Rhythmic RPM Jumping

Malevolous

Malevolous

Greenie N00B Member
Greenie Member
Year/Make/Model: 2007 Mazdaspeed6
Mileage: 80K on Engine / 144K on Odometer
Location: NYC (Sea level)
Concern: RPMs flutter rhythmically between 1K-2K once put in any gear with foot still on clutch
DTC's: P0108 (Manifold Absolute Pressure/Barometric Pressure Circuit High Input), P0327 (Knock Sensor 1 Circuit Low Input), P2006 (Intake Manifold Runner Control Circuit Stuck Closed), P2010 (Intake Manifold Runner Control Circuit High), P0133 (O2 Sensor Circuit Slow Response)
Modifications: Overspeed Built Engine, Mac EBCS, Damond Vacuum Block, Innovate PSB1 Gauge (New), Fused Underhood Busbar (New), Custom 4" Intake (New filter), Garrett GTX3071, Emusa FMIC, Corksport 72mm TB (New), Corksport IM (New), Corksport PI Rail w/ custom returnless lines (1050X Injectors - all disconnected, still tuning on Base Map - all brand new), Autotech HPFP (Replaced all seals with Rey's E85 seals), CPe Exhaust Mani, M2 3" Catless DP, Straight Piping Exhaust, MSWT full wiring kit (New), Deleted EGR/VTCS, Bosch 4Bar MAP (New), Turbosmart CompGate 40 (14PSI Spring - New), Turbosmart GenV Raceport 50mm BOV (New), Blow-through MAF setup (New OEM Sensor), Overspeed Injector seals (New), plenty more but I don't think they're relevant.
ECU/Tuning Software: Versatune
Tuner: Damond
Is the concern intermittent? No.
Can you duplicate the concern? Yes, happened during both previous start-ups so far.
Recent Repairs: Major overhaul since purchase, first startup was a few weeks ago. Relevant "repairs" in addition to new Modifications listed above include: New OEM O2 sensors (up/down stream), SP63 Cleaned/Flowed DI Injectors, Replaced ALL fluids/oils, Cleaned grounding points, New serp belt, New Damond x3 mounts, New NGK 6510 plugs gapped .26 - .28 (Can't remember now), New OEM Ign Coils, New OEM Knock Sensor, and I think that covers it.
Correction:

Just to try and give a better picture, the car's been down for a little over a year for all this maintenance and upkeep. First startup was about a month ago where I logged the Base Map to send to Damond. It initially had P0327, P2006, and P2010. I know the P2010 and P2006 is due to deleted VTCS. I have a Greg Sange relay installed but it doesn't seem to work as I still have these CELs. Planning to plug the black box in to take care of those.

The P0327 eludes me as I measured resistances & voltages to the wiring and to the knock sensor. Everything was in spec as per the FSM. Even replaced the knock sensor with a new one and still shows the code and it can't be wiped with an OBD2 reader, it just keeps reappearing. All connections seem fine and I don't see any corrosion. Yesterday I started to think it could be due to the 14 month old fuel that was sitting in the tank (1/8 fuel level) but I dumped 2 gallons of fresh 93 into the tank before starting it and then topped the tank off with fresh 93 immediately into the first driving log for Damond, so still not sure.

One thing to note, during the first drive with car in gear and foot on clutch, the car still seemed to slightly accelerate as I was going up a small incline. Possibly vacuum leak?

Damond replied that everything looked fine and asked for a 2nd log with a couple of WOT pulls and also recommended I check vacuum connections for that slight acceleration with clutch depressed, but the vacuum leak was too small to be problematic for this 2nd log. When going to start the car up a few days later for those WOT logs the car started slowly, and rhythmically, jumping between 1K and 2K RPMs the second I put it into any gear. Once back in Neutral, it would slowly climb to 2K and then settle back down to 1K at idle. No stutter or anything, the idle is pretty smooth. The car literally just sat for a few days after that first drive/log before this started happening.

I bought an Amazon smoke machine to test for vacuum leaks and actually found a few. Tightened everything back down yesterday and re-tested. No more visible leaks. Tried starting car up for the 2nd time and RPMs still flutter as soon as it's in gear. Hooked up the OBD2 reader and saw these 2 new codes: P0133 & P0108.

