Stay with FI, or go to Carbs?

[robert kenney]

Not to counter Guy's theory.

I'm leaning toward FI causing your idle problems for sure. The high speed could very possible need leaning to compensate for the natural loss in VE/rpm curve. All points above peak torque need a diminishing fuel curve/RPM.

You might gain insight through EGT data but I don't think you currently have thermocouples in the exhaust. If you can acquire a IR thermometer see it all cylinders agree with each other at idle. High EGT = to much throttle plate gap.

The o2 sensor location is prone to only read the two cylinders closest to it and not all 4. I would move it just down stream from the merge. The gauge reading may be seeing data from the two and then the other two or some random mix based on a cylinder going lean while surging.

Sounds like the engine is surging at idle and this is a very common problem in injected race engines that are idling lean. Assuming the TPS has idle (throttle closed) contacts. Are you on the idle contacts? if not re index the TPS. If no contacts then have you tried to tune your idle fuel curve.


Bump up your idle fuel curve some arbitrary % and see the effect. Test run it but let it quit don't manipulate the throttle to keep it running. Make another change minding the surge rpm band making changes within that band.

You are not using MAP or lambda (O2 sensor) in the idle regime. Basically you have a mechanical FI system with 0 to little control at idle.


I'm just trying to get a handle on what you have tried.
Robert

[Guy Croft]

Surely kindof makes carbs look attractive, does it not??!

G

[robert kenney]

Surely kindof makes carbs look attractive, does it not??!

G

That's for sure. Carbs for the street and road, injection for racing.

We commonly convert Hilborn stacked mechanical injected manifolds to EFI. The most critical item is having all 8 ports synchronous. At low speed and idle the 2 batch 2 channel injection won't care if one port is flowing +/- 10% but the afr's sure will. It often takes many hours to get them close enough to drive.

In all cases the system has exhaust sensors and a learning routine that gets better as you drive.

[vandor]

Hello,

Sorry for the late reply, I am having internet connection problems.
I finally had time to work on the car today. I retarded the ignition timing to 10 degrees BTDC, as I had remembered that this FI system has TPS-advance, and with the throttle closed there was a lot of advance. I changed it to zero at idle, so the only advance is from the RPM input. This slowed the idle, so I had to open the throttles a little more. The mixture problem is still there, but these changes improved the engine's running enough that now it does not stall occasionally.
I installed the airbox and drove the car a bit, most of the fuel curve will need to be richened some.

I talked with the FI system manufacturer, and he said the coils already have capacitor mounted in them. The cable going to the crank sensor is not shielded. He said that's fine as long as they are away from plugwires, and the wires and plugs are resistor type. He also said EM interference would manifest itself as an Error message on the controller.
I will see if I can change the cable to a shielded type anyways, as that seems to be a common problem.

Robert,

>The high speed could very possible need leaning to compensate for the natural loss in VE/rpm curve.

The AFR was always between 12:1 and 13:1, and on the last dyno run we did I was able to adjust AFR at WOT to 13:1 all the way from 3000 to 7000 rpm.

I do not have thermocouples, I will see if I can borrow an IR thermometer that reads high enough. BTW, my experience is that those do not get good readings from reflective (shiny) surface, I usually paint a spot flat black and get a reading from there.

We need to go to the exhaust shop anyways, so I will have them move the O2 sensor.

>Sounds like the engine is surging at idle and this is a very common problem in injected race engines that are idling lean.

According to the wideband this one usually idles rich, but then it will lean out for no reason at times. If I try to lean it so it does not idle rich, it will stall when it goes lean.

>Are you on the idle contacts? if not re index the TPS. If no contacts then have you tried to tune your idle fuel curve.

The TPS is a variable resistor, so it has no idle contacts. What determines fueling at idle is the fuel amount specified for that RPM and TPS location. Since the TPS location does not change, only the RPM-fuel value determines the amount of fuel injected.

>You are not using MAP or lambda (O2 sensor) in the idle regime

Correct.

>Basically you have a mechanical FI system with 0 to little control at idle.

Yes. I have tried turning lambda (closed-loop) on, but it did not seem to have much effect on the changing mixture.

Thank you for your input!

Csaba

[kpsig]

In my short experience, cables that are not shielded for signal of crankshaft or camshaft can become a main source of trouble for home made etc projects, inserting electrical/electronic noise and/or spikes, causing ECU algorithms to output unacceptable Fuel values, all over VE table.
You could try a shielded one, just remember to connect the shield only at the end of ECU.
The type of cable I use is relatively cheap typical shielded microphone cable with 4 wires; I pair them so that one pair carries the voltage and the other the signal.

