tweecer

[GTJake]

Well, I did some datalogging and I figured out that my spark curve is actually right on at full load (30*, not 42* like I thought). The 42* must have been a datalog at part throttle. So, good news, I think I can tune a little more out of it by building a MAF transfer and running more spark.

I was wondering if I should play with the injector timing? lol...I have no idea where to start with this thing.

Is there anything I should download to analyze my datalogs better? I know of EEC analyer, but is there anything that is free? What about log analyst? How do you use that?

I developed a surge at start up since I made a new CAI that puts the filter in the fender. Really weird, but I think if I mapped out the MAF transfer than it would start up better? It doesn't die, just surges and finds the desired 848rpm idle. What do I need to do to fix this?

How can I prevent the computer from pulling timing when I shift (not powershifting). Do I change "spark_min_for_tip-in_retard"? Right now it's at 21.

Thanks to anyone that bothered to read this long, boring ass thread. I'm starting to think tuning is not that fun, oh well, I'm learning alot.

Thanks

Jake

 

[Rootus]

You're only running 30* total timing? Is that what Ed recommended? Just curious... I'm running 36* total timing at WOT at higher RPM's.

Dave

 

[GTJake]

lol....I know. That's why I think there is more in it. I really need to learn more about the tweecer and find out how to make some benifical tweecs.

Jake

 

[Rootus]

Are you still using a T4M0 calibration? Although it is more conservative to begin with, the J4J1 would probably make your surge a thing of the past.

Dave

 

[GTJake]

I'm using the J4J1 base. Maybe I should go to a stock J4J1 just for a little while to see if it does anything? What is it that makes the J4J1 more acceptable?

Jake

 

[swingline]

I think I need someone to translate for me...

 

[Rootus]

swingline: J4J1 is the stock EEC for 94/95 Cobras, T4M0 is the stock EEC for 94/95 GT's. Both computers are the same physically, the difference is the calibration.

Jake, I don't know what makes J4J1 friendlier -- it's been a while since I surfed the TwEECer forums regularly, but the last time I heard, nobody could really nail down why the J4J1 works so much better. Ford still knows something we don't, apparently.

If you're surging with J4J1, then I wonder if you are running rich, at least at idle. Longtubes can cause that, but most people don't seem to encounter that unless the ambient temperature is pretty cold (say 50*). I'm running an open loop at idle calibration for exactly that reason (though it's summer now, I could switch back to a CL program). Have you done any datalogging at idle to see what's happening with KAMRF?

You mentioned programming the MAF curve as well -- I think that would be an excellent idea. I found that the one supplied by Pro-M was spot on when I got my 75mm Bullet.

Dave

 

[swingline]

Thanks Dave!

 

[final5-0]

Well, I did some datalogging and I figured out that my spark curve is actually right on at full load (30*, not 42* like I thought). The 42* must have been a datalog at part throttle. So, good news, I think I can tune a little more out of it by building a MAF transfer and running more spark.

I was wondering if I should play with the injector timing? lol...I have no idea where to start with this thing.

Is there anything I should download to analyze my datalogs better? I know of EEC analyer, but is there anything that is free? What about log analyst? How do you use that?

Log Analyst works good and it is free. Not much to figure out after you dl it and load it on your laptop.

EEC Analyzer will analyze your datalog and give you the new inj timing values based off of that datalog. It can help you with inj slopes, inj breakpoints and other things which I have found makes it more simple than using Excell to dump the data into to see what is going on. I think it is worth the small cost and the one time I had a question, I got the reply a couple of hours later.

I developed a surge at start up since I made a new CAI that puts the filter in the fender. Really weird, but I think if I mapped out the MAF transfer than it would start up better? It doesn't die, just surges and finds the desired 848rpm idle. What do I need to do to fix this?

I would bet you are too rich. You can tweec the inj low slope or the first few maf points to pull back a bit of fuel.

How can I prevent the computer from pulling timing when I shift (not powershifting). Do I change "spark_min_for_tip-in_retard"? Right now it's at 21.

As I understand it, the pcm will not pull back spark any more that 21 degrees total when the skinny pedal is suddenly closed from a high load condition. It is not really just a shift retard. You can see it at work in your datalog if you put the scalar value back to the original value of 5. BTW the scalar is spark_min_for_tip-in_retard.

Thanks to anyone that bothered to read this long, boring ass thread. I'm starting to think tuning is not that fun, oh well, I'm learning alot.

Thanks

Jake

Later
Grady

 

[HazMat404]

im glad you asked all of this because ive been tryin to work on my tweecer for awhile now (actually almost gave up for a few weeks due to frustration) but this has got me back into it. i have nothin goin on tomorrow so i plan on workin on it a bit. someone gave me this thread to read over not too long ago.

http://forums.stangnet.com/showthread.php?t=319429

 

[GTJake]

How should I datalog? Should i start it right before I go WOT in second gear and then shut it down right when I'm done? I started datalogging while parked ran up to 2k in first, shifted into second and did a WOT pass, and then did a little in third and the data was really hard to analyze in log analyst. Wonder how everybody else does there datalogging?

