Another M62TU P0021 Thread with Weird INPA Reading

[NolimitE39]

I have read through all the threads I can find regarding the P0011/21 codes, timing procedure, VANOS operation, and INPA. You seem to have had the exact issue I'm experiencing and I was hoping you could steer me in the right direction.

CLIFFS Notes: I had p0011 and p0021. I fixed 0011 by adjusting the sensor wheel on bank 1. Adjusting the sensor wheel on Bank 2 seems to have no effect. I am showing 41.93 degrees in INPA which does not fluctuate at all.

History: I recently did the timing guides along with new chains and ancillaries on my 4.6iS using the GAS master tool set. The VANOS units never exhibited any signs of wear so I didn't touch them. When I started the work I removed and reinstalled the cams on account of me loosening the VANOS sprockets with the cams 180 degrees from where they needed to be because I was going too fast and not paying attention. I didn't want to chance bending the valves so I reset the camos manually. I did the timing procedure twice because I broke the bank 1 oil distribution piece threaded collar when tightening the VANOS solenoid so I replaced it with a new one and retimed the engine. I've checked and rechecked the upper timing covers which aren't 100% aligned with the top of the heads (using victor reinz gaskets), they are about 1 - 1.5mm higher than the top of the heads but Bank 1 is not showing any issues.

Issue: I initially had both the 0011 and 0021 codes. I fixed 0011 by following the INPA freeze frame and analog readings for bank 1 which started at -15, then went to -20 (I turned the timing wheel the wrong way initially), then to -12, and lastly to ~+3. I have tried following the same procedure for Bank 2 but am getting some odd readings. The code freeze frame has since the beginning of my issues shown either 0 or -1. I have isolated the analog value on startup for faster refresh, bank one shows in the single digits both sides of 0 and fluctuates some as one would expect while bank 2 goes to 41.93 immediately and doesn't move however the code (34 / P0021) freezeframe still shows 0 or -1. The car all the while runs perfectly well, idles smoothly and exhibits no other codes or issues. I have tried swapping with another known good CPS and VANOS solenoid on bank 2 which has done diddly squat. Through my exercising I have confirmed that the Bank 2 cam blocks (GAS Master Kit) drop on and off without issue when at TDC with the lock pin in. The shutter wheel is absolutely straight and undamaged, extra care was taken in removal, storage, and reinstallation. The shutter wheel hole lines up with the test port on the timing cover. There was no slack on bank 2 during assembly and timing. I'm at an utter loss right now, I've been screwing with this for weeks, trial and error has brought me the same consistent strange reading in INPA. I don't have access to any other software or scan tools. I have a cheap AutoZone code reader which I am using to confirm that it is in fact still 0021. Any ideas? Any thoughts? Anything at all that may help? I really appreciate anyone taking the time to read this and respond. I'd like to avoid retiming the entire engine but based on my research it may be the last resort.

[JimLev]

P0021 cam timing out of spec with vanos off (over advance bank 2)
Watch this.
https://www.bimmerforums.com/forum/s...a-540tu-engine

I’d start over and retime it after you watch the video.

[NolimitE39]

Jim, I do appreciate you responding so quickly, and I've seen your videos along with about a dozen others on this procedure, not only that but I can now pretty much quote both the Beisan and GAS instructions. I really thoroughly do not believe that my physical timing nor the shutter wheel timing are the problem. I'll end up retiming as a last resort but I was hoping for some input from those who have obtained INPA readings because unless there is something I am missing my bank 2 reading from INPA is very strange being that it is the same number every time but it is not 50 which is the max available reading (at least that's what shows on the screen). I have found 1 other post with the exact same reading and issue from 2017 and reached out to the poster but haven't heard back.

[JimLev]

Are you positive your vanos were fully retarded?
Did you hold the cams with a wrench when you were tightening the cam bolts?
I do have screen shots of my cams using INPA. Will post them later, leaving for dinner out right now.

