So, a MAP sensor, TPS sensor, and Timing spike walk into a bar...

Actually, I'm about to walk into a bar and never leave after chasing this (don't call it a hiccup) issue. It's more like a "bark", and will consistently happen after cruising at a steady (let's say 18%) throttle, then applying more throttle (probably 30-40%) and you can hear the engine cough, it jolts the car, and then resumes moving forward. Tonight it was sharp enough that I managed to capture three different instances on the monitoring graph of my MBII. Watching these individually so that the sample rate is the best possible, see below. The sharp spikes all happened when the problem occurred.

Base info: I've replaced the throttle body (stock both times), MAP sensor, IAT sensor, Intercooler Pump, Cam and Crank Sensors, Front o2 sensors, bypass valve/SC throttle body*, and fuel pumps are new. NO CODES in STAR, fuel trims under 3%.
Mods: 77mm fixed pulley, RaceIQ Tune, 550cc injectors, BWK.
*When it coded for a bad bypass valve about a year ago, I replaced it with a different Mercedes part number per another thread stating it was the same as the AMG one but just without an extra rubber piece. They were identical as far as I could tell and the wire harness fit fine. Since pulling the SC to try swapping in a different one isn't the simplest thing to try, I haven't eliminated that this could be a problem. However when commanding valued in STAR it is responding as it should.

WOT is just great, no complaints there and makes great boost, but obviously something is wrong during normal driving. Timing should NOT be spiking to +15, TPS to 90%, and a boost spike to suddenly drop off. I'm also not seeing MAP readings above 100 kPa unless giving it decent throttle, and I would think that with a 77mm FSP that I would see boost in more low rpm conditions.

Thoughts? Is a spike this short and extreme 'mechanically possible' or do I have an electrical gremlin somewhere?

 

Vacuum leak?

 

I've always used new OEM gaskets any time I've had the SC off, torqued to spec, I've sealed the brake vacuum port permanently and that tube on the drivers side is seated correctly. No codes either, and the problem doesn't seem to be temperature dependent either.

Is it possible to see such dramatic datalogging spikes with a small vacuum leak? Would the TPS really open wide up if the car thought there was a loss of vacuum?

Admittedly I don't own a smoke tester, but the cost of a decent one would probably be more than a couple hours of my time pulling the SC and resealing everything again. I'm happy to do it, but more info would help me feel better about making that call

I'd be curious if anyone else with issues, who saw similar readings, found something small like a vacuum leak as the cause.

 

I was watching Alex on Youtube and he had similar issues. One time it was a vacuum leak and another a special way he had to reset the computer. Watched it yesterday is a matter of fact.

 

Since you have a fixed pulley ....i am leaning towards a tune issue, but...

Is it possible to remove the blower belt and drive it. I know I can do that with no issues but not sure with the fixed pulley.
Would for sure try that and see what happens.

 

I can mostly eliminate the fixed pulley, as it did this when I had the same size but clutched earlier this year. It was worse actually because I'd occasionally get the clutch engagement happen right before or after this and it was like a double-cough.

I emailed Tony but am still waiting on a response, as I've upgraded the car step by step he's sent me multiple tunes and this problem has always been there, so I'm again leaning towards a mechanical issue on my engine's part vs a tune, especially since my fuel trims are all in great shape.

Driving with no boost it is smooth, I just remembered I had an odd occurrence after doing a reset of some sort that I drove the car and it felt way down on power but smooth, hooked up the scan tool and found the MAP wouldn't go over 100 kPa which is atmospheric, so boost wasn't getting to the motor. The only explanation was that the bypass valve was fully open. Unplugged the battery and it went back to 'normal'.

I think maybe I'll man up and pull the SC again, maybe get another bypass valve that matches the OE part number and use some extra sealant everywhere. I got a gasket kit, but I don't remember if that rubber piece on the bypass was new or not and watching Alex's video he mentions that as a common leaker. His concern about power dips after shifts is the torque protection limits in the TCU, my new TCU completely got rid of that and this problem isn't after shifts, it will happen in one single gear.

 

I would then say not mechanical if you have no issue at WOT.
Unplug your ECU and look at all the pins including the connector just to make sure but clearly the ECU is seeing something it does not like.

Also do you have a wide band installed?

 

Sounds good (and easy to check!).

I temporarily ran a dual wideband in the rear o2 ports (my headers are still stock) after my tune to make sure things looked good, but was only focusing on WOT readings at the time, which were fine.

My only thought to support the reason why it's fine at WOT is maybe because it's in open loop, and if there is a small vacuum leak it would only make a small difference when under full boost vs when I'm just cruising and could make more of an impact due to the amount of air the motor is using?

 

Your description of the problem matches my experience EXACTLY. I've been fighting with this for a long time without any luck. I've changed from the 82mm TB back to stock, and it persists. I have a RaceIQ tune too, with a 77mm clutched pulley.

