Megasquirt Installation Continued

TUNING
Big thanks to Stu here, for helping with the tuning, and driving the car, whilst a pig at the start, and also in a variety of ways, well timed and smoothly, so I could tune the map.
We started off with a map by Brent Picasso (www.4agte.com) which is for a 4AGZE + T25 powered AW11. It runs 440cc injectors, and a different turbo, so the map was a wee way out, but it got the car started. :)
Next up was resetting all the constants, including the all important REQ_FUEL.
Calibrate the TPS, save config to harddrive, save to ECU.
Next was tuning, we started driving it, whilst pretty rough, it was possible. One had to be careful not to invoke the TPS accell enrichments, as this could bog the car. We should have really let the O2 sensor heat up by idling first, but I was too keen to get going.
-
We initially tried tuning it on the suburban streets by jumping between the Main Page and the VE-TABLE. Not a good way to go, it's hard to remember what was what, and due to traffic, a smooth run is pretty tough.
-
Next was the motorway drive, this made for long constant throttle sessions, and by using the MegaTune's "Tuning" page, with the 3D map, we started to make great progress. basically just aiming for low voltages off the o2 at low load (low kpa <50 ), and high voltages (0.8-0.9) at high loads (high kpa> 75). After about 40 minutes, the map was pretty good, and one could ease into the gas in any gear, at any revs and not lean out. Backing off the throttle, coasting, and everywhere in between was pretty consistent with suitable o2 readings at each point.
Seat of the pants feeling plays a big part here too.
-
So.. with a pretty good VE Table, in quite a short period of time, we were pretty chuffed with MegaSquirt. Next up was setting up the TPS enrichment bins, so one could just stomp on the pedal. This was a little trickier, but unfortunately soon after beginning this the coil output insulation started packing it in, and spark was jumping through the plastic to the +ve and -ve terminals, wrecking havoc on driving and tuning.
I hope to resume playing with the TPS settings on the next outing. Thanks Stu for driving and tuning help here.
-
Once this is completed, I will upload the config file, this will be to suit a:
Stock Bluetop 4AGE, running no air filter, blocked up manifold water hoses, standard extractors, with 2" exhaust, and static 20deg ignition timing, and 310cc injectors. Good for possibly 75kW at rear wheels. Motor made 85rwKw before on dyno, with stock computer.
I don't expect this power as I'm running static timing and a tired muffler and average extractors, and 1.2mm instead of 1mm gasket.
Hopefully my turbo will be built this week and I can really start having some fun. :)
-
UPDATE - 21/1/04:
The initial map, although very good, was unfortunately set whilst the car was running on cold temp enrichment. As the 3 core radiator, combined with no thermostat would drop the temperatures to 40-50 degrees. Thus enabling the default temp enrichment. I've started retuning this, but need to have someone drive whilst I tune again. I had a brief play with TPS enrichment values to get it into the right ballpark, and even with a bad VE Table, it's not bad. I've used it as a daily driver today... and will do from now on.. so it's a good feeling. :)
-
UPDATE - 22/1/04:
I've been playing around with MegaTweak (to self tune VE TABLE) without much luck. the table it suggests doesn't seem to run too well. Most likely I've not got it configured correctly, but I don't plan on persisting with this now that my table is getting so good. O2 correction is meant to be on for this prog.
I've also been checking through the MegaSquirt Log Visual Viewer, which is very informative, and helped me spot my majorly incorrect TPS enrichment. Which was invoking 75% duty, and thus flooding the car. I guess it's pretty good for tuning anomalies. Make sure O2 correction is off datalogging for this one.
So... I'm back to using MegaTune, which is pretty fantastic, cheers Eric. -
UPDATE - 10/2/04:
Since fitting the turbo, and changing to 445cc injectors, lots has needed to be changed. I've had to fit a different o2 sensor into the downpipe, and unfortunately it isn't that good. Not showing anything over 0.75V. So, I can't really tell how rich it's running. The trick here to be safe is to listen and feel what the engine is doing, and to start rich so it splutters, and back it off until it revs and goes without hesistation. You can still count on being around 10:1, but I think this is a good compromise before jumping on the dyno. You have to kind of make a judgement call, though we have been doing this sort of thing with tricked Bluetop ECU's for a while, so are perhaps getting good at it. Lastly, if the duty cycle is 20%, and you're running at stuff all load, then it's got to be wrong, considering the standard 185cc injectors would probably be run aroun 15% or so for the same MAP reading. Again, a judgement / educated guess is required.
Also, TPS enrichments have had to be knocked right back, I think the 15v/s zone is around 3ms.
The car runs great after about 2 minutes, which doesn't seem to be due to cold start enrichment, and I don't think afterstart either, but I have to look into this.
With the turbo changing the charactistics of the car a lot, I now have changed my MAP bins to be at apex of the points on Matt's dyno graph, ie: to bracket around the big lump of power that his car has on boost. Before doing this, the car would start to lean out a bit as it was coming on song around 5000rpm.
The next outing will be to lean out some of the low down stuff to create better economy, reset base/static timing to 20 degrees, and try tune for 5000rpm+.
