Stay with FI, or go to Carbs?

Road-race engines and ancillaries - general discussion
Post Reply
kpsig
Posts: 35
Joined: May 10th, 2008, 6:41 pm
Location: Greece

Re: Stay with FI, or go to Carbs?

Post by kpsig »

Csaba do you use acceleration enrichment ? If yes, I strongly suggest you turn it off.
Kostas, Greece
Guy Croft
Site Admin
Posts: 5039
Joined: June 18th, 2006, 9:31 am
Location: Bedford, UK
Contact:

Re: Stay with FI, or go to Carbs?

Post by Guy Croft »

Here is the cam map with Ted's Alquati 87 profile overlaid. This tallies with what I know about the various Pittatore/Alquati profiles, a bit more area under the curve than the 77. Note the closing ramp that Ted has carefully mapped by 1 deg intervals.

Forget the effect of running clearance, I am only interested in establishing significant differences betw Csaba's cam and the others.

GC
Attachments
Lift (inches) v degrees of cam rotation for three profiles, Csaba's being 'CV'
Lift (inches) v degrees of cam rotation for three profiles, Csaba's being 'CV'
CV_GC_Ted.JPG (47.12 KiB) Viewed 8091 times
vandor
Posts: 108
Joined: November 30th, 2006, 3:24 pm
Location: Texas, USA
Contact:

Re: Stay with FI, or go to Carbs?

Post by vandor »

Robert,

The reason I want to move away from this system, is that it does not use a table, but simply a set of values for
rpm and another set for TPS. That gives a lot less adjustability than a table. I am hoping that with more adjustability the engine will run better (better AFR control under different conditions).

The airflow through the TBs is nearly perfectly matched across all 4 holes. I checked it 6 times.

The TBs have small vacuum ports, those are all ported together as 4-2-1 for the booster vacuum. The hose is very
small ID, 2-3mm.

I do not have an IR thermometer that would read that high, but I will ask around to see if I can borrow one, thanks for the tip.

Idle speed is ~1200 rpm, so I do not want to increase it any more.

>If it still will not behave make sure the system is realy open loop

Hmm, I could disconnect the sensor from the system...

What do you mean by feedback issues? I can disconnect the sensor. I had already tried it with IAT sensor, but there was no change.

The fuel pressure regulator is not connected to the manifold any more, as MAP is not stable at idle, so I do not want that changing the fuel pressure/AFR. Pressure is now at 40psi, I could lower it, but it would throw off our whole fuel map...

>...the surge seems to corelate to the injector duty cycle swinging between 4% and 5%. That may be huge at idle.

I thought of that, but I hope this system can control duty cycle much better, ie maybe it went from 4.5% to 4.6%, so
if you round it goes from 4 to 5. If indeed it can only do 4.00% or 5.00%, then there is no hope of having the correct
mixture at idle, don't you think?

>Is there any way to alter the injector gain?

I am not sure what you mean by injector gain.

>What is causing the duty % to swing? If the TPS and rpm range is constant the duty % should remain constant in open loop.

Exactly. That is my problem :-)
The ONLY think that is changing is RPM. I have tried to set all RPM values to the same number, but the mixture still changed. I will try it again, just to convince myself.

Kpsig,

We do have TPS based enrichment. I tried turning off the low-rpm one (0-1500 rpm), without any effect.
Thanks,

Csaba
GC book #288
Guy Croft
Site Admin
Posts: 5039
Joined: June 18th, 2006, 9:31 am
Location: Bedford, UK
Contact:

Re: Stay with FI, or go to Carbs?

Post by Guy Croft »

Just an observation - 1200rpm is way too high for idle, you should be 'idling' happily at 850. Ultra-light FW equipped engines might not tick over below ~1200 but is that necessary on yours?


G
robert kenney
Posts: 161
Joined: July 11th, 2007, 2:23 am
Location: La Verne Calif, USA (A)
Contact:

Re: Stay with FI, or go to Carbs?

Post by robert kenney »

vandor wrote: Idle speed is ~1200 rpm, so I do not want to increase it any more.
Never recommended an idle "RPM" increase but as you fatten the idle fuel curve the idle rpm will drop as you go rich of peak. Thus to maintain what ever rpm you chose the throttle angle will require tweaking toward more open. Leave the fuel curves at the idle rpm and a few stages above and below a bit rich. I like to see 50-75 rpm rich of peak in the beginning

vandor wrote: >If it still will not behave make sure the system is really open loop.
Hmm, I could disconnect the sensor from the system...
Only a question, not trying to be wise.


vandor wrote: What do you mean by feedback issues? I can disconnect the sensor. I had already tried it with IAT sensor, but there was no change.

