55 replies to this topic

[sonscar]

In my experience with EFI everyone blames the sensors or ECU and almost ignores the fact that it is an elderly petrol engine with all of the maladies they have always had.If one cylinder was excessively rich and all the other lean it could go two ways,,either always add fuel or always subtract fuel as the lambda reads the average of them all.I would possibly concentrate on the lean cylinders.Steve..

[madazv8]

Hi,

 

An update on this saga. So I managed to replace the O2 sensor with the real Bosch item.  Got everything together and fired it up.  Sadly things were worse  - as per my post about seating the intake. By worse I mean -  MAP readings too high,  higher than the borderline previously.  Lambda 1000mv steady.

 

Couple of reasons for this which could be my own doing....

 

 - I didn't fully remove the intake (trying to avoid dropping coolant as I work in an apartment complex garage).

 - I didn't remove and replace the gasket (yeah yeah  I know).

 

So,  tonight I stripped down the intake properly to inspect everything, and fit a new gasket.  Here is what I found.

 

• The alignment ring from the port between 4&3 was not there. Other one was.
• The flange where the alignment ring is missing was scored up,  and seems to have a circular indentation as is somebody has tried to tighten the manifold with the ring dislodged.
• Found the missing alignment ring between the head and exhaust manifold

 

So my current theory is I have has an un-metered air leak.  but this wouldn't explain the high MAP level - or would it ?

 

Thankfully I was given a spare intake which I will swap over soon. 

 

p.s. What side does the textured face of the gasket go to? asking for a friend.

Attached Files

•  flange.jpeg   53.72K   1 downloads

Edited by madazv8, 06 March 2023 - 12:09 PM.

[sonscar]

Atmospheric pressure is high map.If the engine is sucking air after it the manifold would be at closer to atmospheric than it normally would giving rise to commanding more fuel.As I always understood it.Steve..

[genpop]

Slightly open the rings so they stay in place when you press them in their seats.Put the new gasket on so you can read the texture.When you put the manifold on have a look that there is no gap. Now put on the brackets and nuts. Everything will become good!

[madazv8]

Hi all,

 

So today I got some time and put the mini back together.  Unfortunately the problem persists which is a little frustrating, I was hopeful I was on a good path.  Seems not.  Not sure where I'm going now. Here is what I have done...

 

• Swapped over the inlet manifold for one with clean faces. New gaskets on the injection unit.
• Made sure alignment rings were in place correctly.
• New exhaust / inlet manifold gasket - metal face towards the manifolds as per guidance on minispares website.
• smear of high temp silicone on inlet manifold faces just to be sure.
• Replaced a cracked breather hose (think I split it trying to wrestle it off during strip down).
• Triple checked all other hoses and fuel trap.
• Adjusted the throttle cable as per the manual.

 

All sensors are new and working.  O2 sensor is new Bosch unit.  Symptoms are the same.  Bad/low idle speed. Lambda reading max voltage. No power.  No closed loop. Sprayed half a can of cold start on every gasket face and vacuum hose,  no obvious leak detected.

 

Connected memsfcr and the log looks the same. Attached anyway.   Tried to get memsrosco working,  but it wouldn't connect to the ECU. 

 

Observation:  If I pinch or remove the vacuum line to the ECU when the car is running,  the engine does not react in any way.  The MAP does drop down when I rev the motor,  and it does drop from atmosphere when I start the motor. 

 

Observation #2.  Stepper counts seemed very high after assembly.  They were normal prior to this.

 

Could cam timing be an issue,  but surely that doesn't explain the above?

 

Suggestions begged for, and welcome.  Thanks folks.

Attached Files

•  rename_to_dot_fcr.txt   5.87K   6 downloads

Edited by madazv8, 11 March 2023 - 12:27 PM.

[madazv8]

Clutching at straws,  but I know the car had a cam chain recently.... I sit at a steady 12 inches mercury,  which is not a good vacuum.

Attached Files

•  vacuum.png   436.61K   12 downloads

[genpop]

on the bulk head nearby the inertia switch there are two relais.One is for the electric fan, the other for the lambda heating.Check if there is coming out current to the lambda probe. There should be 12 V between the black and the slate cable at the lambda probe.If not lambda probe will not work.In your mems-rosco directory there is a engl.pdf file read that .You havent set the com port!

According tou your picture  and what you say, the timing chain could be a teeth back!

[MatthewsDad]

I'm working through a similar problem at the moment, but you're ahead of me. I'm impressed by your patience and how methodical you've been (and the advice you've received!), good luck.