It is now completely driveable though, with no splutters or misses below 5000rpm, so very promising indeed. Thanks Matt for your invaluable help with driving and tuning.
-
UPDATE - 15/2/04:
Ignition finally fixed, and static again.. two wires going to 2tgeu module were reversed when turbo was being put back on.. :P ... so now. it'll rev right out...
MEGASQUIRT saved my bacon though... I didn't know how accurate it was going to be for fuelling for boost... so I started around 160 VE, but have since brought it back around 110. But I've got a crazy wastegate problem, as it's only run at 5lbs @ <4500rpm, but with the spark sorted, it's now running 8-10 to 6000rpm.. then 20lbs from there. Thankfully, the MS just ramped up the duty, eventually to 100% and it never leaned out.
So.. I've pulled the fuel back a lot, as MS is pretty accurate for fuelling for boost. Next stage is to chuck it on Robins dyno at Torque Performance and see what the go is.
I don't know whether it's over fuelling, and igniting in the turbo causing the extra boost.. or if the wastegates having problems opening....
-
UPDATE - 23/3/04:
Well.. dyno showed a massive 150kW at wheels, on roller dyno, with 11.9:1 AFR, @ 20 degrees advance and 16lbs boost.
The crazy boost is caused by either a restrictive screamer/wastegate down pipe, or by a lack of throw on the wastegate actuator. Managed to take some of the preload off to give more room for extension.
Currently, it will open from about 0.7 bar, but will creep to 1.1 bar @7500rpm. Running a spring to keep the wastegate shut, and a bleeder to stop it from opening at 5lbs or so. Really need to spend some time smoothing out downpipe me thinks.
On the tuning front. I've been messing with the NA - MAP "zone" values, so that it better covers the peaks in power. Will try to upload a new config soon. Using the Palm software, which is great! Thanks Roger. Also reflashed chip with the HI-RES Code, so I have 1micro control rather then 1 milli second control of injectors. Car is much smoother, and o2 therefore more consistent. Idle is now reading as around 1.1-1.2ms, but doesn't show resolution of microsecond which it actually is. Idle was 1.5ms!!!.
Also, been back through the warmup and acelleration bins, and car is much more driveable for it. Definetely have to keep reapproaching these as you change the VE table. I strongly recommend the HI RES code though, I don't see any reason not to run it.
Trying to find the right hall effect Bosch module so I can run Megasquirt'n'Spark.. but it will show up soon.
Also, been trying to get flames from exhaust. Perhaps not getting enough heat, but even decel fuel setting of 250% didn't give any flames. Have increased fuelling in high-rpm coast/off throttle load points, and set decel at 100%. Will have a play on the track. A decel too low feels like a total fuel cut when backing off at high load, so need to keep >60%. Otherwise off boost a low decel ~ 20% is fine.
150kW at the wheels dyno graph from Torque Performance.
-
UPDATE - 02/05/04:
Changed squirting arrangement to 2 squirts alternating. It was 4 squirts alternating. This has meant idle pulsewidths are around 1.5 instead of 1ms, and it has given me more head room for extra fuel, as I don't need to open and close the injectors 4 times per cycle, but instead hold 2 open for longer. Injector opening time is a low 0.8ms, and the injectors might even be lower then that.
Had the car back on the dyno, with a very smooth hp & torque graph, and much smoother fuelling. Unfortunately the motor is only making 134kW at the wheels now, and is breathing even harder. It's done about 18,000kms now, plus >150,000kms before I got it. Half of that on the stock ecu. So a new motor is in the works.
Also, thanks to Paint (Matt Fearon), for his Norgren valve, as it has allowed me to set the boost to hold ~ 1 bar, instead of opening at 0.5 or so and creeping up to 1.2 bar. Thanks to Tim from GD as well for modifying the new actuator which has more throw and a stiffer spring.
Anyways, i'm getting better at this tuning. As although the car is down on power, I don't feel it when driving... probably due to the greater boost at lower rpm, and smoother fuelling.
-
UPDATE - 27/06/04:
Running a redtop block now, so 10.3:1 compression. Boost still creeps, and with actuator disconnected it was still creeping up to 0.8bar. So, Lance changed the screamer pipe to 44mm tube, this helped a little, and has lowered the pitch of the screamer... I kind of preferred the smaller pipe as it sounds more ruthless, rather then torquey.
Boost is starting from 0.5-0.6 and is creeping up to 0.8 bar. Unfortunately the injectors are actually only 390cc and are unable to fuel for this boost on the high compression motor. Thought it could have been the fuel pump, so changed to another one, but only marginal difference.
Thanks to Paul Dykstra from USA for sending me over some 550cc RX7 S4 injectors. Should be plenty me thinks. And if it doesn't help, then I'll know it's the fuel pump for sure. Dyno day next Saturday, so I better get out to the garage. If fuelling works out then I'll try for 1bar on the redtop too, which I can use the norgren and valve to bleed that much boost, and back at low revs (3500-4000rpm). Will upload the last bluetop map (really good) and the latest redtop map, and the last bluetop dyno graph.

Continue on at Things to be Aware of