Well seeing as the ECU is an electronic device with low voltage sensor inputs, they are very sensitive to variations in V0 (ground)and V+ noise. Make sure ground and power locations close to each other. Also engine bond strapping is intact.
vandor wrote: >...the surge seems to correlate to the injector duty cycle swinging between 4% and 5%. That may be huge at idle.

I thought of that, but I hope this system can control duty cycle much better, ie maybe it went from 4.5% to 4.6%, so
if you round it goes from 4 to 5. If indeed it can only do 4.00% or 5.00%, then there is no hope of having the correct
mixture at idle, don't you think?
vandor wrote: >Is there any way to alter the injector gain?
I am not sure what you mean by injector gain.
I considered as you did the significant digit displayed in injector duty as gross. Depending on injector flow ratings you may need more resolution at idle. "Gain" is the term I used. Basically a constant you can enter to shift the curve as a whole.
vandor wrote:
>What is causing the duty % to swing? If the TPS and rpm range is constant the duty % should remain constant in open loop.

Exactly. That is my problem :-)
The ONLY thing that is changing is RPM. I have tried to set all RPM values to the same number, but the mixture still changed. I will try it again, just to convince myself.
There in lies you issue and not the system as a whole. I do suspect a stray variable. Do you have access to a scope? May even suspect alternator noise and a scope can uncover that.
Robert Kenney # 111
mark allison
Posts: 38
Joined: December 21st, 2009, 4:30 am

Re: Stay with FI, or go to Carbs?

Post by mark allison »

I'd be very interested in seeing an overlay from a 4 channel lab scope of injector pulse rate, ign timing, afr and map
kpsig
Posts: 35
Joined: May 10th, 2008, 6:41 pm
Location: Greece

Re: Stay with FI, or go to Carbs?

Post by kpsig »

Csaba could you make sure that ecu is receiving grounds from sensors separetely and you have not joined all of them somewhere and then send one cable to the ECU.
try to eliminate ground loops as much as possible.
I had faced problems like that especially with lamda and crankshaft sensor.
I am sorry If I repeat those things but I have been through this never ending troubleshooting roadmap that only ends when you revisit one thing at a time and see if it is totally OK (cable termination, conductivity, noise, wire thickness etc).
The fact that fuel enrichment has no effect could be very strange and it could actually mean that you are running very rich and enrichment adds actually really nothing or there is something wrong with TPS (calibration, voltage etc).
Does your ECU offer fueling and ignition compensation according to inlet air temperature and/or coolant temperature?
I was more than once ready to throw away my ECUs, like you said, so what you write here are bringing to me really bad memories of troubleshooting efforts...
Kostas, Greece
Abarthnorway - Remi L
Posts: 207
Joined: June 24th, 2006, 1:39 pm
Location: Oslo, Norway
Contact:

Re: Stay with FI, or go to Carbs?

Post by Abarthnorway - Remi L »

vandor wrote:I would like to know if there is anybody who has been able to get an engine with hot cams, individual throttle bodies, and programmable injection to run well. I just talked to someone who has the exact same setup as what I am working on (Fiat twincam engine, TWM 40mm throttle bodies, hot cams, SDS Fuel Injection), and he said the injection worked very well on milder engines, but once he switched to throttle bodies and hot cams it became very difficult to make everything work. Of course the car runs 'well enough', and has good top end power, but the reason we went with FI is because we wanted driveability like a modern car. Otherwise one could just use 44 IDFs carbs... Two other people at last year's big Fiat meet had programmable injection systems that had some driveability issues.

My question is, is it really worth it? We are thinking what to do next, spend a bunch of time tuning the injection, use milder cams, or go to carbs?

Any ideas, opinions, experiences would be appreciated.
Thanks,

Csaba

Hi Csaba!


I personally know only one person that have set up his car with independent throttle bodies and FI with success.... a very skilled team making a Fiat FIRE 1242cc in an Autobianchi A112 - and where I (proud to say) had some influence regarding the head work. With carbs on the other hand loads of them, even with turbocharging.

After reading this thread from start to end, I am personally tempted to conclude that using and setting up a FI system for a tuned Fiat is a demanding task only for the very serious, skilled and determined tuner aiming for absolute max power. For everyone else the time, money and effort used setting up FI would be better spent on other issues and components. If considering bang for the buck+time & energy used - I would personally conclude that as long as Carburettors are still legal to use - stick to it. Setting up a tuned engine is hard enough in itself, adding to the confusion by introducing programmable fuel injection must be the last resort unless time, skills and budget is close to unlimited.

A set of good carbs, a handful of carb jets and selection of mixture tubes suddently seems like a very easy, economical and attractive alternative.....

But thats my opinion:-)


Best regards

Remi Lovhoiden


Image

Dellorto 45 still looking good 2011


For curiosity some rather interesting pics of the mentioned a112 in race..... (the red one)

Image


Image


Image


Image


Image
GC_45
vandor
Posts: 108
Joined: November 30th, 2006, 3:24 pm
Location: Texas, USA
Contact:

Re: Stay with FI, or go to Carbs?

