1 2 3
Knurled.
Knurled. MegaDork
8/30/19 9:00 p.m.
GameboyRMH said:

Also low-speed tip-in is in general one of the hardest tuning problems to solve. Automated AFR-based fuel tuning tends not to work so well in that area of the map. Fixing it is either a guessing game or a black art depending on how much experience you have doing it...

Da (moto) Boss was annoyed that I couldn't tune a BGN to handle someone punching the throttle with their thigh from a dead stop.

 

Me, I can't even make myself not drive smoothly.  I mucked around with the accel enrich setttings until it would puke soot and it would always k-puh.

 

I did some more reading very recently (thanks to an EAE thread on rx7club.com) and there's a way to get an asynchronous acceleration enrichment pulse that SHOULD make things a lot better.  Doing two squirts/cycle instead of one should also make things easier.

Knurled.
Knurled. MegaDork
8/30/19 9:03 p.m.
Keith Tanner said:Understood, but there's no reason for the wideband to power down during cranking. This post was not a diagnosis of your specific problem but a reply to morello15's comment.

It depends on which wideband you are using.

The Innovate MTX-L will reset every time the voltage drops below 12v, and go into a warmup period in which you will get no useful data.

 

I have never seen a car that didn't drop below 12v when cranking...

Keith Tanner
Keith Tanner MegaDork
8/30/19 10:18 p.m.

Another reason to avoid Innovate widebamds, I guess. Good thing they’re cheap. 

NickD
NickD PowerDork
9/1/19 9:14 a.m.

Well, confirmed it is not anything related to the coolant temp sensor or the supercharger. Swapped in a new ECT sensor, because I had one on hand, and had no effect on driveability. Then, was driving my niece home and the blower belt shredded itself, so I had no supercharging going on, and had the same stumble. Also learned I will not be deleting power steering anytime soon. 

alfadriver
alfadriver MegaDork
9/1/19 9:51 a.m.

So once you get the calibration file sorted, can I suggest that you figure out a way to mount your laptop in the car?  Record every drive from here out until you get it perfect.  

morello159
morello159 Reader
9/1/19 5:57 p.m.

Or, if you don't feel like strapping your laptop to your car and have an android phone, a usb-to-OTG adapter with MSDroid will allow you to datalog with your phone.

Should be able to upload datalog and and MSQ (tune) files to google drive or dropbox and share them here. 

TurboFocus
TurboFocus Reader
9/2/19 2:21 a.m.
alfadriver said:

The real key is to understand what to change when:

1) rich immediately, lean after

2) rich immediately, rich after

3) lean immediately, lean after

4) lean immediately, rich after.

Those are the 4 general transient fuel situations- where the first part is an indicator that the mass component is too high or low, and the second part is the time constant- whether the compensation is too long or short.

Just to clarify because I'm self learning tuning and I wanna verify the correct actions.

Rich is high/long and lean is low/short?

High low would be changed in your MAF/MAP sensor chart and low short would be in your transient fuel section? What's changed in transient fuel?

 

I'm booting up my sct software to get the transient fuel names, standby

alfadriver
alfadriver MegaDork
9/2/19 8:11 a.m.

In reply to TurboFocus :

So in terms of how low of MAF/MAP change to deal with- that's really up to you- at some point, small changes wont cause small errors.  And the big assumption for transient fuel is that the steady state part is really good. 

For the "what's changed"- it really depends on the code- even from MS2 and MS3 there's different options for what can be adjusted for these transient events.  All of them, though, will have have some kind of mass of fuel added component and some kind of time component.   The nice thing for most of you, the requirement for making it good is just lack of stumbles- so if you are rich to, say, 12:1, it will still run great.  So the the basic adjustments should run ok.  

Not looking at other products, one really want to do it well, the MS3 code as X-Tau transient fuel- which is pretty close to what is used in the industry.  And it also includes some rather nice cold adjustments.

Anway, TF- post what your options are, and we can work through them.

Knurled.
Knurled. MegaDork
9/2/19 8:28 a.m.

In reply to alfadriver :

MS3 has several different acceleration enrichment modes.

 

It will do X-Tau.

 

It will do EAE, which models how much fuel from each injection gets lost to the runner walls, versus how much how much of the fuel clinging to the walls gets pulled off and drawn into the combusion chamber.  There are two separate curves for stuck to walls vs. drawn from walls, and there are also temperature curves and RPM curves.

 

It will also do a generic "accelerator pump shot" curve like the pump cam on a Holley.

 

It will also do a MAP-based "accelerator pump shot".

It will also do both throttle position and MAP based, with a percentage blend that you can set.

 

 

morello159
morello159 Reader
9/2/19 8:56 a.m.
Keith Tanner said:

Another reason to avoid Innovate widebamds, I guess. Good thing they’re cheap. 

Actually, my AEM does the same tihng... A bit of a doofus moment for me, but of course it's getting power while cranking (the gauge says BATT while turning over). I guess the voltage drops below whatever threshold it has, and it heats after the vehicle starts. Still, makes cranking and cold start tuning difficult.

alfadriver
alfadriver MegaDork
9/2/19 8:56 a.m.

In reply to Knurled. :

I should have noted that there were many different options.  MS2 also had different options- basic pump shot, EAE, and EAE with lag compensation.  It's really matter of which choice solves the problem well enough.  If the basic compensation isn't enough, go to the next complex one.

Paul_VR6
Paul_VR6 Dork
9/3/19 2:08 p.m.

I agree fully that one should start with the "basic" AE strategy first and move on from there. If you want an OEM level of AFR control across loads then X-Tau and EAE do work well when time is taken to tune them. If all you want is things to work well and feel good, then TPS only AE can get the job done most of the time, with the most other cases being small engines with big throttles. 

