Stay with FI, or go to Carbs?

[timinator]

Hi Csaba, sorry about going off point. Let's recap: No map,egt,knock sensor,reliable timing light,and an afr that does not respond to changes in calibration. Reminds me of tuning my first hotrod in 1970, except for the SDS system.Can we assume that the engine is back together and still running the same as before? If it is, let's try an experiment with the timing. You stated that when you reset the timing to a lesser advance that the idle became more stable. So hook up the vacuum gauge and set the timing to 10 deg. ATDC. If the vacuum picks up with this new timing you just can join the ranks with the rest of us that have said ," How the heck did I get that wrong?." Hope this effort at least gives you some new information. Good luck,Tim

[vandor]

Hi Tim,

I am sorry about the super-late reply. Free time is just hard to come by.

We have tried using a MAP sensor, where the pulse width was calculated mainly from the RPM and MAP signal. However,
the engine had such a narrow vacuum range that there were not enough points to precisely tune the mixture. A normal
engine has 10" of vacuum driving on the highway, and the software has, say 15 data points from 0 to 10" of vacuum.
Our engine had 3-4" of vacuum during normal driving, and there were only 5-6 data points between that and WOT - not
enough to control the mixture. That is why we switched to a programming that uses TPS and RPM as the main data
to calculate pulse width. Interestingly in this mode there is a 'throttle advance' that substitutes vacuum advance, where one can program in more ignition advance at smaller throttle openings.

Initially we had a problem with the MAP signal pulsing, but that was solved with a restrictor in the line. We tried to add a half liter plenum, but it made no difference, the restrictor did even out the MAP signal.

: No map,egt,knock sensor,reliable timing light

I can read the MAP signal, but it is not used in calculating the pulse width. I can use a timing light, but I need to set all the timing values to zero. The only point in using a timing light is to see if there is a difference between actual TDC, and what the FI system 'thinks' is TDC, as that is dependent on magnets in the crank pulley.

:Can we assume that the engine is back together and still running

It's not, we are having the head and manifold flowed on a flow bench. Results in 10 days or so.

Once the engine is back together I can do your timing experiment :-)
Thanks,

Csaba

[mark allison]

Have you figured out why the vacuum is so low? I'll assume you have installed the new cams, did that change the vacuum readings?

[vandor]

Mark,

actually I have not yet checked the vacuum signal with the GC 3A cams, I will do that once it is running again.
The car idled better with the new cams, so it is possible that there is more vacuum.

Csaba
Auto Ricambi, LLC 124 Spider Parts Specialists
http://www.autoricambi.us

[berjohansen]

Hi

Im sure you have thougth of this but i just cannot help myself.
I work with motorbikes. I have more than once had problems using non resistor ngk plugs (i.e B8EGV instead of BR8EGV)
This has caused all kinds of amusing problems like Winkers on a piaggio not working AT IDLE but worked fine with a bit of revs.
For some reason the winker relay is located in the ecu on this (clever?). Changed to resistor plug, problem solved.
For some reason the plug makes enough noice to mess up at idle but somehow its ok further up the rev range.
Just a thought.

[vandor]

Hi,

I don't think you can even buy non-resistor plugs here. I know a few years ago a customer was looking for some, and could not find any. So, yes, the plugs are resistor plugs :-)
Thanks,

Csaba

[vandor]

Sorry for taking this long to post an update, life gets in the way...

We have the cylinder head flowed with and without the intake manifold, and although not as good as Guy's heads, it does flow well enough.

I have measured the pressure in the exhaust system at different throttle openings, and it never goes over 1 psi. At lower openings the gauge needle vibrates badly, I assume this is because of the exhaust pulses.

As an experiment, we have installed 44IDFs on a Pierce manifold (the same manifold we used with the FI at first), and we will dyno the car with this configuration.
bye,

Csaba

[Guy Croft]

It's not suffering from back-pressure then Csaba.

Pressure-waves, sure. As for adverse wave effects upsetting the power output I would not want to second-guess that at this time.