The only thing I can add at this point, is that SOMETIMES my check engine light comes on, and while it's on the car actually drives NORMAL. That leads me to believe it has something to do with the tune, and it runs fine in open-loop mode, but that's just a guess. I haven't been able to pull a code while the CEL is on, as it usually goes off ( and the not-a-hiccup returns ) the next time I drive the car.

 

Interesting, and hopefully we can solve this! I have a notion I might upgrade to an 82 if I have to pull the supercharger (because why not?!), at least do the snout boring modification so I have options.

The only CEL I've ever had is the B28 implausible MAP reading (P0106), and that was on a LONG boring highway drive (4+ hours). I pulled the freeze frame data at the time and it showed 37% load, closed loop, 56 kPa (so about 6.5 psi of vacuum if I'm correct), and 78 mph / 1900 rpm. IATs were 122 deg (it was the middle of summer), timing at -22.5. Long term fuel trims 1.6%, and short term 1.6% and -6.3% on bank 2. It cleared itself the next day, so maybe that is what you've seen (I just happen to keep my scanner in the drivers door pocket so I pulled that data immediately after it came on). That much vacuum given my speed and load (I would have guessed I would be higher, or maybe into boost under those conditions) led me to believe the MAP was reading incorrectly.

I did a little research and decided to replace the MAP sensor, didn't really experience any difference though (and I used the Genuine Mercedes part #).

I've also tried the idle modification in STAR that the 82mm guys recommend, also making no difference so I set it back to stock 650 rpm.

 

Joe if the light comes on the code should be there regardless no ?

 

Yeah, I guess you're right. I just haven't ever plugged in the scanner. Maybe I'll go do that and see if there are any stored codes.

 

More logging this afternoon. I had one instance earlier that the MAP spiked but TPS did not, which leads me to look at the bypass valve suddenly closing then opening again vs a TPS problem.

STAR doesn't show you the opening % of the M16/7 actuator, but you can watch the two voltage values and drive. Not knowing which voltage signal is which, here is my take: The flap is open at idle, stays open as you slowly accelerate away, then I got a hiccup as it was closing, then it stayed closed while driving and decelerating only to open again when fully stopped. If this is the case, then the other thing I found was that when I accelerated harder and did not get a hiccup, the flap opened up again almost all the way, which seems like something it shouldn't do.

So in short - my 'clean' accelerations, at least around my neighborhood not going super fast, the bypass valve is opening and not allowing boost. Is this due to a problem with the actuator, or another issue that the ECU is instantly protecting things? Despite having a fixed pulley and tune for it I can still see the supercharger clutch commanded on and off, and when I've checked it stays on until coming to a stop, so you would think if the ECU is protecting things it would also disengage the clutch, but I'm not sure what nuances went in to Tony's tune for the FSP.

 

Here's the vid of the bypass operating (and not far into it you can watch and hear the hiccup ocurring).

It's a big ask, but if anyone happens to be driving around with STAR and wants to watch these same voltages you would be my hero! It's in control units-drive-ME-complete list of guided tests-M16/7 actuator and then watch using actual values.

I'd be very curious if the behavior of a smooth car is the same or not.

 

Alright, pulled the ECU today and all the pins look great. Wiring looms have wetness from the TCU harness leaking 2 years ago but it does not extend to the connectors and I could see no breaks in wiring etc. I even pulled the cover off of the ECU to look at the circuit board and it looks dry.

To add to the confusion, I had three distinct recurrences of the problem driving into work today. Once, the TPS spiked wide open. Second, the boost spiked from vacuum to 8 psi. Third, the timing shot from -20 to +15. This was all with 30-40% throttle around 30 mph, nothing extreme. I had monitors on all three of these signals, and for each occurrence the others remained the same. This is just too weird, unless the sample rate was slow enough watching 3 signals that it just happened to miss the other two parameters spiking at the same time.

I think I may invest in another wideband to plug in somewhere and see what the AFRs are doing, however it seems unlikely a tune issue could cause such weirdness when I'm sure Tony has done a bunch of cars w/ 77mm pulleys and larger injectors.

 

It looks like you are using the innovate logger from the screen shots. You can log TPS and Bypass as well as factory Map by using their analog cable. I can send you what I did or if you are on the Facebook AMG pages I uploaded a document for all the connections.

I had the code issue with the 82mm tb, still do, but nothing else, no lean condition. Put the 74mm back and it ran fine.

I have the new Zerocat TCU and my issue is all but gone. Only shows up on shift from 2-3 and only at small openings and transient openings, say 15-30% tps at a slow to medium acceleration ramp rate. Very slow or very fast it is not present. I feels like the bypass is too slow to close.

Injectors are mapped using load as a calculation (or load is affected by what is used to calibrate for them) so it may be at lower loads we are into tables that are conflicting with what the expected Map data should be.

