ww wrote:4cylndrfury wrote: KA or SR 240???I would never put a 4 cylinder in a 240Z, not even an SR20DET. It's just wrong.This is an L28 block w/N47 head running the original Dual SU manifold converted to dual 55mm ITB's with 2 750cc injectors and a basic MSA 6 into 1 header.
I have pictures of it done.. and it put MORE power to the ground than the modded l28 and handled better because it was a true front-mid design with the sr engine behind the front wheels
I think I found the source of some of the stumble, which turned into lots of stumble everywhere, which turned into undriveable: The GSXR's TPS decided to have a poo. "3500rpm at 65kPa with a closed/open/moving/stationary/staccato throttle? Ok - here's your fueling, good luck with that....."
Now to see if I can find another quickly.... Daily driver's body work is delayed another week, and I'm back to work on Monday.
I'd post video, but you wouldn't be able to hear the ITB's over my swearing at the car.
G
Latest maps (still in progress, eh?! Made some more refinements to SA and VE/AFR bins, and still (perpetually) tuning). Air pressure here is 98kPa, so I really haven't been right at the top of the table much. The Acceleration control has been the most annoying to tune.
I found the DIYAutoTune.com MAP Daddy dual MAP sensor helped some with the ITBs. I also put one on the Sportster because it is smaller than a regular one and we do a lot of up and down mountains with it. It still didn't completely solve my weather change issues on the 20v, but it helped. I think there is (or was) some question as to if the code was worked out right or not for dual MAP.
I find it better/easier to tune alpha/N and use the onboard map for realtime baro correction.
For accel use tpsdot, they're small enough not to need rpm based.
Paul_VR6 is spot on, for anything with ITB's the Alpha-N algorithm is going to be the better route, and tpsdot will be best for AE on this setup as well. It's not really about people giving up on speed density and running Alpha-N, it's really more about choosing the right tool for the job. An ITB engine generally has a massive amount of throttle body flow capacity compared to the engine size. What this means is with tiny little throttle movements you're at atmospheric pressure in the intake. So by the time you're at maybe 15% throttle it's atmo pressure in the intake, and speed density would think you're at full load, but you're not. As you continue to open those throttle plates more air will be able to flow. Alpha-N on the other hand uses the TPS to determine load, allowing you to fine tune the difference between 15% throttle and 50% throttle, even if both have a MAP reading of 95-97kpa.
Tables for an ITB vehicle look vastly different from a single throttle body vehicle. First off you'll set your VE table up with extra resolution down at the low load range. Set a few rows every 3-5% load apart, as at low load on an ITB car 3-5% is a big step. As you see the engine reaching higher KPA readings you can start to space the rows out a bit more, working on up to 10-14% apart at high load, where larger throttle movements make little difference.
OK, did a little screenshot work to illustrate, and maybe this will help someone a bit. Here are the fuel and spark maps from a Lotus7 clone (same car we had up at The Mitty) with a 2.0 Zetec crate motor and Jenvey ITB's. Stock internals/cams/everything except the intake and exhaust. And I'm not sure the exhaust is much of a benefit really.
First the ve/fuel table:
Notice the small increments in load at the low end of the load range, and the larger increments at the higher end. Also notice the big slope in VE even at low RPM, this is a fairly unique characteristic of ITB's. They just flow right from the get go.
And for reference, the spark table on this motor:
Not too much to say there, except to maybe note that modern motors don't need a ton of advance to make peak power, and guessing can be dangerous. How many people might have guessed this would have needed closer to 30 degrees? It made peak power where it is and needs no more advance, any more and it would start loosing power and potentially knock.
One question on AE-- I noticed you're running MS2, are you running the latest MS2/Extra code by chance? If so, you can still run the old standard AE code, tuning an extra shot of fuel for different rates of throttle movement. With some tuning you can get that working well. But there is a better way now... A way based on tons of math and code and crap that you don't care about, all you care about is the magic. Turn on EAE, read about it at http://www.msextra.com/ms2extra/MS2-Extra_Tuning_Manual.html#EAE
There are 6 tables that are tunable, but realistically usually only 2-3 that might need very fine adjustments. If you have a wideband, fully tune your base tables first, then turn on EAE and read and follow the steps at the above link, a beautiful thing will happen ;).
Note-- it is important that your base tables are right first, you can do this (fuel at least, not ignition really) on a closed course, but a dyno will be your best friend here.
Good points, as always. I second/third/trillionth getting your base tables tuned before you go playing with accel at all, it just muddys the process.
The 'shape' of the map in 3d is a good point as well. A well tuned Alpha map will look like you folded a paper on its diagonal. The little bumps that you do see in it are just areas of high ve where happy resonance tuning is occurring.
EAE would have been perfect on the few that I have tuned on MS1 that needed to use rpm based accel. It's more a brute force method that's never right, just good enough.
One of these days, I want to remap the 20 valve with Alpha-N. But I've been running it with speed density for like 3 years now and I have that tuned to about as good as it will get that way and I hate to start over with a new tune. I got 33 MPG on the last 2 tanks, so for whatever reason, it's working right now. I'm using Alpha-N on the Sportster, primarily because the vacuum take off point on the throttle body is in the wrong place and I didn't figure that out until after it was all together. It is working quite well, though. My wife says it's like driving with an automatic transmission: Just put it in gear and go, and you don't have to shift.