Thanks for the update,

G

[vandor]

Last weekend we dynoed the car with the 44IDFs. The power is back! Actually the car only made a few more HP, however the power does not drop off around 6000rpm, but around 7000 rpm. It's nice to have 1000 more rpm of powerband :-)
I can't really say why the throttle bodies did not flow as well as the 44IDFs.
We experimented with jetting on the dyno, however the results did not always make sense, or were very consistent. The next day Jeff spent an afternoon experimenting with jets and got the car running very well!
bye,

Csaba

[Guy Croft]

Very interesting Csaba. Good progress. In reality one would expect to get about 10% more power from injection as there is no venturi in throttle bodies so getting the same on carbs as inj is quite telling.

Is it perhaps possible that the air velocity thru the throttle bodies wsa too low - ie: too big in fact?

That could account for the poor filling at higher speeds.

I'd like to know your jets and chokes please on 44 IDF.

Did you open out the runners on the IDF manifold? That is vital, they don't flow near enough in standard trim.

G

[vandor]

Hi Guy,

The TBs were 40mm, while the 44IDFs have 38mm chokes. Would 2mm less make that much difference?
I don't know.

We did not port the manifold, only matched the top with the 44IDFs. It's not a waffle manifold, so I do not know how it compares, but it's made for 40IDFs, so likely it is not large enough. Good point. We may gain top end power by porting it.

The final jetting ended up being 60 idles, 135 mains, 190 airs, 55 bleed backs. Interestingly the AFR meter shows rich (~12:1 AFR) even with leaner jetting, but the car still had stumbles. Now it runs extremely well, and pulls strongly to 7500 rpm. It gets there fast, I almost over revved it. I do not want to find the ignition cut at 8000 rpm!
bye,

Csaba

[Guy Croft]

38mm chokes too big. The fuel signal is poor with that barrel-venturi ratio. Change to 36mm. Get that inlet opened out because its strangling the airflow. You will inevitably find the mains need to go up and the air correctors down in size. You should not need 60 idles 45s should be plenty big enough. You definitely need to be on F9 emulsion tubes too.

Your carbs vs FI I don't know if we will ever know the true answer. My view, as I remarked yesterday to a Polish magazine (Classicauto) who are running a feature on TCs and site member Miro's 131 works replica - that I'm collaborating with:

....'If you are a National level champion sure go fuel injection and then pay another 50% of the cost of the engine for all the hardware and dyno calibration and 10% more power with a bit more driveability. Otherwise you are totally wasting your money even if it is flawlessly calibrated. Either that or you’re suffering from some weird clinical disorder that compels you to tune engines by numbers with a laptop. Take one of mine on Webers against one of yours on FI and mine will leave yours for dead. Because the money you have wasted on gismos, I will have spent on things that make my engine faster and more reliable'...

G

[vandor]

Hi Guy,

My bad, the chokes are 36mm, not 38s. Agreed on the F9s, they are on order, but we had 3 sets of F11s, and nothing else...
I bet it will be fine with smaller idle jets with the F9s, but as it was anything smaller than 57s, and it was stumbling.

Yup, amazing how simple the IDF install was compared to the FI...
bye,

Csaba

[timinator]

Hey Csaba, great to hear you are up and running. When will you be re-installing the fuel injection so we can find out which is better, carbs or the SDS fuel injection.
Tim

[robert kenney]

Hi Guy,

. Agreed on the F9s, they are on order, but we had 3 sets of F11s, and nothing else...
I bet it will be fine with smaller idle jets with the F9s, but as it was anything smaller than 57s, and it was stumbling.

Yup, amazing how simple the IDF install was compared to the FI...
bye,

Csaba

Csaba,
You have a similer issue to what I had in the beginning with the Spyder. Now 50 pilot idle screws about 1 1/2 turns out with 145 mains and 200 correctors. Before with 140 mains and 190 air correctors it had a short but pronounced dead spot
The main air corrector size has a big effect on pilot/main transition. The 200 size will improve the transition early emulsion formation when the pilot gives up and until the mains take hold.

Hope this is usefull.