[NolimitE39]

100% that the VANOS were fully retarded, worked them for a good 5 minutes each to get any residual oil out. The VANOS actually laid on a shelf for about 2 weeks because I didn't have time to reassemble. I held the cams with the wrench when torqueing. This is why I am stumped because I did everything right and after the initial screwup I was even more careful to check and triple check each step. The current INPA readout is what's making me really think that its not a physical timing issue. Even if the timing was way out the live reading would still fluctuate some, mine is dead set at 41.93 so unless that's the weird limit value set by the extra precise German engineers I think I have an issue with the DME or the sensor circuit or something else is at play. Enjoy your dinner. Cheers.

[R Shaffner]

As I understand it, the code will deactivate the VANOS. That might explain the steady reading on bank 2. (Except I thought it deactivated both VANOS units.)

I too got the P0021 code after doing the guides in 2019. At first it was intermittent -- it would take a few days for the code to pop after resetting it. Then it got more frequent and annoying. I pulled the cover, marked the timing wheel location, moved the wheel 1/8 to retard the wheel slightly and...still got the code. I went back in, started to adjust it a little further, but then double checked it with the test hole. From that perspective it was already too far retarded. So I put it back toward what the test hole suggested, which nearly matched my mark. Then, not wanting to just repeat the same test, I turned it a little past my mark, advancing it slightly. I put it back together and voila, no more codes.

I know. I know. Some will say I just got the directions backward, between advancing and retarding. Except I didn't. I even had pictures to prove it, but I decided it wasn't worth it, to try to convince the Forum that advancing the trigger wheel further actually worked to kill that code. You can believe it if you want, and try it if you want. Up to you.

One thing I firmly believe, that others will probably agree with, is that the timing cover can be attached in slightly different positions, which can throw these things off. You said your cover was a little high, which means it's a little off. That means the test hole is a little off too, right? And you said it lines up with the trigger wheel, which means the trigger wheel might be a little off too.

If it were my car, I'd try different trigger wheel positions, probably after making sure the upper cover on that bank was located as well as I could get it.

(Actually, now that I think about it, I'd probably try adjusting that upper cover with the cam cover on. Loosen 1 or 2 bolts for the cover far enough to see which way it's biased. Loosen the other bolts slightly. Then push/pull the cover, or lever it, to position it slightly differently, then re-tighten, reset the code, and see what happens. Might not work but it's worth a try.)

[JimLev]

All those minor error stack up so as R Shaf said the trigger wheel isn’t set correct.
Once the DME sets a code for the vanos they are both disabled so it’s weird that your reading 41º. It can’t move that much.
I thought I had a screen shot but it was the Rough (LU something) I had a pic of.
I remember looking at it while reving the engine, it went from about 0º to maybe 12º.
I’ll be using INPA again tomorrow, I’ll get a pic.

[NolimitE39]

The Camshaft Values in INPA (top lines of Analog 3 Screen) show the actual state correct? More onto that weirdness is that the VANOS Angle Values match exactly the Camshaft Position Values. Below is a link to an image posted by BorderM3 in 2017, this looks exactly like my INPA screen.

https://www.bimmerforums.com/forum/a...2&d=1514479808

And here is the thread where it was posted:

https://www.bimmerforums.com/forum/s...11-p0021/page6

[JimLev]

Some pics from my tests this morning.
This first one is with the engine off and the key in pos 2, everything powered up.
INPA updates slow so it was hard to keep the RPM steady waiting for updated with one foot on the gas and the other one outside the car so I could snap some pics.

68893C7B-235B-4D20-85E8-6F1382CDF044.jpg

This is with the engine running after the Sec Air Pump (SAP) shuts off.
I was starting to rev it up a bit here.
A52DFEAA-EF98-4165-8962-9A28FC44C9C1.jpg

This is at 4K RPM.
6B845FAF-2249-4FC2-B2D5-B86B452493B3.jpg

Just curious, what’s your’s read?
3C1038FF-51FC-4877-AF01-51219AC3E40D.jpg

[NolimitE39]

I just got TestO working and I think I have the results I'm looking for. The singal from cam shaft sensor #2 was steady at 41.93 as before so I unplugged the sensor that's in the timing cover and plugged in an extra sensor I have and laid it on the coolant reservoir. Wouldn't you know it, on startup the reading was at 0 and stayed at 0 until I tested it by waving a wrench over the sensor at varying distances and speeds which varied the signal, the sensor is sensitive to 10ths or perhaps 100ths of a mm so a slight change in the sensor position (UTC position) may be all it takes. I'm going to go beat the UTC with a mallet (figuratively speaking) until its seated and retest. Will report back. God I don't want to retime this thing again. Definitely not certain of anythign at this point but at least this is a diagnostic step in some direction, is it the right direction? We shall see.

