partial throttle surging

[Sand]

***
Year/Make/Model: 2007 speed3
Mileage:~180k on chassis, ~10k on my garage built engine
Location:SOCAL: moderate temp, no humidity, no altitude
Concern: Car seems to surge at partial throttle. Probably a coincidence, but it seems to happen in fourth gear mostly
DTC's: None
Modifications: Built engine, but still on stock turbo for the short term. I do have a Mazdaspeed Cold Air Intake, turbosmart 50/50 BOV/BPV, HPFP internals, cleaned and flow tested injectors with some brass seals(forget the brand), one step colder plugs gapped to lord knows what (maybe like 26?? i really dont remember, but i remember it was a widely recommended gap). hopefully not forgetting anything important...
ECU/Tuning Software: I have flashed an old COBB OTS stage 1+ MSCAI map (from an older style accessport) as part of the troubleshooting. The car used to buck hard after surging with the stock map and does not do this with the OTS map. My guess is that it was going into some boost cut kinda thing as the OTS map, among other things, raises the boost limit a bit.
Tuner: COBB OTS
Is the concern intermittent?: fairly consistent
Can you duplicate the concern? yes
Recent Repairs: I have tried changing to a COBB OTS map, checking for vacuum leaks (not 100% confident in my testing as I could not get my starting fluid by the PCV valve under the intake manifold), cleaned my MAF, installed the AEM air straightener for the MSCAI. Symptoms have not changed.
Correction: TBD
***

Amplifying info:
The car surges pretty hard multiple times at seemingly random intervals while under constant partial throttle. This has been an issue that has plagued the car since i first got it back on the road (bought it blown up maybe a year ago). It used to surge so hard that it would hit some sort of fuel, spark, or throttle cut and just about send my head into the steering wheel. I have learned to drive around it while I was fixing other issues, but its time to track this down...

After I tried a COBB OTS map and a few other things to no avail, I took a few datalogs on ye ol Accessport V2 (low update rate, sorry...) and noticed that the Accelerator Pedal Position would stay pretty constant (in orange below), but the "throttle position (%)" (in gray below) would jump wildly. Note: I am still on the COBB OTS map at the time of data recording

CSV attached as well

Throttle duty cycle (green) seems to also go crazy with throttle position (%), but a little different...

I dont know what the difference is between position and duty cycle, but sometimes one leads the other. I feel like if I had an AP V3 with the higher update rate, itd be easier to tell. Maybe one is like throttle position commanded by the ecu and the other is actual position seen by the throttle position sensor? I feel like if these are both actual throttle position, then maybe my throttle body is bad?? if one is commanded throttle position and the other is actual throttle position, then that means for some reason the ECU is telling the throttle to open up...

Most other variables stay pretty consistent: AFR holds steady ~14.5 through this section, no knock retard, but do not know if STFT (green) and LTFT (orange) are normal or not


Mass Airflow g/s (yellow) goes kinda crazy too, probably expected


and the wastegate opens up too (blue), probably expected


I am a data analyst by trade and love this kind of stuff, but am out of my depth here and could use some troubleshooting ideas before I just start replacing O2 sensors or the throttle body or something... I can also try to take some more logs with the accessport (i may not have had all the variables turned on in the accessport) or try out the datalogging thing with my cheapy OBD2 scanner and Torque Pro (have not done that before) to try to get like sensor voltages or something.

Also if you end up looking at my datalog, I am very curious at what you are looking for and why you might come to a certain conclusion!

Thank you in advance

 

[L337TurboZ]

Can you take KOEO log and post a graph of your Throttle % vs Accelerator pedal %

You'll want to make sure one of them isn't out of sync and cycling them while the car is off may change the data you get.

If they check out then check your throttle body connection for corrosion or bad pins. Maybe swap a different throttle body out to be certain if you have access to one.

To me it just looks like the throttle body motor is on its way out and it can't maintain it's specific spot.

 

[Sand]

Unfortunately, I dont have a nice scan tool that can trigger a KOEO (Key on engine off in case anyone else comes along this thread, and yes I had to google it lol) unless theres another way to trigger it... maybe plug in my PS2 controller and go up up down down O X X O select start

I have a friend who has some spare parts. I tried his throttle position sensor (little plunger thingy on the end of the intake manifold kinda under that wiring and vacuum line mess by the HPFP, no luck. Ill see if i cant borrow his throttle body next. Hopefully that fixes it, if not i may need some more expert opinions

 

[L337TurboZ]

You can take logs with the access port with the vehicle not running.

The "plunger thingy" you're talking about isn't a throttle position sensor. That's part of the variable runner control system of VTCS as it's called.

You have two throttle position sensors built into the throttle body assembly. Both are 5 volt reference. One starts off at 5 volts and as you go to 100 percent throttle it drops to 0 volts. The other goes from 0 volts to 5 volts rising with hitting the pedal.

They did that since it's an electric throttle body and you don't want a failure that could cause the motor to stick wide open. It's a redundant fail safe. The same principle is used in your accelerator pedal.

If the ECU sees that there is too much of a difference between the throttle position sensors (TPS) or Accelerator pedal position sensor (APP) then it sets a DTC.