Without logs it's hard to say but usually lean on tip in will pop/stumble and rich will just feel very unresponsive. WITH a datalog with AFR much more can be done, you will have to setup fields in MLV for TPSDot if it's not a default (MS1, MS2?). 

NickD
NickD PowerDork
9/3/19 2:27 p.m.

So, I took a pretty good data log this morning (I think). Started up from cold, about 50 degrees out, even caught a stall in there. Not sure how to sure it on here so that you guys can analyze it. I did look at the AE options in MegaSquirt but they might has well have been written in Cyrillic for as much as I understood. I'm pretty sure I'm not going to want to use the TPS-based one setting, as I am still using the 1.6L throttle body which doesn't really have a TPS. 

Paul_VR6
Paul_VR6 Dork
9/3/19 2:41 p.m.

Well, we have now been transported back into the past where throttle bodies do not have sensors. 

MAP based AE is inherently laggy, and a bit harder to tune. It can work OK on stock(ish) engines, but stock block/tb with a SC should be able to be tuned.

Do you have your msq and msl hosted somewhere?

NickD
NickD PowerDork
9/3/19 2:52 p.m.
Paul_VR6 said:

Well, we have now been transported back into the past where throttle bodies do not have sensors. 

MAP based AE is inherently laggy, and a bit harder to tune. It can work OK on stock(ish) engines, but stock block/tb with a SC should be able to be tuned.

Do you have your msq and msl hosted somewhere?

Yeah, I should really do the variable TPS setup. Then I could run the bigger 1.8L throttle body (or an even larger aftermarket johnny) and likely pick up some power too. A project for this winter perhaps.

As for the second part,  how do I do that?

Stefan
Stefan MegaDork
9/3/19 3:02 p.m.

In reply to NickD :

Dropbox works pretty well for that.  Or even Google Drive, just make the files/folder shared.

NickD
NickD PowerDork
9/3/19 3:31 p.m.

https://drive.google.com/open?id=1UfWacI7TwBwDgowxXHfCjZ24zON0578g

Does that work? Man, I'm learning all sorts of E36 M3 here.

Paul_VR6
Paul_VR6 Dork
9/3/19 3:33 p.m.

Will check out a little later tonight.

NickD
NickD PowerDork
9/3/19 3:50 p.m.

The first 30 seconds or so, the AFR will be at 7.50ish because the wideband is warming up

alfadriver
alfadriver MegaDork
9/3/19 3:55 p.m.

Did you get the calibration downloaded from the car?

edit- posting the file like that works- just looking at it.  Need more data, though... laugh

NickD
NickD PowerDork
9/3/19 4:03 p.m.

In reply to alfadriver :

I only have a three mile commute, and it's literally turn left out of my driveway, drive three miles at 55mph, then turn left into work. This was also without a blower belt or power steering. And it steers like a Mack truck without the P/S, so I didn't want to drive too far anyways. Got a new belt on it today.

alfadriver
alfadriver MegaDork
9/3/19 4:04 p.m.

A quick look at the file- there's no transient fuel at all.  I'm not entirely sure how the calculations work- as the enrichment is at 100% all the time- but being flatlined like that just says it's not doing anything.  The only compensation that is doing anything is warm up enrichment.

Still, next time, can you get the a/f meter running before starting it?  

NickD
NickD PowerDork
9/5/19 12:59 p.m.

In reply to alfadriver :

Well, that would explain why it stumbles, because the ECU is trying to play catchup to what's going on, instead of getting ahead of throttle changes, correct? I was also irked to find that the tuner had set my rev limiter higher than I wanted. When I was putting this thing together, TDR warned me to stay under 7500rpm for the sake of my (stock) connecting rods. I had set my rev limit at 7200rpm to play it safe. When I went into the the computer to data log, I just happened to stumble into those settings and saw that the tuner had bumped it to 7450rpm, despite me saying I had set the limiter to 7200rpm for longevity's sake. I had never really noticed, because most of the time I went up against the limiter was at autocrosses, where I was too busy looking where I was going instead of looking at the tachometer. But that definitely made me mad.

 

Have not had time to do any more logging, because I've been busy with the other kind of logging, sawing and splitting firewood for a family friend. Not sure if I can get the wideband to kick on sooner, because it's on a cranked power circuit, and even if it wasn't, it turns off if voltage drops below 12V. I could preheat it before starting to get it on sooner, but Innovate says you really shouldn't do that, harmful to the wideband.

alfadriver
alfadriver MegaDork
9/5/19 1:13 p.m.

In reply to NickD :

The risk for turning it on early is that you may get liquid splashing onto the sensor.  While it's a risk, for sure, I've never had it break a sensor on me- that includes doing a lot of -20F starts.  

accordionfolder
accordionfolder Dork
9/5/19 1:16 p.m.

In reply to NickD :

Where is the soft limiter? I'd data log - I've got mine set to 7400 - and that takes it to 7200 with the soft limiter kicking in, and then spark/fuel cut. Previously when I had the limiter set to 7200, the car wouldn't see over 7k w/ the soft/hard limiters kicking in. I might be screwing up the exact words, but that's roughly what my settings said.

That said, I have a built motor with valve springs/rods that's waiting in the garage that I can't wait to put in and see what it does a bit higher up (without over-spinning the rotrex - I think it was good to 7800 RPM before that according to their calculator w/ my pulley).

1 2 3
Birthdays
Our Preferred Partners
tdrkUWRq2j2DIRHnMe0mgIh64kzgRsF02xY5nB2H9aYsVSkxwpMh0fDkR2Ea3V06