[JimLev]

Testo reads back info much faster than INPA. I did download testo a while ago but couldn’t get it working.
Are you sure your trigger wheel isn’t bent? There is only about maybe 1/8” from the end of he cam sensor to the trigger wheel.

[NolimitE39]

I just pulled the UTC and trigger wheel completely out and laid it down on my two work benches and the kitchen counter, its perfectly flat. So the only other possible culprit with the readings I'm getting is the UTC not being seated far enough. I've ran out of time today so will pick back up tomorrow. I took the week off between projects at work to try to get this and a slew of other things around the house done, I've only messed with this and done one thing with the house, needless to say I'm frustrated but as long as I haven't hit a complete dead end I'll continue trying to figure it out. I'm reading up on hall effect sensors now.

[R Shaffner]

There's still a chance that it's a wiring issue. (You swapped solenoids, as I recall, but there could be something wrong with the wiring, the connector, etc.) Still, I'd try the route your on of repositioning the UTC and the trigger wheel.

[NolimitE39]

You're right in the fact that I haven't tested the operation or signal to the VANOS solenoid that's on the bill for tomorrow if realigning/reseating everything doesn't work. All of the testing points to the sensor not sensing the shutter portion of the shutter wheel or rather sensing the constant interference from the inside solid portion of the shutter wheel and thus providing a constant signal materializing in a reading of 41.39. I wish I had a guide to all of the values so I could actually see the raw data coming out of the sensor. I know it exists I just haven't found the information on the internet yet. I ran the engine with the cam sensor unbolted and just stuck in the hole, I could get it to fluctuate by moving it in and out but it was obviously erratic since I cant maintain mechanical millisecond timing with my hands. So again to me it all points to the sensor picking up the inner portion of the wheel resultant from the UTC being too high. Before performing all this testing I would have never guessed that the tolerance is that tight. BMW could have made the wheel will a bit less meat in the middle and such issues wouldn't exist. As for the wiring, I did trace the wiring for the cam sensor and VANOS solenoid through the engine top boxes and all the way into the main loom that goes into the firewall. I would imagine any damage would be between the main loom and pigtail since that's the part that gets messed with most often.

[JimLev]

Speaking of wiring (thanks R Shaff) have you checked the voltage at the cam sensor?

Cam Connector.JPG

If you have access to a scope you could look at the signal from the cam sensor.
Dscn2571.jpg

This is at idle.
DSCN2637.JPG

This probably isn't your problem, just throwing it out there. The tu and non tu engines use different cam sensors. The tu's are longer.
Dscn2568.jpg

[NolimitE39]

Not yet, I've been trying to find the values. Jim, what else do you have, lay it on me. I love raw values, they tell me so much more than INPA or TESTO. I come from the Z32 (300ZX) world where everything was raw values or bust.

[JimLev]

I'll see what I can dig up tomorrow.

The car before the 540 was an '87 Z31 300ZX that I bought new, had it for 14 years.
I had 2 engines for it, the original got the royal treatment, ported and polished heads and intake, ISKY cams, headers, nitrous, chip, etc.
I had rebuilt it using all the stronger Z32 bearings and rings.

I eventually sold the car (after I bought the 540) with the hopped up engine, minus the nitrous, and sold the second engine to a guy in NJ.


300ZXpic7.JPG

DCP_1085.JPG

87_300ZXa.jpg

[NolimitE39]

That's a clean looking Z31. I have a TT that I built with all forged internals, bored over to 88mm IIRC and GT28RS BB turbos (hopped up disco potatoes). Its in the garage waiting for some attention once the E53 is back on the road.

[JimLev]

I always kept it clean. That pic was when I sold it.