Jake

 

[GTJake]

Here's a pic from log analyst of that pull I was describing above. As you can see it's very hard to see anything from it.

Jake

 

[Rootus]

To see tip-in retard, you'd probably want to graph a very small interval of time, and include throttle position on the graph. I agree it's tough (read: impossible) to see on that bigger graph.

Dave

 

[rockin_rick]

It doesn't seem that graph would be very helpful as setup (for what you are trying to do). With RPM on the X axis, it's telling you what the SPARK was at any given RPM in that datalog. It doesn't tell you what the load was, TP position, etc. So you have no idea of what the throttle position was at those spark values. It doesn't display the information in a chronological order.

Try putting ET on the X axis, and have SPARK and RPM and TP (in this order) on the Y. Sometimes you have to play around a bit with the order that you add the items, as Log Analyst will scale things less than optimally. When you add items, the first on the Y axis will be assigned it's units on the left side of the graph, and the second will be assigned it's units on the right side. Then the rest will be on the left. So if you put RPM first and TP second and SPARK third, you'll end up with SPARK barely showing in the graph because it scaled with the RPMs. I usually try to put AFR first (I have a WB datalogging into the EVP) , and I try to put RPM's second since it's range is so much higher than everything else you usually look at.

I've also just started tweecing, and I try to datalog 100% of the time. That way I can go back and look at the data if something 'happens' while driving. You can always just not look at the data if you don't want to later, but if you don't datalog, then you don't have the option.

Even if you are not buying EEC Analyzer, download it. It has very informative help files. It will explain the theory behind injector timing.

I believe that tip_in_retard is activated when you step on the throttle. It just so happens that is usually when upshifting. I believe that it retards anytime you 'step on it'. I'd imagine that there is a minimum "delta volts/delta time" (rate of change in how quickly you push the throttle) 'breakpoint' that causes the EEC to check and use the tip_in_retard scalar. Anyone know what that breakpoint is?

Rick

 

[tmoss]

Does this help at all?.............

WOT TPS Breakpoint voltage - Throttle position sensor (the sensor on the throttle body) measuring how far open the throttle is - when to use WOT functions and Spark Adder for WOT - adds or subtracts spark to all WOT functions you can enter a positive and negative number. Spark Adder for WOT - This value is added to the WOT Spark Advance vs RPM function only during WOT operation.

So, total timing at WOT is WOT Spark Advance vs RPM plus the Spark Adder for WOT.

 

[krahkin]

Sorry to hijack. Tmoss check PM

 

[final5-0]

Just a few thoughts about what works for me as far as crunching the datalog to get the big picture.

A wot pull is the hardest cause things happen at a very fast pace and sometimes the speed is quite high.

Spark can be done in second gear cause you don't need as many data points. For a second gear pull I usually have calcon set up to be able to press the button to start the log. I go to wot at say 1200 to 1500 rpm and just before I mat the skinny pedal I press the button to start the log. At the desired final rpm I usually have pressed in the clutch and allowed a bit of speed to die then press the space bar to stop the log.

Working with the wideband to dial in the af ratio does not work so good with a second gear wot pull cause you don't get enough data points. Third gear works good and fourth is excellent for mid range and higher but it is kinda scary as the speed gets on up there.

A word of caution on the speed

For third gear wot pulls I usually move to fourth gear and coast a good bit after the blast to scrub off some speed (I'm at around 95 mph at the top of third) before taking eyes off the road to find the laptop spacebar.

When I get my af ratio close I will make one or two fourth gear wot pulls to about 4500 - 5K rpm for a very detailed look at all but the very last bit of the rpm range but the little woman is riding shotgun and working the laptop so I can keep eyes on the road.

You can always look at the raw data in Excell or Log Analyst and you can delete the parts of the log you dont need to help you focus in on what you are looking for.

In Log Analyst the smallest resolution on your laptop will allow you to see things in more detail.

If using EEC Analyzer to dial in inj's, after you have crunched the log, you can press the alt & print screen keys, then open paint and paste the screen shot to save it.

A readme file in the logfile folder or a folder of notes helps cause after a while you don't remember what all those log files are for.

Later
Grady

 

[GTJake]

What values should I datalog. Right now I usually do: load, %load, MAF kg/hr, MAF voltage, RPM, spark advance, throttle status, MPH.

If I make MAF transfer with the info I got from PRO-M will that change my A/F ratio?

I downloaded EEC Analyzer. Are you allowed to dump a datalog in there without pay? It worked for me and I'm wondering what kinds of options you get when you pay the $20?

Jake

 

[GTJake]

Also, is there somewhere where I could find basic definitions to what some of the terms mean in the tweecer and EEC Analyzer, like MAF and then it has really large numbers, peaking at 954. But I have no idea what the 954 means. So yea, if there is a link that I could read up on that would be great.

Jake

 

[greg@berkeley]

If I make MAF transfer with the info I got from PRO-M will that change my A/F ratio?

the maf transfer function directly affects a/f ratio, but i don't know exactly how much it will change it compared to the stock curve. when i was on the dyno, we adjusted the maf transfer function by different percentages at different rpms to richen it up more or less at different points. it worked perfectly. just remember to make sure you keep the curve smooth if you start playing with it.