[142 Guy]

Thomas:

I just had a quick look at your log files. I need to spend some more time looking at the logs; but, these two events suggest something is odd with your ignition.

At 53.3 seconds



At 133.4 seconds



The white trace on each lower graph is your base ignition advance. Engine speed and MAP are changing small amounts; but, your base ignition timing is going through rapid swings greater than 10 deg. Do you have a very sharp change in your ignition map around 900 RPM? I looked at other places on the log with similar RPMs and MAP values and I didn't see similar movement in the ignition advance (I scanned the log quickly so I could have missed things).

I don't know whether this variation in the ignition timing is what is causing your problem; but, it is something that you should check out. I also note that your TPS is reading -4% (presumably closed). I don't think that would cause your problem; but, it can create some problems with the throttle cut depending on what settings you have for throttle cut. I try to program mine to give 0 to 1% closed and around 99% fully open. There are some rounding off things that occur in MSExtra with the A to D converters so you can't get certain values in the settings.

 

[Ram99]

The big jumps (53s and 133s) of the spark advance are caused by the idle advance settings. It turns on with speed lower than 1050 Rpm and allows a regulation range between 4 and 18 deg.

The TPS signal is strange sometimes. I set to -1% the close signal and 100% full open. The close signal change sometimes to more negative when the engine is hot. I did not investigate in that issue.

Edit: I do not use a idle valve jet. The two flanges on the manifold for the air valve are closed.

 

[142 Guy]

My error on the ignition. I was thinking that I had selected SPK: spark map 1 which is the raw spark map data. Hazards of trying to think past my bed time.

I had a look at the second log you did with fixed ignition advance and the speed oscillations are still there so the speed oscillation is not being caused by the ignition timing calculation in MSExtra.

The majority of my logs in the last two years have been records of cold and hot starts as I refined my starting settings. I could only find one log with higher load engine operation and there is nothing with sustained engine operation around 2000 RPM. Any steady speed recording I had were all above 2300 RPM where you don't seem to have a problem. I did have one section where the engine speed was around 2600 RPM and then dropped - looks like a 1st to 2nd gear shift and I found a similar spot on your log with approximately the same change. The MAP and fuel PW values versus RPM are very close so I think the engines and the fuel injection are operating roughly the same in that region. Your Ve value is a lot different; but, I think that is because of your fixed fuel pressure. I can post those if you are interested; but, based upon the following I don't think they are that important.

I expanded the display scale on your data log around 153 seconds where you are getting some nasty RPM changes



Red is MAP, white is RPM, Yellow is fuel PW and green is Ve. As an approximation, if other things are not changing the calculated fuel pulse width is

PW = k (Ve)(MAP)

Where k is a constant that accounts for air temperature and all the other fuel corrections which should not be changing during that time period on the log. RPM does not affect PW calculation directly, only indirectly because Ve is a function of RPM and MAP. If you look at the log results, PW is hardly changing at all. The Ve values are very steady and the PW value is just about directly proportional to the MAP value which is what should happen. Since the fuel delivery is hardly changing (and is only following the change in MAP) I don't think your engine speed fluctuation is being cause by the fuel injection calculations. I think the "mega squirt box" is doing the correct calculations and your problem is somewhere outside of the "box".

Do you still have your fuel pressure sensor connected up and if so, can you log fuel pressure while you are running (I can't remember whether there are any unused analog inputs on the MS). Perhaps there is some strange problem with the external ignition components of the injector wiring - but that would be odd that it only occurs in that specific range of operation.

 

[Ram99]

I think the "mega squirt box" is doing the correct calculations and your problem is somewhere outside of the "box".

Do you still have your fuel pressure sensor connected up and if so, can you log fuel pressure while you are running using are running (I can't remember whether there are any unused analog inputs on the MS).

I did some more checks and in the meantime I believe the same. The problem is somewhere outside of the MS. I do not have the time to check the external parts at the moment but I will do it on next Thursday.
I will prove all the ignition parts (cables, spark plugs...). The injectors are okay. I tested the flow rate again (51,51,51,52ml).
I will also check the supply voltage with the scope during oscillating. I do not expect the problem there but I will check it anyway.
The fuel pressure is available in the log files. Search for „FuelPressure“. The value looks stable at 2,1bar (30psi).

 

[142 Guy]

I was muddling around in my trunk yesterday and I noticed that on the extractor vent outlet under the rear window the paint had been scraped off of a couple of the little bridges / grills in the outlet opening. A little more examination showed that the front edge of the trunk lid was just grazing over a couple of the bridges as the trunk lid was closed. The little bridges have a slight outward bulge to them and the amount of bulge is clearly non uniform since the trunk lid front edge was not contacting the bridges on either side of the scraped piece and the distribution of the scraped bridges was rather random across the extractor vent opening.

I did an immediate fix to the problem by positioning the trunk lid on its mounting holes so that the lid sat further back about 3 mm which eliminated the contact problem. However, the back edge of the trunk lid is no longer nice and flush with the back of the rear 1/4 panels. While doing some open and closed testing on the trunk lid, I noticed that as the trunk lid is approaching the closed position, about 50 mm from full closed there is a fair amount of fore - aft play in the hinge mechanism. As you are closing the lid there is a slight natural tendency to push a little forward on the lid and the play in the hinge mechanism is what allows the contact to occur. If I very carefully lower the hood and pull back on it ever so slightly the contact does not occur. With the lid just held barely open (latch has not engaged) I can move the lid for and aft about 3 - 4 mm quite easily. I tested the hinge pivots when the lid was fully open by pulling up, down and sideways on the lid and the and there was no obvious slop in the mechanism. Given that the lid does not get opened and closed a lot, I would be a bit surprised by significant wear.

I am curious as to whether other 142 and 144 owners have similar for - aft movement in the lid as it approaches the closed position? Given that signs of wear were not obvious, I am wondering if the kinematics of the hinge mechanism conspire to amplify the normal clearances in the pivots - there are 5 pivots in the one linkage so 5 small clearances might up adding to a lot of slop in a particular alignment. The intent of my question is that if this for aft movement in the lid is a normal feature, I am not going to pursue new replacement hinges as a possible fix.

It occurs to me that the other possible fix is to take a brass drift covered with a couple layers of leather along with a ball peen hammer and massage those high bridge points in the vent opening a bit so that they are lower. The paint already requires a little touch up so as long as I don't go extreme and flex the steel so much that the paint cracks I should be good - unless the contact point then moves to a new spot!