hi, just a thought, are the engine mounts good? If these or one of them is failing perhaps the vibration is being transfered from the engine to the chassis and making it seem an engine running problem. A wild guess on my part.
Steve PD

Maybe that is the case. Any idea how to tell if one of them is going bad?

I haven't seen a defective engine mount myself but guess that with the engine running, bonnet up, stand in front, manually vary the rpm and see if you can feel or hear any knock or unusual feeling/sound from around the engine mount area.
Perhaps take the weight of the engine on a trolley jack, handbrake on with with rear wheels chocked, try moving the body slightly in relation to the fixed held engine may expose failing engine mounts/lack of flexibility/unusual noise.

Please note I am guessing, I am not a mechanic and safety must be your first concern. I am guessing and offer my thoughts out loud only, I think other forum members may have experience of identifying/replacing failed mountings and hopefully will pitch in with advice.
Good luck, I hope the problem turns out to be a simple fix.
StevePD

Hi Steve, thanks for the advice - I will be careful. I do not recall any sings of bad engine mounts while driving last season, typically there is a kind of bang noise while releasing the clutch so it is not my prime suspect at the moment. I will check the mounts when I run out of the ideas.

By the way, I am starting to suspect an O2 sensor. I watched this video about a Honda Civic where a guy diagnosed bad oxygen senor:

The way my B. reacts to gas pedal release is disturbingly similar to this case. It was so similar to symptoms people tend to assign to idle control valve malfunction that I originally went that path and looked for the issue there, but with no success.

Also, here's another video (Rover 25 this time) which also shows a very similar warm idle noise that I get . (this was also O2 sensor issue).

I am thinking about uplugging it and see if it makes any difference. It will have to wait till the weekend though.

Thanks!

Hi,
Good detective work and research. Let us know how you get on.
StevePD

nowing that is very dificult to diagnose any trouble just based in "strange noyses" or weird simptoms even wen we have the car in our presence more dificult it is wen they are just told in a forum ... we really can just have a wild guess that even so could help , if not , just for excluding some cause ...
i just don't understand why not buy a cheap diagnose tool in ebay and download multiecuscan ( the FREE version) that identyfies imediatly any engine fault !??? being so cheap and saving so many time ...:bang:
sorry for the lecture

except when you are not even able to connect to barchetta due to their weird obd system and converters/adapters needed!

Btw anyone knows the part number/where to find Engine mounts for barchetta?
Mine are busted ( got told when the car was checked over), but not able to find where they are sold without knowing what's the proper part name/number of them.

what ever you do don't buy a diagnostic connector from China, it's just a waste of time and money. I finally learned that the hard way after buying two cables which were junk. I finally bought everything from Gendan and it worked perfectly with the Multieuscan program to turn off my seat belt light. Even used it on my buddy's 97 Alpha,he brought over from Italy, no one in Ontario, or Canada, for that matter has any software for these cars.

sorry but i can not agree with you !
all the hardware you need is in here
http://www.ebay.co.uk/itm/3-Pin-OBD...465945?hash=item43fb369059:g:JeMAAOSw44BYWRss

i both one and i allready conected it sucessfuly , i was able to see live data parameters from engine unit , read fault codes and even read a faut code from airbag unit that able me to identify the problem, solv it and erase the code reseting this way the airbag light in the dashboard ...
and i'm almost sure i'm not the only one around here !

i felt some dificultys in installing the software (drivers, comunication settings) and make it work at first try , because i'm not an informatic fan , not to say i'm really dumb in this issues !
but after an all day strugling :bang: i did it
so if i did it , any one else can . "YES, YOU can"
with alot of search in the net and this forum , there are alot of topics related in the forum ...
now i have a very usefull tool for cheap price that saves me alot of time and troubles and i advise every one to do the same .

search for it ! and if you have any question or need some help about it feel free to ask , i will help in everything i can and i'm sure that others with more experience than me will to .