As per Damond's advice, I am going to try spraying some MAF cleaner today around the intake couplings and see if the RPM changes at all while idling. I'll also check on the MAP sensor connector and see if it's all secure.

I know this is a lot of info but I wanted to be as detailed as I could to help paint the full picture. Also attached the log of the RPMs fluttering in case it could be of any help. Everything above with "(New)" was replaced during the overhaul.[/COLOR]
 

Attachments

Last edited:
Update: MAF spray didn't seem to reveal any leaks and the MAP sensor was connected properly. Third start-up and it still jumps between 1-2k. Not sure what else can be done now.

2nd Update: Rather than make a new post, I had the idea that maybe the way my vacuum lines are run is a problem. I've never dealt with so many lines, or vacuum blocks, before so maybe my routing is the issue, or the way I've run my catch cans is an issue.

The Damond vacuum block I'm using is setup like this: Left side port blocked, Right side port to Intake piping nipple (Post-turbo), Bottom left port to BOV, Bottom center port to EBCS center port, Bottom right port to PSB1 AFR/Boost gauge.

Also the EBCS has the right side port VTA and the left side port going to the wastegate.

Catch cans are the Damond Stage 3 kit. I have one vented can bolted to the framerail near the trans. Top hose runs up to the VC rear port and bottom hose runs to the left port on Damond PCV plate. The 2nd sealed can on the passenger side has the top hose run through a check valve and then into the CS IM port, and the bottom hose runs to the PCV valve on the right port of the Damond PCV plate.
 
Last edited:
Update 3: Bought a Power Probe 3 and ECT3000 to try and locate a break in the wires and keep getting inconsistent readings. Wasted 2 days just trying to trace the 2 knock sensor wires. I decided to give that up for now and try doing a continuity test from one end of the wire to the other. The orange wire with blue stripe has continuity all the way up to the connector under the dash and the black wire with white stripe also has continuity up to the fuse box. I unplugged the main fusebox connector on the left side to test all pins there with the black wire and the terminal that showed continuity was the thick red wire with yellow stripe. I believe that wire goes to the starter. Does that sound normal?

Sent from my SM-N986U using Tapatalk
 
As a test, pull the filter on the vented can and block off that port with something to prevent fresh air entering the engine.
If that works, you *might* be able to get away with something like this, but if your blowby numbers are large the car will run like crap:

upload_2024-5-31_9-5-45.png
 
Enki said:
As a test, pull the filter on the vented can and block off that port with something to prevent fresh air entering the engine.
If that works, you *might* be able to get away with something like this, but if your blowby numbers are large the car will run like crap:

View attachment 28989
Click to expand...

Thanks for the advice - so right now, I basically have the entire intake/battery area connections disconnected with all the wiring exposed. You think this is more of a vacuum-leak type of issue? Should I just connect everything back up and plug up the catch can and try driving it? I tried avoiding the setup you linked so I could keep the intake tract oil-free lol but yeah, I'll give this a shot and see if it all works out. For what it's worth, the car had only one catch can installed when I bought it, but the can didn't have a lot of oil in it, so I can only assume the blow-by isn't too bad.

Also, just curious, you think the vented catch can could be causing the Knock Sensor, Map Sensor, and O2 Sensor codes?
 
It's possible this is related to the angle your PCV/Check valve is at; if vertical, I could see it opening up sporadically. I had similar RPM hunting issues whenever I would go clutch in from a pull due to my PCV setup at the time, but it would settle once I came to a stop.

Codes seem unlikely to be related.
 
Makes sense, thanks for the direction. So since codes likely unrelated, I'll keep searching for this invisible short and plug up the catch can before buttoning everything up again.

Do you have any special tips looking for shorts? I tried continuity already (which there was) and a visual inspection looks good, no tears. ECT3000 stops beeping to signal a short, then when removing parts to visually inspect the area, the beeping magically comes back.

Sent from my SM-N986U using Tapatalk
 
It would help if you posted what the codes actually are; everyone having to come here, read this, then google them themselves is not really ideal.
 
