ENGINE WORK  July, 2007

July 17 - some STi users are also complaining on the STi list of hard starting.  I had experienced hard starting back in April.  The last time I started it, it didn't start as easily as it had at first, but it's never started as hard as it did back in April.  I attributed it to being too rich.  The other users, Robert Paisley, and Jan Eggenfellner seem to think the problem is being too lean.  Jan and Robert recommend making a PRIMER circuit by installing a momentary switch onto some of the injector leads, to ground them out & provide a manual primer for the engine.   The idea of a manual primer on a modern fuel-injected engine is such a band-aid, is the opposite of why I bought this engine, and shows what a kluge this engine really is.

I still think the problem is "too rich", not "too lean".

It always started instantly, although I always did pre-heat the engine for an hour or so before starting it, as it was so cold outside. I started it at the end of April, after redoing the wastegate.  On the 29th, it started normally.  On the 30th, after having fiddled around running both pumps beforehand and during start, it started VERY hard and ran very rough after it started (for maybe 15 seconds). (see web site engine pages those dates for details).  I later attributed that to being too rich due to having both pumps on.  I can't remember now if it was cold enough then that I pre-heated it on April 29 and/or 30.  The last time I started it, June 21, I definitely didn't do any pre-heating and as I recall it started "slightly" hard. It wasn't the instantaneous start it originally was (both with & without supercharger connected), but it wasn't much cranking; maybe a few seconds.

July 18 - I did a couple engine test runs last night, mainly to test the possible starting problems.  The results were mostly inconclusive.  It hadn't been started since June 21. I turned on the ECU, then one pump, then immediately hit the starter, using only 1-2 second hits on the starter   0.5 hr

First hit = no fire

Second hit = fired, no start

Third hit = started, ran rough for about 5 seconds.

After running it awhile, then shutting down, I later tried a restart testing the "too rich" idea. If anything, a "too rich" situation would be exacerbated by the engine being warm. I turned on the ECM, turned on both pumps, waited about a minute, then hit the starter. I held the starter button for about 3 seconds, then the engine started OK and ran normally.  0.5 hr

If it was a "too rich" situation, we should see some black smoke, and I have never seen any.  So, I dunno - I guess it must be "too lean" , as others have said, and I wasn't seeing it during the winter engine runs because I was pre-heating my engine until the oil pan was slightly warm to the touch, because it was about 0 F outside.  When outside temps got more reasonable in late April, I quit pre-heating and then started noticing the harder starting.  It was only one time, after sitting quite awhile, that it was really HARD to start.

I was thinking last night that one of the few things that changed, besides the need to pre-heat due to ambient temperatures, between the time it always started immediately and the time it began starting harder was that I had ADDED the ECM memory line (to support the OBD port).  Maybe just coincidence and seems counter-intuitive that adding the memory line would make it harder to start, but just thought I'd mention it.  Last night, I did some starting tests - mostly inconclusive.

Randy Crothers has observed that the engine seems to start much easier when using short hits on the starter, rather than long cranking times, so I'll keep that in mind.

Also, I still have the question about the STi idle speed.  Jan said it idles at 1200, but I don't believe it.  I've asked on the Internet and was told it's in the 700 rpm range for people with the STi car.  I've also gotten emails from at least 2 other people, telling me the car idles at around 700 rpm.  When I ran the engine last night, I was showing about 620 rpm when warm (and EIS tach p/r set to 4).  One person who emailed me the idle speed of his STi car also suggested checking the idle with the EZ-Scan 4000 OBD scanner tool.  Imagine my surprise when the OBD scanner told me it was idling at 1240!  That's the value Jan said, and that's the value I'd get if I set my EIS tach p/r (pulses per revolution) to 2, which is the value Jan & others say I should use.  And 2 is the value that makes sense for me on a 4 cylinder 4-stroke engine.  But I just don't think it's idling that high, and all the STi car owners tell me their idle is in the 6-800 rpm range.  It all makes no sense.  Still trying to get it sorted out.

July 19 - Jan sent an email to the STi list with the following tips:

The primer Robert installed is the easiest way, of test to see if the bypass should be closed prior to shut down.  The primer is more universal for any kind of temperature.
Less throttle will always make the engine start easier.
Continued cranking never work since additional starting fuel is only injected for a short time.
If you ran your own starter solenoid wire, and did not attach it to the ECU, start wire, bypassing the original wire in the harness, you will have starting problems.

and of course he couldn't resist throwing in some of his brand of "improving customer relations":

For those that are waiting for STI upgrade parts right now, I had a program going 2 years ago,  We had parts out for anyone participating.  We can still get parts out, but not for a while.  Shop is far too busy completing current engines.

An H6 customer wrote me and mentioned the cooling and starting problems he's having.  I had thought all the big problems were only with the STi, but apparently the H6 has cooling and starting issues, too.

July 20 - I ran the engine again last night for a couple racer buddies from NY who stopped by on their way to the track.  They also bought my race bike, so I guess I am definitely done racing for the foreseeable future.  I didn't race last season or this season, anyway, mainly due to trying to focus on getting this plane finished.   0.5 hr

Temps were in mid 60s.  Even though it had only been 24 hours since it was last started, it took about 5-6 short (~2 sec) hits on the starter button to get it to start.  Once it did start, it ran rough for 5-10 seconds, then straightened out.  This is certainly warmer than the engine would have been last winter (even after pre-heating), when it always started immediately.  If too lean, it seems it should have started harder last winter, even with pre-heating. 

Also, both of the visitors know a lot about engines, cars, motorcycles, racing, etc.  They were quite amazed at some of the details of what a kluge this engine is.  They were especially dismayed that the supercharger drive system has no idler pulley.  They took one look at the radiators and said there's no way those 2 little radiators can cool this engine adequately.  They also really liked the concepts of swapping out the cam sprockets, grinding the supercharger housing to make it fit, no closed loop operation, no ECM warning light, etc.  They were incredulous about the "primer" circuit proposed by Jan.

July 31 - still monitoring the situation regarding marginal cooling issues and Gen 3 PSRU & SC overspeeding issues, exchanging lots of emails with other STi users.