Killing timing and opening the bypass seem to me to be a way of killing torque so we may be at the mercy of the tcu ecu interrelation with regards to traction control. I don't have the logger anymore but can help you with set up if needed.

But then again i did stay at a Holiday Inn Express last night so who knows? ;')

 

Thanks, yeah those screenshots are with the Carsoft MBII which can access a lot of things, but not bypass. It doesn't log either, I have a bluetooth and logging program on my laptop but the sample rate is slooooow.

The zerocat tcu definitely helped get rid of that torque management issue, shifting feels SO much better, but this issue is not during or near shifting, it's the same gear as I go to accelerate more after cruising (and not enough to signal a downshift). I can keep it locked in manual mode in 3rd or 4th gear and get it to act up.

When I drive home I'm going to see if I can pull up injector duty cycles and see what they are doing, mine are 550cc so shouldn't be taxed under these conditions, but maybe the data can point me somewhere if it doesn't look right.

 

Just a data point, my 630cc injectors are at 60-65% max duty cycle. I would expect you to be in 75-80% range tops.

 

UPDATE:

I think I've found a definite path to start heading down. Took off the TB and ran the STAR actuator test of the bypass valve while actually watching it, and it's just fine. Responds immediately and accordingly. So the next thing I did was prop the main throttle body up so I could watch it from the cabin and ran actuation tests on it, and here's where things got funky. Standing voltages pass, then it asks you to check with the kickdown operated. I slowly depress the pedal and by about 25-30% through the TB is wide open and stays for the long journey to the kickdown switch. Voltages match too, the TB voltages climb to max by the time I'm 1/3 of the way through the pedal, and the pedal voltages continue to climb until I reach the kickdown.

Then, I opened the menu 'throttle body learning' and it states to turn the ignition on for 60 sec, then off again for 10 sec. About 30-40 seconds in you hear the TB click a few times and the butterfly moves back and forth, but only opens about 20% of the way, it doesn't open fully which I'd expect it would to 'learn' it's entire travel distance.

I have videos of this but you probably don't need to watch them with my description.

So....I'd normally say replace the TB, but I have already with no change, both stock TBs and the newer one was a known good unit from a guy who upgraded his supercharger. Any thoughts? Could this be some error in tuning from Tony that changed the operating voltage range of the TB? Could my accelerator pedal be bad (even if the voltage sweep for it is pretty linear)? Strange things are afoot.


(Injectors are fine, running WOT in 4th gear at 2k rpm I'm at about 15% duty, obviously it would be better to have a higher speed run but traffic sucked and if it's that low under boost at lower speeds I'm not concerned).

 

You really need a scope to accurately look at the signals. It looks like the MAP, and TPS, are momentarily losing the signal return, ( low, or close to ground ref). This would cause the signal out, to pull up high.

 

That would make sense, and I agree as I hate the terrible sample rate through OBD2. It just seems like the 'scale' is altered as I'm getting both voltage signals on the pot switches acting normally, just topping out way too early.

P.S. hello from a fellow Snohomish County resident! 🙂

 

After looking at your first post, you show an indicated 40% throttle, and then it reports to the ECU, an instant 87% throttle opening, without you moving the accelerator pedal. I would monitor both TPS signals, and the accelerator position sensors, and verify correlation. I doubt the throttle itself is even moving much, during this event. There would have to be throttle codes, and a MIL, plus even default, in either case, unless the ECU is wacked. Grounds on the left, right cylinder heads OK?

Yes, Snohomish County, where the action is. .

 

Seems very strange that it has to be in a certain condition to fault so maybe those exact voltages between the throttle body, peddle and map sensor are triggering a ECU fit for a split second.

Can you down load your factory tune just to make sure before you start really sleeping at the bar ? LOL

 

Just downloaded factory tune, no change. Plugged in my old TB, also acting exactly the same.

My theory - As I went through the TB learning reset (key on 60 sec, off 10 sec), I watched as both throttle bodies snapped closed, and then only partially open, then closed again. Right after the TB does its thing, the bypass throttle body does the same thing (and I can watch it with the TB off), but it is snapping fully closed, then fully open, then back again. My logic tells me the TB should be doing the same but isn't.

So, if it thinks it's doing a full sweep, but isn't, that would explain why partial pedal movement brings it to fully open as its scale is calibrated to this narrow voltage range. If this is the case, the problem is with the learning function of the ECU and the potentiometers are just doing their job (and why there are no fault codes, and the voltage testing values in STAR are all within range).

New and stock tunes both doing the same thing, so it's not the tune. To me that leaves crazy internal ECU fault, or bad wiring somewhere? I wish there was a way to mount a tiny camera inside the TB while driving haha.

Any ideas?

 

So just so I'm on the same page....there is no problem with the blower belt off and no problem under full acceleration correct.?