Trials and testing Will van Gemert
Turbo Mini Forums

After having received a lot of tips and advice from several people on this forum my engine is starting to get running. I thought it would be good to start a new thread for further testing now all, more or less standard problems (as far as I know them), seem to be tackled.
My setup is a bit different from the others on this forum as it is a 1380cc n.a. engine with a 286 cam and high lift rockers.

Edited by paul wiginton on 10th Apr, 2011.

Will,

This should be a "post topic" not "post poll"

Easy mistake, hopefully one of the Forum Mods will see and change it :)

Post a few more details of how you have the MS setup and your most recent logs since fixing the fan problem and take it from there.

Schrödinger's cat - so which one am I ???

Paul,
You are completly right was intended to be a topic not a poll.
Megasquirt setup:
Control algorith: speed Density ( will try alpha N later is perhaps better with my cam?)
sequential siamese injection , dual values.
Ford edis coil pack

After fixing the fan problem I Installed a new double membrane fuel pressure regulator. I hope this wil help get the pulse width up a bit as they are still to small. With the previous runs I had a fuel pressure of 4.5Bar over the injectors (calibration problem with my pressure gauge). I am now aming at 1.75 (injector minimum) at idle. The double membrane will bring fuel pressure to about 2.5 at full throttle.
had the engine running for a minute or so and it seemed to run better at idle. Rocker cover gasket popped out so have to fix that first.

Changed from poll

I seriously doubt it!

Thanks for correcting my mistake Paul.

I can not seem to get the pulse width up above 1.3 msec at idle with the 1000cc injectors they just seem to be to big. The engine idles but afr’s are almost uncontrollable at idle. With an estimated power output of 120-130 bhp I will need the 1000cc injectors at full throttle. Tomorrow I am going to revert back to the 460cc injectors. These are to small for the engine above half throttle but at least I should be able to get a good idle with these and have the engine running at part throttle and do some further testing and setting up. I was hoping I could do without staged injection but now I not so sure anymore. :(

It will be interesting to see if small injectors make a difference in your case.

I installed 680cc ones initially (just to get it running and learn the code) and then increased the size to 1000cc (for the same reason of wanting an estimated 120HP before moving to staged injection) and the idle was only very slightly worse once I had the right fuel values and timing.

The 1000cc one were more sensitive to getting the balance right between "openning time" and "req'd fuel" and I ended up with it idling with pulse widths around 1.6 - 1.7mS.

I'm using a standard fuel pressure regulator, linear rate (sense line off the plenum) set for fuel pressure to be 3bar above the plenum pressure (3 bar is the recommended setting for my injectors). I'm not quite sure what kind of regulator you are describing ?

The only other difference is the cam, I suppose mine is reasonably "mild" (I'm using a N/A MG Metro cam and standard rockers).

Schrödinger's cat - so which one am I ???

The 286 cam and high lift rockers will not idle well anyway, but you should be able to get it running as well as a carbed engine.

I would load up the VE table whilst it is running and see if it is running in a flat part of the table. You want stable pulse widths to get a good idle.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Returned to 460CC injectors.
Managed to get the engine to idle at good AFR's (for the first time!) at 1500RPM. pulse width's are around 1.6 to 1.7 msec. With a completely flat VE table below 1500 RPM and a map of 500mbar the engine goes 'off cam' very quickly below 1500 and stalls much faster than with a carb. I think it should not be completely flat I will do further testing on this tomorrow. The throttle response is much better than with the carb so that’s good news. With the carb the engine idled also stable above 1500rpm only but then it did not stall at 800 but was very lumpy between 800 and 1500.

Rod,
did you really have opening times of 1.6msec with the 1000cc injectors? Even with an injection pulse of 1 msec I still had AFR's below 12 with those injectors. The injector opening time I used with was .75msec with injector opening times of .6 msec I could not get the engine to run stable.

Schrödinger's cat - so which one am I ???

alpha n ,or throttle position, instead of map should allow it to run lower rpm gem ,your map is possibly going too low to run at areas below 1500 ?

regards robert.

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

Rod,
With the 1000cc injectors my req'd fuel setting was 1.8msec injector open time was set to 0.75msec.
Pulse width as shown in tunerstudio was around 1.05msec and AFR was around 11.5

Robert,
I am first trying to get the best settings with fuel density after that I am going to see if alha-N is the better choice with my cam. Below 1500 rpm my map indeed goes down rapidly.