[Guy Croft]

Csaba - once you've got the FI working as best you possibly can - that engine really ought to be developing over 170bhp @ 7000+ & around 144-146lbf ft.

If it doesn't the cams have to go I suggest. The name of the thread says it all - if I used my cams in that motor it would make 174bhp @ around 7250rpm with that torque on 44IDF and about 10% MORE power on FI. I offer that merely as a hopefully helpful benchmark at this time.

G

[mark allison]

my bet is cam issues that may be amplified by the exhaust restriction. Very interesting thread

[parrish]

I recognise all the previously mentioned problems, ive fitted an EFI/ECU sysem to my 8V TC and am having problems getting a good set up.
Im following this thread with interest.
Steve
Photos:

[vandor]

Steve,

I envy you for your ability to use sidedrafts!

Guy,

I had tried to post this a few days ago, but my internet connection disconnected...
The muffler is one of those where the entry is not in the center, but it is offset. Other than that offset, it is a straight through muffler. It's inlet OD is 62 mm, so I assume ID is ~58mm. The exhaust system's OD is 57mm, so I assume ID around 53mm.
There is a 'resonator' in front of the rear muffler. If the dyno results prove unsatisfactory, we will remove it to see if it is a restriction.
I am going to the muffler shop tomorrow to have the header downpipe redone and the oxygen sensor location changed.
Thanks,

Csaba

[Guy Croft]

Steve - nicely dressed unit but you should remote that filter and get a sealed cold air duct in there.

The red stuff looks brilliant, I do like a bit of color underhood, but you know, now now everyone is going to be asking 'where did you get those?' Esp the bits with A_B_A_R_T_H on them!!

G

[vandor]

Hello,

Well, the muffler shop was closed, even tho they said they'd be open... maybe my Spanish is not as good as I thought :-)
I'll go there on Monday.

Anyways, I looked at the tips we bought and were going to have installed, and I didn't like what I saw, see the picture. It seems that this design would disrupt airflow and be restrictive with all those pieces hanging into the airflow. Any thoughts?
Thanks,

Csaba

[mark allison]

We've found a lot of disruption in flow using resonators or mufflers with baffling such as your tips have.

[WhizzMan]

Widebands will show lean mixture when you have misfires, even tho you are mixing rich. The sensor simply measures the amount of oxygen in the exhaust gasses and you will get oxygen there if you have unburnt fuel and "air" coming out of the exhaust port into the exhaust. Can you verify with a 4-gas or other measuring technique so you can also measure unburnt fuel in the exhaust gas? That will at least give you an indication if you have a mixture problem or an ignition problem. Of course, bad mixture can cause spark problems as well, but you will at least have an explanation why your measurements are all over the place.

I was suggesting that you at least set up the EFI for the new injectors and the changed cam profile before looking at erratic idle behavior. Idle setup is an art on EFI and usually only can be done successful when the rest of the setup is dealt with, at least partially. As I mentioned above, if you have misfires, just measuring O2 to assess mixture won't work. I read that you already tried changing the ignition advance and you got some more stable idle out of that, even tho you had to open up the throttle a bit more. I would suggest you keep using that approach until you have found a stable setup, and then work your way back to better timing/mixture from that point. With EFI, if you change more than one thing at the time, you may find your setup to be completely off and you'll have to start all over from scratch.

[vandor]

Hi Whizzman,

> Widebands will show lean mixture when you have misfires, even tho you are mixing rich

Thanks, I did not think of that, but it makes sense.

> Can you verify with a 4-gas or other measuring technique so you can also measure unburnt fuel in the exhaust gas?

Hmm, I have no idea who would have that. Well, emission stations, but I doubt they'd let me use it...

>I was suggesting that you at least set up the EFI for the new injectors and the changed cam profile before looking at erratic
> idle behavior.

I will do that now that the idle is stable enough that it does not stall.
Thank you for your comments!!

Csaba

[Guy Croft]

Considering there are two exit pipes you can probably assume that even with the disruptive internal surface the cross-sectional area - compared with that of the entry pipe - is big enough not to cause back pressure. What the mufflers themselves are like is another matter and I really think you should get of that 'so-called' 'resonator'..

Where silencers ('mufflers' in the USA) are concerned function must dicate form, not the other way round which is why I exclusively recommend very 'functional' (ie: no 'frills') & over_and_over race-proven George Polley's AX891, one or two depending or requisite noise level/blown/bhp.

G