I should qualify my 'for anything with ITB's use alpha-n' statement above with 'on an n/a vehicle'. On a ITB+turbo car Alpha-N is still the way to go but you need Speed Density as well for when the boost comes on, so you need one of the blended Alpha-N/Speed Density methods, which MS does support. Not that anyone here is talking about doing this, but figured I'd mention it....
How do I set up MegaSquirt to run Speed Density for low throttle, and blend into Alpha-N for high throttle? This seems to make the most sense to me.
Help me Obi-wan, you're my only hope.
Ok, I'm finding my answers at www.msefi.com
My eyes are stinging from wading through the data there. Not as user-friendly as I'd hoped. I'm sure glad this car isn't my daily driver.
I've now got MegaSquirt configured (haven't attempted running yet) for Speed-Density at low throttle and blending to Alpha-N higher up. I'm not sure whether I need to be using dual tables or what, but I'll get there.....
Helpful links that took me a while to find (including an updated/corrected ini file):
When is it time for Alpha-N or blended Alpha-N?
(to be continued)
What code are you using? Is this a MS1 or MS2?
MSII v3.0
In that case, you'd use the alpha-N to MAP blending table. Most cars on B&G code use alpha-N at low RPM (where a big change in throttle angle may not move the MAP all that much, or cam reversion can be a nuisance) and transition to speed density at high RPM. You use the alpha-N MAP table to create a "pseudo-MAP" reading as a function of throttle angle.
Played a bit with Alpha-N today.
At least until the neighbour came over to complain about noise. Again. And listing off all the incidents of noise for the past three months. Always fun. I hope my kids grow up to like really loud cars. Just to be spiteful.
But I digress.
There's a really good article on MS2 SD/Alpha blending that I can't find right now. Though the methods you need to use to get it to work are fairly advanced and manual.
... scratch that, it's here: http://77e21.info/mstuning.htm
Yeah, I found that one and have studied it and studied it for the past few weeks. Very informative.
One month later now, and I've ditched Speed Density and am now re-tuning for Alpha-N. I give up.
If I datalog a sedate drive with lots of cruise, MLV will tune everything appropriately lean for cruise. Cruise can see upwards of 93.5kPa.
If I datalog a more "spirited" drive with lots of full throttle, MLV will tune everything appropriately rich for WOT. WOT at my altitude is 94.5kPa (though my ITB's can generate over 97kPa somehow....).
So, I either run lean or pig rich - I can't have a happy medium here. So I'm ditching SD and running Alpha-N.
Right now I got it to start and idle. Nothing else. Hahahahahahaha. Now to get it to drive again. Re-tuning right from scratch again. Wheeee.
And boy howdy, those ITB's are loud. I'll get that video up once it's going again.
G
Yeah, that was what I figgered you'd eventually do. I'd change my Rolla over to Alpha-N, but I have it running OK on SD if a low pressure front doesn't come through and I don't want to have to start over from scratch.
You can get more then atmospheric just from pressure waves in the tb's.
Each of the 'peaks' in the tq curve were well over 100kpa on a 99kpa day.
Car is out of the garage again for another year of playing.
I spent quite a bit of dinking around with Alpha-N and discovered the TPS was poo. I bought a brand new TPS for the GSXR600 TB's I was running. It was also poo.
I then machined up a fancy adapter to mount a Ford TPS off a dead engine from work, only to discover it was poo.
The next three Ford TPS's from the wreckers were also poo.
The local auto parts store took six weeks to still not get me one. The local Ford dealership wanted $300 for one. Poo.
Over winter I adapted a GM sensor, which was rock solid and stable (unlike their stock value).
And coolest of the cool, the latest version of TunerStudio has an "Auto Tune" feature that you can physically see and observe improving the engine as you go. Assuming, of course, that your SA and AFR tables aren't poo.
It's running VERY nice and strong now, with little more than an afternoon of driving. Very cool.
Current tables, running pure Alpha-N on a "stock" low compression big port 4AGE with ITB's and EDIS:
Understanding, of course, that I haven't a clue what I'm doing.
Looks lean unless you have 'incorporate AFR' turned on.
I haven't posted a recent tune map anywhere in a long time, so here it is. This is my (hey googlers) Megasquirt 20 valve 4AGE Alpha-N Spark and VE Fuel Map (s) for ITB from last week. I'm still playing with them, of course, but this is a fairly good tune, I think. Couple of coming off cruise stumbles where the EGC switches over, but they are getting smaller. I can't get the mileage back up where it was with the MAP sensor based tables, but it runs a lot better on Alpha-N and it doesn't fall on it's face when a low pressure front comes through. I need to clean up the ignition map a litle bit yet at the bottom. I've been playing around with it some. Also, I think I'm leaving some power on the table with timing at the top, but I'm overly cautious there as this is my daily driver and I don't want to blow a piston.
Paul_VR6 wrote: Looks lean unless you have 'incorporate AFR' turned on.
I have "Use Combined VE/AFR table" selected in AFR Table Fuel Calc Usage.
Should I not be doing that?
You'll need to log in to post.