Hi JMSG,

I have started with checking if there are any codes and there were none. I did not do it on the cheap interface as I had a poor experience with such tools in my other car. I had asked my mechanic to do it for me while I had him change the rear shocks. No codes whatsoever.

Note that in the videos that I linked in my previous post there were no reported codes as well.

I just wish it was as that simple as plugging in with the interface and finding the cause.

Anyways, I will get to the bottom of this, but need some time to do it. I will be reporting progress on checking the oxygen sensor.

@itaa5x

I have found a guy on yt who made his own mounts for his Toyota MR2. He just used the old housings and cut off the worn rubber. Then he poured a polyurethane into them. I think it requires certain level of DIY mastery, but maybe it's worth considering given the fact it will be harder and harder to get them from any source.

[ame]https://youtu.be/2_0BYvTkZ0o[/ame]

Cheers!

HI there MichalM !
starting to say that not being an expert in the mather but with some experience in informatic car diagnose , not just multi ecu but also vagcom opcom and scantool ; i realise something , there are specialised diagnostic software that are exclusive for a manufacturer like is the case of MES for fiat group, vagcom for volkswagen group , opcom for GM group... and there are general diagnostic software like the scantool that can diagnose every manufaturer that complys with ISO rules ! so , what is the diference ?! with the general software i diagnose every one , why to buy one for every manufacturer ?!!
the diference is that the manufacturers are oblied by law to comply with the ISO protocols but they create specific fault codes of their ones besides the general ones that a general non specific software would not read !!
and this is allways evoluting ...

Resuming . do you noow wath software your mecanic used ?
if he used a specific software and found nothing wrong , probably the problem is not in engine it self like you suspect , if he used a general purpose and the lambda sensor is not broken but just defective there is a chance that poped out another specific fault and he couldent read it !?
note that i'm just guessing
you say the videos you posted there was no fault codes , because the MIL ligth was off ??
not all the fault codes triger the MIL !
It's not easy , there are issues and malfunctions very complicated even for expirienced mecanics that work in the oficial dealer and have acess to all information and tools .

My point is that a diagnostic tool is really very usefull in our days , every mecanic has one !!
And please keep posting the evolution of your problem , i'm learning to

About the engine mounts , i have searched in " mister auto" online shop and they didn't have any for barchetta
then i tryed out searching for fiat coupe 1.8 16v with the similar engine and they have them and also for alfa 146 1.8 similar engine to , i'm not sure if they are the same but it's a chance ... you should try searching for punto hgt and lancia lybra 1.8 vfd they all share the engine so there is a chance !
maybe on EPER you can see !:idea:

Hi

My 1998 B had some idle and stalling issues when slowing down, so I did the following:

New battery - old one was over 10 years old.
Cleaned the Idle air control valve with Carb cleaner.
Took of the MAF sensor and cleaned with electric contact cleaner.

Seems to have fixed the problem and a car starts on first crank. Still idles a tad rough, but nothing too bad for an 18 years old car.

The only hitch was i lost one of the U shaped wires that pushes onto the electric connector block to stop it coming off. Does anyone know what they are called or where to source them? I'd upload a pic, but chrome keeps crashing when i try.

Ta
B

Hi Guys,

I thought I would give you an update with my case. As I was able to connect to ECU I think it is not the temperature sensor as the temperature readings on cold start and when it gets warm are reasonable. In other thread I have mentioned that I suspected lambda probe because of very low voltage it was reading when the idle reached around 950 rpms and it coincided with the moment when hectic idle showed up. However just few minutes ago I tried disconnecting the lambda and there was no difference whatsoever. There is a quite unusual heavy smell in the fumes that I thought also could be the catalytic converter getting too rich mixture. I can only tell you how much this surprised me as I was nearly betting my money on the lambda.

At the moment I am quite lost with the next steps I can take. The only remaining options I see is checking the fuel pressure or trying to swap ignition coils. I don't trust the diagnostics embedded in the ecu to catch misfire.