You should be able to graph data with a scan tool at around 50 percent throttle the sensor lines should cross like an X. If they don't, or they drop out or seem erratic then there could be faulty wiring, faulty throttle body or pedal, or a bad ECU.

It takes some diagnostic work to narrow it down. Not trying to rip on you but if you don't know what you're doing take it to a shop that does. Instead of putting his throttle body on your car, put yours on his if his is running ok. That's a much better way to eliminate if yours is at fault. If his starts running bad you'll know what the failure is.

 

[Sand]

--Response to L337:
Suggestion attempted but unfortunately, the throttle does not open when i depress the accel pedal with the engine off.

I understand the confusion, the plunger thingy it is not a throttle position sensor in the traditional sense but is referred to as such by parts diagrams (example: https://edgeautosport.com/mazda-oem...7-2013-mazdaspeed-6-2006-2007-cx-7-2006-2012/). Not trying to get in an argument and appreciate the reply, just want to point it out for anyone else that comes along this thread.

great suggestion! I was able to borrow a nicer scan tool (see update below) and will try that out when i get a chance. I was able to trigger a KOEO and KOER test with no issues detected

Of course, I was jut hoping to get some free help from the good folks on the internet . Taking it in is definitely a possibility, but I have always done my own stuff and would prefer to keep it that way if I can help it. Unfortunately, he is currently in the process of an engine swap, so I cant test my throttle body on his car and hope I can resolve this issue prior to.

--Update:

I was able to borrow my friends throttle body and it did not fix the issue.

I also tested for vacuum leaks with some starter fluid, cleaned some connectors, cleaned my MAF, installed an air straightener in my MS CAI. I also tried driving around with my MAF and subsequently my MAP unplugged as a troubleshooting step, but as expected, the car ran so bad I was not really able to tell.

I was able to borrow a nicer scan tool as well and found out during a late night testing session that while Accelerator Pedal Position stays constant, Desired Throttle Position goes crazy along with Actual Throttle Position meaning that, for some reason, the ECU is telling the throttle to open way up. So, this makes me think it is either something wrong with the ECU itself or the ECU is getting some funky reading from some sensor that makes the ECU think it needs to crack open the throttle (wiring issue or sensor going bad). The fact that it is so consistent has me scratching my head. In the past, wiring or sensor issues have seemed to cause issues more intermittently.

I will try L337's suggestion with comparing the two sensors in the throttle body. I also received another suggestion that maybe the ECU is falsely sensing an increase in load and is trying to compensate by adding throttle. I will check it out anyway but am not super hopeful as I think it will be hard to tell if the load is increasing because the throttle opened up or if the throttle opened up because load increased. Maybe the next step is to try my friend's ECU to rule that out.

Any other suggestions are greatly appreciated!

 

[L337TurboZ]

Using a different ECU won't work cause it won't be programmed and you'll be unable to start it.

Have you tried load testing the wires between the ECU, Throttle Body and Accelerator Pedal?

See this post about how to load test wires. Should be the last part posted

https://mazdaspeeds.org/index.php?threads/how-to-diagnose-thread.12295/

 

[Easter Bunny]

Try unplugging the ECU and plugging it back in. Sometimes the separate a little bit and make things act screwy

 

[Andrew Sheffield]

Has anyone found a fix to this issue? I’m having the same exact same issue and it’s driving me insane! My pedal % stays steady but my throttle position and throttle duty goes crazy. It almost feels like my pedal is delayed or hits a brick wall on the percentage then once it reaches past that point it jumps to that percentage, Almost like a dead zone. It all started when I rebuilt the motor and put in a gen 2 transmission about a year and a half ago. I can send datalogs if needed but please respond if you know a fix or how to find a fix.

Mods-
Corksport hpfp internals
Hps short ram cai 2 3/4 pipe size
Turbo smart dual port BOV
Corksport catback
Corksport turbo inlet pipe
3 port Cobb boost controller
Damond motorsports egr delete
Cobb access port
Vtcs delete
Gen 2 transmission
Fidanza lightweight flywheel
Cst4
Tail external 44mm wastegate
Corksport exhaust manifold
Corksport Charles’s down pipe
Jbr fmic piping kit
Tr8 intercooler

 

[Sand]

Apologies for the lack of updates

I bought versatuner during black friday 2023 and was fiddling with stuff last winter. I found that under Accellerator Pedal Calibration > APP Mapping xxx Gear, the stock mapping is exponential:

and when i made it linear:

the surging went away and I have not experienced this surging since (prob 15k miles)

I see you have an accessport. if you can get ahold of an accesstuner license, im sure there is something similar...

That being said, I feel like there has to be a better explanation... maybe there is a narrow band of "reported APP %" low down in the "Actual APP%" that this occurs at and making this table linear makes it difficult to hit this funky spot where surging occurs?? not curious enough to continue troubleshooting though, I am just happy to have found a workaround

 

[Andrew Sheffield]

I appreciate the update, I’ll try to find out how to get ahold of accesstuner. I’ve looked In the past but I couldn’t find any purchase or download links. I’ll ask Justin over at freek tune as well to see if he can change the table to a more linear map.