Your vanos should read 10.5 ohms. It’s PWM controlled, the signal going to it will look similar to the cam sensor signal I posted above.

[NolimitE39]

So I pulled the bank 2 UTC and shutter wheel. Locked all 4 cams and crank and unbolted and retorqued the VANOS on bank 2 all of which went back together in exactly the same place that it came apart. I also pulled the VANOS solenoid backflow valve to confirm proper operation. Also retested the VANOS solenoid on the bench and 12V makes it move and it retracts on its own. I am positive that the timing is set correctly. I measured the 3 pins on each of the cam sensors which showed battery voltage on the power, continuity to ground on the ground and the sensor showed fluctuation between 5ish and 8ish volts (I had to use my Klein MM700 because I don't have an oscilloscope). Upon first startup TestO was showing both banks within acceptable ranges (-2 to +2) for bank 1 and (1 to 8ish) for bank 2 which leveled out after about 30 seconds to ~ +1.x on both banks (EUREKA I yelled). I shut the car off because I didn't have the cooling system filled. Turned ignition to on and went to read the codes. Guess what, code 34 is there again, time stamped with the current time, still showing 0 degrees. CEL came on upon the following startup and TestO again shows bank 2 cam sensor reading 41.39 though now when the car is shut off it goes back down to 0 and upon startup actually fluctuates some between -20 and +41.39. I traced the signal wire for both cam sensors into the DME, no breaks, no damage.

If the VANOS gear failed internally somehow (not sure what could break in there catastrophically) could it do this? The car idles very well, not like Bentley but it is a Dinan S3 so I wouldn't expect it to be buttery smooth (and it never was previously). I'm stumped completely stumped. Anyone, someone, PLEASE HELP!!!!!

EDIT: Both VANOS solenoids in the X5 (engine I'm working on) read 12.6 ohm, they are warm though so it could be contributing. Both solenoids in the 540i (currently parked awaiting guides) read 10.5 ohms. Same part number for both engines confirmed with RealOEM, guess Ill go swap them out and see if it does anything.

[JimLev]

I took the readings with my Klein MM1000.
Are you sure when you set the crank to TDC it wasn’t 180º off? I’ve seen it done more than once.
Bank 1 cam lobes tipped toward each other with the harmonic balancer set correctly and the #1 piston at the top is when you know it’s correct.
I also don’t think there is anything inside the vanos that would cause your issue.
I’ve got some pics of the vanos opened up on another computer, I’ll post them later.

[NolimitE39]

I am 100000000000% positive on the orientation. I did that when I initially disassembled the engine, realized it after I loosened all of the cam gears and dint want to take the chance on damaging valves so I had to flip the cams manually (remove and reinstall). The car runs very well, idles, boosts, has gobs of torque and no belt slip since I installed a griptech pulley on the SC. That's the biggest WTF for me, everything seems to be working well except that GD sensor which causes a CEL which wont let me pass inspection. Honestly if I could drive it i probably would and just keep my laptop running in the passenger seat to keep an eye on the vitals.

things just got more WTF.

I ran the car in the garage and watched the bank 1 and 2 cam sensor output. once the car gets over 2k rpm the signal goes from +41.93 to ~ -20, from there, the higher the RPM the closer to 0 is goes until it hits ~ -8 at around 4k RPM and stays there through the rest of the rev range. Any ideas?

[JimLev]

These are from a 540, your (4.6?) probably has a metal ring with pins, not plastic like shown in the last pic.
Very strange the vanos shows 41° as it can't turn that much.
Here's what inside the vanos if you cut it open.

image011.jpg

image012.jpg

image013.jpg

image014.jpg

image015.jpg

[JimLev]

More
VanosTiming.JPG
VanosTiming_1.JPG

[NolimitE39]

I'm currently working on my 4.6iS X5 but I also have a 540i6 which is sitting in the garage because the guides on it need replacement. So currently any parts I need I am borrowing from the 540i. What does the plastic inside the gear do? I've been trying to figure it out from the pics and descriptions on Beisan Systems, if this plastic breaks apart does it allow the piston to move more thereby allowing mover 20 degrees of adjustment?