Megasquirt tuning

[lshark21 at pacol.net (Ed Taylor)]

Dan,
       Was this Andrew's bunny?  Sounds like his setup.  I can't wait to 
see it in action.

Ed






At 01:59 PM 11/4/2003 -0500, Dan Bubb wrote:
>Just spent Saturday tuning a 2.0L 8V Rabbit GTI with Eurospec head, 280 
>cam, TT race headers, 50mm ITB's and Megasquirt.
>
>I guess the best way to put this is, it was alot more difficult to tune 
>this fairly radical engine than it was to tune an ABA with cam and exhaust.
>
>We started off using a 486 laptop. A 486 can run the Megasquirt software, 
>but it's so slow it's hard to see where you are when you're tuning on the 
>road and everything is bouncing around. Also, the screen needs to be at 
>least 800 X 600 to see the tuning screens and an active (TFT) screen is 
>going to help alot. We switched over to my relatively new laptop after one 
>short run. You can pick up a killer laptop (for this application) on ebay 
>for less than $200. It's worth it to save the aggravation.
>
>Check for ignition trigger stability by watching the RPM display on the 
>Realtime display before attempting to drive with or tune MS. If you don't 
>have an accurate trigger MS doesn't know if it's fueling for 900 RPM or 
>5000 RPM due to spikes. Some people have been able to pull the signal 
>straight off the coil, but it appears on VW's that this will give 
>problems. If you intend to use one of the MS ignition derivatives the best 
>bet is to pick up the hall signal directly in conjunction with a pullup 
>resistor (1K ohm between the hall signal line and switched 12V). If you're 
>using stock or KS ignition you can pick up the signal off of pin 6 or 7 on 
>the ignition module. This is one of those items you need to get right 
>before you can proceed.
>
>If you're running ITBs or perhaps if you just have a really wild cam, 
>you're going to see the MAP signal jump around quite a bit even at high 
>engine speeds. It appears easy to fix this by putting a volume (old style 
>inline fuel filter) in the vacuum line to the MS. You may also need a 
>restriction in the line. A 0.031" mig welder tip has been suggested. I 
>would highly recommend adding these if you need to before starting tuning. 
>With a jumpy signal you're tuning more VE map points at once and since 
>some positions on the VE map have pretty steep gradients it makes for less 
>effective or accurate tuning.
>
>When I was tuning the ABA I felt pretty good about tuning using a narrow 
>band O2 sensor. With the wild ass 2.0 w/ITBs it seemed less effective. 
>Maybe it was that the VE map was so far off to start with. In any event 
>the WOT tuning looked good at .8-.9V, but after installing the WB-O2 it 
>looked too lean. So, here I guess that most of the tuning can be done with 
>a NB-O2, but when you get to WOT you should have a WB-O2 or plan on 
>hitting the dyno early in your MS installation if you have a fairly 
>radical engine.
>
>One other thing to look out for is if you should change to a different TB 
>or ITBs make sure your KS ignition switches are operating correctly. 
>Without the idle switch closing your base timing will be way off and the 
>ignition response coming off idle will be to retard the ignition instead 
>of advance it. This caused a really bad stumble moving off from a stop. 
>So, make sure all the other systems are operating correctly before you try 
>to tune MS.
>
>Last, once you get some tuning under your belt and are getting into the 
>WOT tuning make sure you keep an eye on your injector duty cycle. No sense 
>tuning WOT VE entries when the injectors are maxed out!
>
>Overall this particular tuning experience was only mildly satisfying 
>considering the expectations which were awfully damned high. The KS switch 
>issue (which was not even thought of until the next morning!) hurt power 
>and caused some really dramatic drivability issues!
>
>None of this dampens my enthusiasm for Megasquirt. It just points out 
>there are things that need to be taken care of before you get serious 
>about tuning.
>
>Dan
>
>
>
>_______________________________________________
>Scirocco-l mailing list
>Scirocco-l@scirocco.org
>http://neubayern.net/mailman/listinfo/scirocco-l