in the bin itself.  

There's a seperate dedicated chip which does the basic knockdetection, making reprogramming knockdetection extremely difficult.  


Anything above a load of 8.02 ms will use the value corresponding with a load of 8.02

You can rescale the loadaxis to get more refinement in the load regions above 8.02.

Edited by Piet
Link to comment
Share on other sites

There actually should be some maps that set up a filter for knock dection. They are setup at the factory using a special setup. It's not clear what those maps do exactly, and unless you have a way of measuring in cilinder pressure it would be very hard to tune them. 

That beïng said, I also doubt that the forged internals would change the knock frequency enough to render the knock detection routines ineffective. They might produce extra chatter, but that would have to be in the same frequency band as the knock filters are tuned to. It's possible, but I wonder how likely that is.

A trained ear can detect knock via a det can , so you could try to use that as a control measure. I suspect the stock routines on a stock engine are more sensitive than one can be with a det can though.

Link to comment
Share on other sites

@Piet, I should've been more clear that he meant my forged Wiseco pistons, not the conrods, that can alter the sound internally. As in different sound from the different metal composition and optionally more/less piston slap. Besides that I know there's some false positives as when I mishandle my Spec 3+ clutch accidentally, I can see knock detects at 1000 RPM..

I was looking into installing another knock sensor and hooking that up to an amplifier so I could listen to it myself in-car, as I'm still having knock issues where I shouldn't have them (at timings <5 deg BTDC) at this power level, even with my re-torqued sensor. Maybe the sensors are bad? I don't hear pinging in-car at all.

Edited by Boxman
Link to comment
Share on other sites

I can immagine that piston slap  could produce false knock.

Although it should be more of a problem with the engine cold then warm then.


Most knock you can't hear inside the car while driving unless severe.

I had a Gizzmo knockdetector installed for a while, which worked reasonably well.

No harm in replacing your sensors though, they are quite crucial.

Link to comment
Share on other sites

@Boxman  I would look into new knock sensors.  I don't remember what MY your car is but I believe there was an updated sensor/harness for all 850s.  Don't recall the details off hand but this is what i installed



found the bulletin on ipd's site


Edited by gdog
Link to comment
Share on other sites

  • 2 weeks later...

I've been doing a bit of playing with wideband regulation and PID tuning. I have a comparison of 6 logs from the dyno with different PID settings for each. 

RI = Reglersteigung: I (0xF1C2)

RP = Reglersteigung: P (0xF1C3)

KI = Lambda I Factor (assumed)

KP = Lambda P Factor (assumed)

KD = Lambda D Factor (assumed)

Dashed red line = STFT

Green line = AFR

Reddish brown line = target AFR

In retrospect, I probably should have taken all the extraneous data of the graph, but I didn't think of that until after it was done and I don't feel like doing it again...


This wideband control utilizes a separate map for the VE table and target AFR map. We were fighting with some sort of fuel defciency, and knock enrichment fuel (the yellow line) disables STFT.


Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.