Post by vandor »

Guy,

Yes, 1200 rpm is too high for idle, but if I lover it the engine dies when it goes lean. As is, it will go to ~750rpm when it goes lean.

Robert,

>...as you fatten the idle fuel curve the idle rpm will drop

It idles very rich (11:1-12:1 AFR) most of the time, then goes lean at times. I don't think it needs to be richer than that?

>Only a question, not trying to be wise.

I did not take it as trying to be wise. I am not that easily offended :-)

>Make sure ground and power locations close to each other

Not sure what you mean by that.
I probably should add an extra engine ground wire, I usually do that anyways, as the stock ones can be suspect after 30+ years.

>Depending on injector flow ratings you may need more resolution at idle.

That is why we went from 30 lb/hr injectors to 24 lb/hr, thinking it will be easier to control the mixture.
It did not seem to make much difference.

>Basically a constant you can enter to shift the curve as a whole.

This FI system has a knob that instantly alters the pulse width by as much as -50% to +50%. It's there to make tuning easier, and once that is done it is disconnected.

>Do you have access to a scope? May even suspect alternator noise and a scope can uncover that.

I am not that deep into electronics, but I do have a friend who is, and he does have an oscilloscope.
Mind you right now I am taking the head off to get ready to install GC 3A cams.

Mark,

>I'd be very interested in seeing an overlay from a 4 channel lab scope of injector pulse rate, ign timing, afr and map

Will see how dangerous we can get with that 'scope.

Kpsig,

>Csaba could you make sure that ecu is receiving grounds from sensors separetely and you have not joined all of them >somewhere and then send one cable to the ECU.

Actually there is only one or tow ground wires coming from the wiring harness, I assume it grounds all the sensors and the ECU? I followed the EFI system manufacturer's instructions, but will double check. I did the install over a year ago, so I do not remember all the details.

>I had faced problems like that especially with lamda and crankshaft sensor.

I looked into changing the wire for the crank sensor to a shielded type, but the current wire goes into a sealed sensor.
The only way to add a shielded wire is to cut the current one a few cm from the sensor and ECU, and solder on the shielded wire. Would that be worth it, even if the ends by the sensor and ECU are not shielded?

>Does your ECU offer fueling and ignition compensation according to inlet air temperature and/or coolant temperature?

Fueling compensation for both AT and CT. AT is not adjustable, CT is programmable.

Remi,

Thank you for your comments, and the pics! I love A112s! Our apartment complex' caretaker had two of them when I lived in Rome.
bye,

Csaba
GC book #288
timinator
Posts: 116
Joined: March 9th, 2011, 5:20 pm

Re: Stay with FI, or go to Carbs?

Post by timinator »

Hi Csaba, I have empathy for your experiences with F.I. tuning. It took 8 hrs. a day for a month to learn how to manipulate my AEM PEMS. If you don't mind there are a few questions I would like to ask. Are you using a distributor less ignition system adapter with your timing light? What is the volume of the plenum you constructed to connect to your map sensor? What manufacture's knock sensor are you using, and are you controlling fuel or timing with it? Are you using leaded racing fuel? If you have time to answer any of these questions it would help me understand what is causing your tuning difficulty. Also, WhizzMan was correct in saying idle afr is not of much importance at this point. Atmospheric air will partially fill your performance exhaust at idle and the wide band sensor will read it. Tim
vandor
Posts: 108
Joined: November 30th, 2006, 3:24 pm
Location: Texas, USA
Contact:

Re: Stay with FI, or go to Carbs?

Post by vandor »

Hi Tim,

Thank you for your questions.

>Are you using a distributor less ignition system adapter with your timing light?

No, the FI system manufacturer said that if the light is set to zero advance it will read correctly.

>What is the volume of the plenum you constructed to connect to your map sensor?

We are currently not using a MAP sensor. When we did use one, there was no plenum, just a restrictor.
I tried a ~0.3 l plenum and it made no difference in vacuum signal fluctuation at all. The restrictor did.

>What manufacture's knock sensor are you using,

Not using a knock sensor.

> Are you using leaded racing fuel?

No, this is a street car, we are using unleaded 'street' fuel.
Thanks,

Csaba
GC book #288
timinator
Posts: 116
Joined: March 9th, 2011, 5:20 pm

Re: Stay with FI, or go to Carbs?

Post by timinator »

Csaba, thank you for answering my questions since your spare time is hard to come by.
kpsig said about your wide band sensor to "...take them out for normal running." which I assumed meant that he had knowledge of your car that I hadn't read in your post. Bosch states that the service life of a wide band sensor is 100,000 miles with unleaded fuel. Innovate MS states sensor life with leaded fuel is 15,000 miles. That is why the question about fuel. For general background on WB sensors it is recommended to ground the WB sensor and the ECU together on the head/engine block to eliminate distortion in the signal. Also an un-powered WB sensor will be quickly damaged when exposed to hot exhaust gases.