Any ideas?

please replace your lambda sensor. They are not expensive and easily attainable.They are the number one culprit causing rough idle in North America. it feeds information to the ECU on fuel requirements in every situation with extremely tight tolerance. We now have them fore and aft of the cat.:bang:double trouble, but cleaner air, hopefully

Thanks for your response eramosa,

My original intent was to replace it, but shouldn't unplugging it be a good test to check if lambda was the cause or not? I expected the ecu to set some kind of richer mixture and bring the idle rpm's a bit higher but make them steady. That did not happen. Or was it because the ecu needs some time adapt? (I kept it running 5 mins on warm idle total).

By unpluging it wille runing should pop out a error code open circuit for the heating element of probe !?
the signal of oxigen level would be 0V interpretated like lean mixture by ECM !!?
also oxigen sensor only works after being hot !

Okay, I spent couple hours in a garage and I still suspect the o2 sensor.

1. I did some checks on the coils and all seem to easily produce voltage for the spark to jump over ca 15 mm gap in a tester I have. I didn't see any missed sparks. Per the tester setting 15 mm is more than 20kV, so... I think I am good.

2. Then I have finally found an easy way to test if the ICV is not stuck. I noticed it is only possible to see the shaft moving when the motor was assembled with the valve and I was only make this observation when I was looking inside the valve case and through the holes. The motor seem to push the valve shaft by maybe 5 mm when 12V is applied to its pins. I kind of expected it to open more. I have also noticed that when you turn the key the ECU sends ca 4V to it, then when the motor is running cold ECU sends ca 6-7V to achieve 40-50% of ICV full movement. Anyways, I am not sure if I can fully exclude the ICV, but this makes it as a last suspect on the list.

3. Although I was not suspecting it at all I did also check the TPS. The ECU could see the throttle angle change and it was continuous and smooth change. By the way, I have seen something that made me wonder: if I was pressing the pedal to the metal I could only open the throttle by 67% (repeated couple of times) and when I manually pushed on the lever in the engine compartment I could open it by 83%! (repeated as well). Is this normal?

5. Today B. had a good day and after all these checks I connected to ECU and kept monitoring o2 voltage. Since there was barely perceivable vibrations for a moment I thought that maybe I un-stuck the ICV by connecting full 12V to it. When it got warm the o2 voltage was oscillating at 0.4-0.8V I had my fingers crossed, but now and then I saw voltage going down in rapid pikes and then vibrations occured for brief moments. I have increased RPM slightly and after it dropped down to idle lambda voltage dropped too and became 0.3-0.2 with small variations in that range. I think it is either bad or is going bad from time to time (maybe gets too cold = heaters issue). This makes the engine run unstable. What other thing could explain being normal to rich and after adding some gas and going back to idle becoming lean? I have no clue . I am thinking about one last check before I spend money on the new sensor: checking the fuel pressure. I am writing this all so that hopefully someone will benefit from it and will be less lost than I am with this issue.

EDIT:

JMSG, as always I appreciate your input. I have assumed that there is an actual difference btw the open circuit and lambda giving 0V. I have seen someone saying that the former makes the ECU assume constant fuel mixture (slightly rich) and the latter cheats the ECU into thinking the mixture is extremely lean. I had hoped I would cause the first situation and this would be my confirmation of the sensor issue. It is funny that actually ECU did not pop up any errors. It may be that I have not waited long enough with unplugged sensor for the engine to get warm and start taking its readings.

So, I realize this thread is old, but the problem is still highly present in there cars.

I was about to start i thread on this topic, but found this one. Now I´ve read it top to bottom, but is not sure of the final outcome. Seems like there are about 7 different things that would give the same problem with irratic idle and stalling.

Is there any conclusion to what was the most probable cause of it?

Sincerely, Seb

please check the fuel regulator valve at the end of the fuel rail. If its leaking than that's your number one culprit