[madazv8]

on the bulk head nearby the inertia switch there are two relais.One is for the electric fan, the other for the lambda heating.Check if there is coming out current to the lambda probe. There should be 12 V between the black and the slate cable at the lambda probe.If not lambda probe will not work.In your mems-rosco directory there is a engl.pdf file read that .You havent set the com port!

According tou your picture  and what you say, the timing chain could be a teeth back!

Yep I did that,  still does not connect

[genpop]

what interface do you have? Some arduino

 usb-ttl converters work with 3V, they do not work, you need a 5V one.Some nameless as well. I prefer the one in the pdf. Always publish the csv file, you may read it with the standalone memsfcr and you can easyly change it in excel for your needs..fcr files you may read only in the web based version!

[madazv8]

what interface do you have? Some arduino

 usb-ttl converters work with 3V, they do not work, you need a 5V one.Some nameless as well. I prefer the one in the pdf. Always publish the csv file, you may read it with the standalone memsfcr and you can easyly change it in excel for your needs..fcr files you may read only in the web based version!

I think I understood about none of that   Your advice is appreciated,  sadly I am not enough computer literate to understand what you are recommending.

Edited by madazv8, 13 March 2023 - 07:41 AM.

[madazv8]

on the bulk head nearby the inertia switch there are two relais.One is for the electric fan, the other for the lambda heating.Check if there is coming out current to the lambda probe. There should be 12 V between the black and the slate cable at the lambda probe.If not lambda probe will not work.
 

 

I will check that,  thanks.

 

EDIT:  12V feed confirmed.

Edited by madazv8, 08 April 2023 - 10:36 AM.

[madazv8]

Still working through this issue,  been a while as I was busy with work. 

 

Had a mate come over to sanity check what I think I am seeing. Results from todays play around are:

 

Noticed the fuel spray into the throttle body is very heavy,  the flap looks soaking wet,  is this normal ? The injector does pulse at times, but does not look like a fine mist cone.

Tried an injector from a spare throttle body,  but it is different in appearance !? even though it is same shape.  Leaked badly out the top seal so couldn't test it. According to google it is a Renault Clio SPI injector ! Is this a mod ?

Pulled the vacuum hose off the air cleaner,  and left it open,  car idles much better.  O2 sensor danced around a couple of times but then reverted to max voltage.

Squeezed fuel pipe to make car stumble,  no response from O2.

Rigged up a syringe to the MAP sensor, along with a gauge, and pulled a more correct vacuum.  ECU cuts the fuel completely and car stalls.

 

 

Images - Clio injector on the right.

Attached Files

•  mini SPI injector.jpeg   23.09K   1 downloads
•  clio injector.jpeg   15.28K   0 downloads

Edited by madazv8, 05 April 2023 - 12:01 PM.

[madazv8]

OK, so this is escalating every time...  this weekend I decided to check the low vacuum reading with another gauge just to be sure before I started digging into cam timing. I am absolutely certain there is no vacuum leak, I have checked everything, countless times.

 

Glad I did this as it seems the cheap e-bay gauge I was using has too much dampening inbuilt.  When I connected a more expensive gauge I got some new information.  Instead of a steady low reading which the e-bay gauge was indicating, the new  gauge was rapidly dancing around between 13-16 inHg.  Still low,  but an important distinction. I am inclined to believe the new gauge as it was fairly pricey and sold by a major auto store.

 

The dancing needle smooths out after I give it some RMP and the needle also behaves as it should with throttle blips.  So according to numerous online diagnostics guides,  a dancing needle like this means some issue with the valvetrain.  The movement is to rapid for it to be a single sticky valve, which then points to worn valve guides. This is cited in various websites.  Before I venture into a head off job,  would like some opinions please.  The trouble I have with this diagnosis is that I do not get any smoke at all on cold start up - but surely with  worn valve guides I would expect smoke on start up as oil seeps down?  Can the guides be worn and I don't get smoke ?  What do people think?

 

p.s. I changed the o-ring on the Renault Clio injector mentioned in the last post.  It runs exactly the same as the mini injector (stationary anyway).  And they seem to be available on e-bay from unknown sources.

 

Attached Files

•  timing.png   70.68K   0 downloads
•  valve guide.png   70.57K   0 downloads

[sonscar]

I associate fluctuating vacuum readings with burned or poorly seating valves.Might be wrong.As I said earlier EFI faults are often a sickening motor.Steve..