Enki said:
It would help if you posted what the codes actually are; everyone having to come here, read this, then google them themselves is not really ideal.
Click to expand...
Good point - original post updated, thanks. Sorry for the novel, I tend to do that.

Sent from my SM-N986U using Tapatalk
 
New update: Checked wiring diagram (Thank you Metal Members section!) for Knock Sensor routing and come to realize that the Orange/Blue & Black/White wire I've been tracing actually belongs to the "Variable Swirl Shutter Valve Switch" ... the Knock Sensor has the same exact connector with a solid Yellow and solid Brown wire, which I don't even remember seeing anywhere in the engine bay. This could explain the RPM fluctuating, the knock sensor codes, and the VTCS relay "not working". Not sure if it's also related to the MAP sensor or O2 sensor codes. Planning to look over all connections over the next week and report back if this is what the issue was, in case someone else does the same thing I did.
 
New update, got it all plugged up and put back together again, hopefully with all the correct plugs connected. Started up after leaving battery disconnected so all codes are definitely wiped. No code at all during startup. Let it warm up a bit to keep an eye on things and after ECT was above 190, did a few quick revs to check for any issues. Eventually got a code but only P0108 (Manifold Absolute Pressure/Barometric Pressure High). Maybe the baby oil mist from the smoke test?

Sent from my SM-N986U using Tapatalk
 
Enki said:
MAP sensor issue or you're not tuned for the sensor you have. Yes, that's a thing.
Click to expand...
Thanks for your help. The MAP Sensor should be included with the Base Map I received as Damond asked for the Part # of the one I have, but I'm just assuming. I'll confirm with them today. If it is tuned for, is it safe to clean it with some CRC Electric parts cleaner?

Sent from my SM-N986U using Tapatalk
 
Enki said:
If it's new, shouldn't need cleaning. You could also have a vac leak; post a log.
Click to expand...
Vac leak is possible, I had a couple small ones before. I found them using a smoke machine from Amazon. It creates the smoke using baby oil. The smoke actually left a light coat of baby oil along the inside of the intake piping. That's why I had the idea to clean the MAP, maybe it got a light coating of that oil on it.

Is an idling log OK, or do you want to see some driving? Revs?

Really appreciate your help [emoji482]

Update: I think I found the issue - Pic attached. There's no winning with this car.


Sent from my SM-N986U using Tapatalk
79ed8e66015da574a9989a7802244dd1.jpg
 
Last edited:
RIP. Idling log should be fine if you have the code at idle too. Engine should be warm though.
 
Update! I spliced in a brand new MAP pigtail last Friday. Have started the car and let it idle to get to operating temps. No CEL and idling seemed smooth around 1K RPM. As soon as I put it into any gear, the RPMs rise to about 1500 and stay there - unlike before where they would flutter. This happened both times when starting and idling the car.

The second time, for some reason, I decided to pump the clutch a bit and put it into gear. I discovered that *sometimes* when doing that, I would be allowed to put it in gear without the RPMs rising, but it didn't help 100% of the time. I sprayed MAF cleaner over all the silicone couplings and around the IM/TB, but the idle never changed. Even went over all bolts with a wrench, but everthing is tight, so I don't think it's a vacuum leak. When I plugged in the code reader, it showed a pending P2188 (System too rich at idle). I erased it with intentions to try idling it again today and taking a log for Damond. I'll post the idle log back up here, hopefully it will help to figure out what's causing this.
 
Managed to pull this log tonight. The logs are split because I popped a CEL P2010 (Intake Manifold Runner Control Circuit High) and reading the code interrupted the recording. First log is cold start bypass and 2nd log is some operating temp revs, and gear shifts while holding clutch, note the RPMs climb with no accelerator input (although throttle position reading climbs by about 1.5% each time). Hoping this might reveal something? There's also a portion towards the end that shows a slight rev during the RPM climb, and that would cause slight oscillation (2 or 3 cycles) of fluttering RPMs before it would dip back down to 1K and rise again to a steady 1.25K - 1.5K RPM.
 

Attachments

You must log in or register to reply here.
Back
Top