Little progress after another evening of testing I managed to get the engine to idle at 1000 with AFR's around 13.5. Changed something I did not mention before. To reduce oil consumption I had the sump evacuated by manifold pressure this caused the engine to stall very suddenly below 1500 rpm. above 1500 rpm manifold pressure is almost not affected by this connection but below 1500 it all gets very instable.
I am now seeing some strange things when I tune the VE table to have the outers to run at 13.5AFR then the inners are running slightly richer. When I tune the tables to have the outers to run at 14.5 the inners are running much leaner this however also causes the engine to sound smoother. Does anybody has an idea what is causing this? I tried playing with the injection timing ( fixed for both at 30) Changed the inners to 90 and the outers to 0 but did not make any difference.
Rod, The fuel pressure regulator I use has a double membrane. with a normal regulator fuel pressure is kept constant across the plenum as you said. A double membrane causes the fuel pressure to be higher at low manifold pressure (or at boost in your case) giving a bit of extra fuel by overcompensating the manifold pressure. I have set the regulator to 3 bar at a map of- 50kpa this then results in 3.5 bar at a map of 0kpa. With 50kpa boost for example it will give a fuel pressure of 4 bar this will then cause the flow to raise sqr(4/3) =15%

I am also experiencing that the ego measurement of the inners is much more instable than the ego measurement of the outers. What could be causing this? In average it is following the outers but it is fluctuating 3 times as much. Both are of the same type ( Innovative LC-1).

Will, in the above order,

1 - Your req'd fuel = 1.8, open time = 0.75 giving a pulse width of 1.05 is quite strange.

Quoting from Paul in another thread when I was having pulse width problems (but it's somewhere in the MS manual),

PW = REQ_FUEL * VE * MAP * E + accel + Injector_open_time

and

E = gamma_Enrich = (Warmup/100) * (O2_Closed Loop/100) * (AirCorr/100) * (BaroCorr/100)

..... you must have an incredibly low MAP or E value. Next time you try, put those gauges on the TS screen and do a screenshot.

2 - That kind of regulator is what we call "rising rate" over here. I'm not a fan of them as most injectors are designed for an optimum single pressure. But in your case I can see the advantage of getting more flow at high power but it may be the low pressure at idle is not closing the injectors fast enough.

3 - How are your two sensors configured, ie what is your manifold arrangement (picture). If it is an LCB (you said it was a N/A engine) is the outer sensor in a single tube (one cylinder only). If so the inner sees twice the flow and may just be in a turbulent area.

Schrödinger's cat - so which one am I ???

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

You are going to have problems with pulse widths that low. You are very much in the non-linear area.



I know we did when we first tried to use 960cc per port.

You could always try the "hybrid" mode. It uses a single pulse for the two cylinders on each port. Set it to engage above 500rpm say. It uses VE Table 3 and Injection Timing Table 3.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

I don't know how well the single pulse of the hybrid mode will work with a 286 cam and its relatively big overlap.

Jean

http://www.jbperf.com/

Well, I've been looking at simulations using fairly high overlap and I'm of the opinion that it will be OK. Subject for another thread maybe.

Worth a try I think.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Definitely worth a try, especially if you have promising simulation data. I was just voicing a concern.

http://www.jbperf.com/

Correct Robert it gives higher fuel pressuere at higher manifold pressure

Rod,
1 - With the 1000cc injectors I had to go down with my VE to 6 and then I still had AFR's of around 12 so you are correct about the very low values. But its the pulse width that does it. Higher ve and lower req-fuel does give more normal numbers but will not change the controllability or am I seeing this wrong?

2- you've got a point there will do some additional testing with fuel pressure a bit higher when I have the 1000cc injectors back on the engine.

3- I indeed have an LCB the outer sensor is in the single tube of cyl 1 about 25cm down. the sensor for the inners is put at about the same distance from the manifold face downwards.

Paul, Jean,

Looking at my test results up to now I think its fair to conclude that sequential siamese with two 1000cc injectors does not work very well but that is what you all knew already I think. I will switch back the 1000cc injectors to see what the hybrid mode setup will give. Any clue what a good value to start with would be?
Do you have any experience with a manifold evacuated sump together with fuel injection? My PCV valve was stuck open so I will get a new one and will connect it up again to see if that solves my problem.
I don’t know if I will come to testing it today if I won't it will not be before Friday that I've got time for it.

I suggest that you leave the VE table as it is and try an injection timing of around 90 degrees at idle.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

Did two tests with the 1000cc injectors and siamese hybrid mode. First test was in speed density second one was alpha N mode. Alpha N seemed a little bit more stable in respect of idle rpm. Siamese hybrid seems to be working better in both options compared to sequential siamese (at least at idle). With the 1000CC’s idle rpm is about 300 lower and afr’s are I think reasonable Both option however seemed almost insensitive for injection timing I changed injection timing from 50 to 100 degrees and saw hardly any difference almost looks like it does nothing! Timing values however do change in megalog viewer so that would mean that the actual timing does change.
Saterday I will do some further testing with the 1000CC injectors and will take the car out for the first time to see what it does under load.
I have the MSQ file and logfile attached for the Siamese hybrid fueldensity test. At the beginning of the test injection timing is at 50 at the end its at 100 I would have expected a much bigger change in AFR’s between inner and outer.
Because I only tested idle do not look at any values outside idle conditions

Attachments:

15.01kb

14.96kb

10.01kb

84.53kb