A bit long..
So, ever since i bought my '17 CX-5 in 2019, I've been really frustrated by
1. a launch, tip-in hesitation that gets REALLY annoying in hot weather,
2. very sluggish throttle response at slower speeds and 87 octane fuel,
3. along with some slight but noticeable surging in acceleration in city driving.
By the time it got to 40 mph or faster, and running 89 or higher octane, throttle response was decent, but not urgent. 91 octane was by far the best overall drivability. Yet, the around town drivability has been distracting.
In almost 3 yrs time I had carefully measured, and eliminated as
root causes;
1. MAF curves for 4 or 5 sensors (there are tangible variations among sensors),
2. careful measurements of primary and secondary ignition (virtually perfect waveforms and very rare single count misfires)
3. careful measurements of upstream and downstream sensors. Virtually perfect and good as compared to other skyactivs
4. careful tracking of fuel trims, ignition timing and other ODB data. Pretty much spot as compared to other skyactivs
5. on-rail cleaning of injectors.
6. after making my own in-cylinder pressure transducer, compared crank sensor (CKP) timing to TDC. Perfect; 20 crank pulses = TDC.
7. I previously measured both cam sensors (CMP-I, CMP-E), but didn't find any known good waveforms that referenced cams to CKP pulses on the various "free" waveform sites. Until the last two weeks...
A tech on Rotkee published a scope capture with CKP, CMP-E and CMP-I together, for the conditions of idling with the intake cam VVTI motor power disconnected and the exhaust cam VVTI control solenoid disconnected.
I measured my CX-5 in same fashion and discovered that the exhaust cam signal was spot on at 20 pulses of the CKP wheel (same as Rotkee waveform), however the intake cam signal was at 37.2 CKP pulses, as compared to the Rotkee waveform intake cam signal , where it occurs at 35 CKP pulses (exactly 90 degrees different than exhaust cam under those conditions). I double and triple checked the measurement. The intake cam was indeed 2.2 CKP pulses late (retarded), which is about 13 CKP degrees or 6.5 CMP degrees retarded. See attached image.
The next questions were why? How does the PCM *not* set a code for this?
After studying the service manual and internet photos of the valve train and closeups of the camshaft pulse wheel, there seems only two possibilities; the timing chain was off a tooth for the intake cam, or the intake cam pulse wheel was pressed onto the cam incorrectly, or the pulse wheel had shifted. As for why the PCM didn't flag this, unknown; maybe within some operational tolerance?
I any case, the PCM is able to command a certain cam position (in degrees) and the VVTI system complies perfectly. What the PCM apparently doesn't know, is that the cam pulse wheel is lying about the physical location of the lifter lobes.
So, after noodling on this a while, it seemed that I could either pop the valve cover, check the pulse wheel position on the cam (not sure if there are alignment marks), and then attempt o move the pulse wheel. OR I could fiddle with the cam sensor and shift its position to detect the pulse wheel EARLIER. I chose the latter.
I calculated needing to shift the nose of the sensor about 3mm UP, which meant about a 1.8mm gap at the top of the sensor mounting flange. I loosed the bottom sensor bolt and wedged in a small horseshoe shaped piece of circuit board (~ 1.7mm), and applied a dab of adhesive caulking to keep it in place.
So...a SHOCKING change in overall throttle response; most dramatically at low speeds, but definitely noticeable at freeway speeds. I wont have a chance to measure the CKP/CMP-I waveform for possibly 10 days to see how close to 35 CKP teeth the CMP reference pulse is.
My long term plan is to remove the valve cover and see whether it is a wrong pulse wheel position or if the timing chain is off a tooth. Then decide what (if anything) I want to do about it. Depends on how close to nominal my DIY intake cam sensor shim is working, I think.
Anyway, for anyone else who has a skyactiv vehicle that seems much more sluggish than an another CX-5 or Mazda they've driven, this might be worth looking into. I will note two different Mazda dealerships looked into this specific complaint and claimed all was normal. No codes, fuel trims normal, cam timing PIDs all normal, no misifres, etc.
View attachment 313453