When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
SpitfireS: That's pretty interesting.
Now you should try and change back to the other sensor, or even better have someone else change and it without you knowing
i'm still waiting on my new map sensor and tie wrap housing to arrive from HondaPartsNow but in the mean time i wanted to record my MAP data vs RPM and see if whacking/cleaning my MAP and then zip tieing it would do the trick
1 - before we cleaned the MAP and went out for a test drive the data log of the MAP vs time looked super unintelligible... the graph was all over the place and there was no clear correlation between acceleration and increasing MAP. unfortunately i forgot to take a screenshot of these graphs but take my word for it, the value jumped up and down all over the place constantly
2 - after we cleaned the MAP and zip tied it then checked the log the smoothness was NIGHT AND DAY. i also believe i can feel the same thing you guys are talking about, the car felt so much smoother after we cleaned the sensor almost like a brand new car. looking at the data logs you can see there are clear peaks and valleys now as i accelerate and let off the gas. this data is actually valuable whereas before the data was too sporadic and you could not use the MAP line to distinguish when i was accelerating
3 - unfortunately i still ran into the dread misfire after we cleaned the MAP. it took about 20 3rd gear pulls on the freeway but it finally started occurring again. this time i got a solid CEL... same bullshit, 5 stored codes 0300, 0301, 0302, 0303, 0304 and 1 pending code 1399.
so im kinda back to where i started (still have my appointment booked at evasive coming up in ~10 days). we tried to see if there was a correlation with the MAP values when the misfiring occurred but it didn't look like the MAP spiked too much... the one thing we noticed was that when the map didnt spike most of my pulls would plateau flat then drop off whereas during the misfire pulls the MAP would ripple up and down before dropping (basically not a clean, flat plateau like the other pulls). we aren't sure if this is cause or effect... aka is the MAP sensor posting a value a tad too high which is then causing the misfire etc... or is the MAP fine it's just rippling bc of the misfiring that's occurring?
again we're going to test it after installing the new MAP sensor on wednesday when it arrives... until then i'm still stumped
Last edited by ericmdub92; Jun 23, 2018 at 08:34 PM.
^ First: THANK YOU for sharing the data, data is good!
It's good to hear cleaning does not damage the sensor, it would be nice to hear what you did to clean it.
Reading online about it, the MAP cleaner manuals all say to spray some on a rag and wipe and NOT spray anything into the sensor directly.
Looking at the sensor, how much is cleaning the outside going to do, I mean, the actual sensor is IN the housing.
I guess if you spray some of the cleaner into a cup and then pour it in that 'tube', shake a bit and let it dry out might work?
The ripple in MAP producing the CEL is in the seconds time range?
Did you feel / hear anything during this?
Or notice something like a different exhaust tone or something?
(I guess not as you would have mentioned it...)
If the MAP is leading and the ECU injects/fires based on it, it would be 'silly' if a rippled MAP would make the ECU follow that ripple (make the engine run like a misfire) and then show a misfire.
One would hope the ECU is smarter than that ;-)
You can drive around in second gear in VTEC all day and no problem?
4th gear?
Here are 2 screenshots of a short drive I did an hour ago.
Resolution is about max 2 datapoints per second but it wasn't fixed, as in, it changed a bit (I guess) based on value change.
Asking flanders: what is the max data transfer from the ECU?
Here I used what you suggested, RPM, TP and MAP.
When using those 3 values, is twice a second what you get from the ECU or is my ODBLink the bottleneck?
For some reason the TP never goes below 5% and never above 91%.
The OBD scan tool is a OBDLink, the software is ScanXL, connected via USB to a laptop.
I'm not sure the resolution is high enough to actually see a difference between MAP's with this ODBLink.
The car still feels good, not that the ECU corrects itself for the different MAP signal or something.
Maybe a ECU reset will fix it IF it starts to 'get back to normal'.
P.S. The whole set is also stored as a .csv file but I don't have any software at home to open/edit it.
I will take it to work tomorrow and see if excel can make nice graphs.
^ First: THANK YOU for sharing the data, data is good!
It's good to hear cleaning does not damage the sensor, it would be nice to hear what you did to clean it.
Reading online about it, the MAP cleaner manuals all say to spray some on a rag and wipe and NOT spray anything into the sensor directly.
Looking at the sensor, how much is cleaning the outside going to do, I mean, the actual sensor is IN the housing.
I guess if you spray some of the cleaner into a cup and then pour it in that 'tube', shake a bit and let it dry out might work?
The ripple in MAP producing the CEL is in the seconds time range?
Did you feel / hear anything during this?
Or notice something like a different exhaust tone or something?
(I guess not as you would have mentioned it...)
If the MAP is leading and the ECU injects/fires based on it, it would be 'silly' if a rippled MAP would make the ECU follow that ripple (make the engine run like a misfire) and then show a misfire.
One would hope the ECU is smarter than that ;-)
You can drive around in second gear in VTEC all day and no problem?
4th gear?
Yes, the ripple in the map is in the seconds range, it's not too clear in the image i attached because there are around 5 pulls there but the first one has a good example of a flat plateau and the 4th one is where the ripple happened - you can see how the first just flattens then droops down while the 4th has a sharp spike upwards before dropping...
When this happens I can feel the gas pedal/engine go limp (like it's cutting off fuel pressure and has hit an RPM wall). Further pressing the gas pedal while the engine CEL is flashing produces a sound that is like a clicking/crunching sound (probably metal), almost akin to a toned down version of the sound the starter makes when the battery is dead.
I agree in that I hope the ECM would be smart enough to have the MAP sensor spikes lead it to misfiring, just don't know what the reality of that scenario is... I think from what I'm reading it seems unlikely that the misfires are caused by my MAP at this point, probably something else..
And yeah I can go around in VTEC all day in 1st and 2nd. Now i can do about 5-10 pulls through 3rd gear as well, it takes at least 10-15 pulls at the top of 3rd before this happens.
Here are 2 screenshots of a short drive I did an hour ago.
Resolution is about max 2 datapoints per second but it wasn't fixed, as in, it changed a bit (I guess) based on value change.
Asking flanders: what is the max data transfer from the ECU?
Here I used what you suggested, RPM, TP and MAP.
When using those 3 values, is twice a second what you get from the ECU or is my ODBLink the bottleneck?
For some reason the TP never goes below 5% and never above 91%.
The OBD scan tool is a OBDLink, the software is ScanXL, connected via USB to a laptop.
I'm not sure the resolution is high enough to actually see a difference between MAP's with this ODBLink.
The car still feels good, not that the ECU corrects itself for the different MAP signal or something.
Maybe a ECU reset will fix it IF it starts to 'get back to normal'.
Regarding the TP... my OBD software reported my TP as 5% with the pedal completely untouched and ~91% at WoT - might be normal?
Last edited by ericmdub92; Jun 24, 2018 at 07:44 AM.
Reason: change 96% to 91%
The older pre-DBW ecu has very slow communication, I think mine was around 10Hz with some cheap chinese OBD-adapter.
I'ven changed to a PRB ecu now with kpro which got a nice 100Hz refresh rate.
Here is a 20s window with steady throttle input for some kind of reference, looks pretty clean (especially for an ITB setup )
~5 and 91% TPS is perfectly normal to see.
I've use csved for nice and easy .csv file editing.
Another quick update... pulled out the fuel injectors today and cleaned the O-rings/the openings on the intake manifold. Taking them to RC Fuel Injection here in Torrance, CA for cleaning and will throw them back in the car when the new MAP sensor finally comes in the mail.
)