I was sorry to read that you are not using a MAP or knock sensor. You simply cannot have enough sensor input. I consider a good knock sensor to be the most valuable data that can be had. It allows you to spend your time tuning and enjoying your car instead of endlessly repairing broken engines.

Using the TPS to control fuel on an IR manifold engine is an accepted strategy, but useless in sensing engine load. Engine load sensed by the MAP sensor is necessary to determine ignition timing. The problem with using a MAP sensor and IR manifold is that there is no volume to dissipate the pressure pulses coming from the opening /closing of the intake valve. When you have a common plenum manifold there is the volume of the space below the carb and all four runners to the intake valves to calm things down. So constructing a plenum to duplicate a common plenum manifold is required. Each runner must be plumbed with suitably sized tubing to the plenum to which the map sensor and fuel pressure regulator can be attached.

This probably is a good spot to let others add their opinions. Best regards, Tim
Guy Croft
Site Admin
Posts: 5039
Joined: June 18th, 2006, 9:31 am
Location: Bedford, UK
Contact:

Re: Stay with FI, or go to Carbs?

Post by Guy Croft »

Sorry but a knock sensor will categorically not detect the damaging knock frequencies unless tuned specifically for that engine and place accordingly.

If this regime is not followed the sensor will just not respond (even if the engine is knocking itself to bits which is of course way too late) and I know this because I tried some years ago to design an aftermarket knock sensor with the aid of a Plessey electronics expert. You just cannot mix-match the senders. Picking up the signals is the easy bit but getting the right signals is fearsomely difficult - which is why no dyno shop uses them on test.

There are several breeds of knock but irrespective, the best anti-knock protection you can get is the same old stuff I always post, 1) keep ex temps down with a good free-flowing header/muffler and power-optimised mixture control, 2) low (74 deg C) engine temp, 3) ambient inducted air, 4) careful igntion timing calibration to borderline power-loss (induced knock), 5) right octane grade.

Power goes down if insufficient or too much advance. Power goes down with over-rich or over-lean fuelling. That is the main job on the dyno. Go either way and adjust accordingly. The drop in power may be immediate but the decrease will not be dramatic. You need to find the optimum. If the engine knocks - you've held the adverse setting (usually ignition - ignition induced knock) too long and not noticed the sudden and very dramatic power drop. If you don't see that - she's not knocking.

The trouble with messing with knock sensors that were not original equipment (and even they may not work properly if the engine is modified) is that you won't know for sure if the engine is actually not knocking - or just the sensor is not working. The only way to be sure is to make it knock (which can be very difficult and whilst trying let's not confuse mere pre-ignition with knock as they are not the same although pre-ignition can lead to knock and pre-ign is not that damaging whereas critical knock can punch holes in things).


GC
timinator
Posts: 116
Joined: March 9th, 2011, 5:20 pm

Re: Stay with FI, or go to Carbs?

Post by timinator »

Hi Mr. Croft, Let me thank you for allowing me to become a member of your forum. I have enjoyed reading the valuable information you have assembled at great effort to help educate the public about engine development. What you said about your experiences with knock sensors is undebatable. Excuse me for being vague as to what I was trying to describe to Csaba. I was indeed referring to pre-ignition, or more accurately increased voltage signal from the 2nd generation Mitsubishi knock sensor that I use. Since we do not have your experience or resources we must use what we have. Having had a failure involving detonation I spent many hours inspecting my data logs for clues as to what preceded the event. I noticed a correlation between a rapidly increasing knock sensor voltage and the onset of detonation. Accordingly, I set a maximum voltage limit well below that value as a threshold for a 15 deg. retard of ignition timing with a 500 crankshaft revolutions time limit before allowing a slow return to desired timing. I have not had a subsequent failure. It could be just good luck rather than a valid strategy, but it seems to work. Again I stress that I don't disagree with you in any fashion. Simply trying to help Csaba develop his own strategy for engine tuning by relating my experiences albeit with less than precise language.

Best regards,Tim
Last edited by timinator on March 23rd, 2011, 1:42 am, edited 3 times in total.
Guy Croft
Site Admin
Posts: 5039
Joined: June 18th, 2006, 9:31 am
Location: Bedford, UK
Contact:

Re: Stay with FI, or go to Carbs?

Post by Guy Croft »

Very good Tim,

can't argue with your experience in that regard and thanks for describing it in constructive detail. I am sure it will help Csaba and others.

Glad you like the forum, always nice to hear that said,

G
Post Reply

Who is online

Users browsing this forum: No registered users and 32 guests