Camshaft deviation wandering slightly?

Recently I replaced the bank 1-3 cylinder head of my 996 3.4 engine, because of an intermix issue (see https://rennlist.com/forums/996-foru...l#post20296153 ) In the process I timed both banks. Installed the engine again and drove it around 300km since. Idling was fine initially, but got rougher over time. So I started monitoring some parameters via obd. Unfortunately I did not take the camshaft deviation directly after installing the engine, but I took it a few days back �300km after the rebuild. When I checked, the bank 1-3 camshaft deviation was around 5.4 degrees. This was constant at that value with a running engine, but when I stopped and restarted the engine it went up to around 5.6, stayed there when running, but when restarted it again, it moved to 5.8. So for a few times, when I restarted it, it moved a few degrees. However, restarting after it was at 5.8, didn't move it anymore. (Bank 2 values I couldn't measure, but I have a European car, as I understood it they only have 1 crank sensor?) The other day I test drove the car to see what would happen and halfway through the ride while I pulled away from a traffic light, I saw the deviation move from 5.8 to 6.35. Stayed there for the rest of the drive and doesn't seem to be moving anymore now. Does not throw a CEL or error, but rough running value is 15.6 and idle jumps al little bit (not problematic, but enough for me to notice.) I'm kind of looking for hypotheses now. Anyone had something similar happen? What would you reckon could cause this? My current thinking/Hypotheses; Maybe the camshaft sprocket wasn't tight (tightened 4 bolts to 13nm, but perhaps the sprocket itself 'settled' after. Also did not loctite the bolts.) Validate this by pinning the crankshaft to time, remove camshaft plug from exhaust camshaft and checking alignment of the camshaft mark. (if things are off it's gonna be fun trying to time bank 1-3 with the engine in car...) Something with the variocam system not working right Validate this by removing the both camshaft plugs and visually checking the alignment of the cams IMS shaft sprocket is moving around on the shaft Don't consider this a likely option (something with occams razor.

The orings on the actuator have probably shreded and the piston is sticking. @SlakkerRacingDev sells a.rebuilt service for the actuators. If u r in europe slakker works closely with Hartech and maybe they do it too. The gear could be slipping too but thats pretty rare... unless assembly was not done well when the head was installed.

My money is on the camshaft sprocket tightness that you mention. Best of luck.

Quotes source post: 20367955

I'm from the Netherlands, so that might be an option. Just wondering, is there a way to test that actuator thingy without (completely) disassembling things?

Actuator is bad. It needs a replacement, expensive and hard to do. But you have done it once.

Quotes source post: 20368086

Can I ask what data points point to the actuator and if there's any way to test that without complete disassembly? If it makes any difference, pads have been replaced around 3.5kkm back. (Just want to form a good understanding of what happened here and what I can learn from it)

My uderstanding is that if the VarioCam is slow (usually actuator seals broken) you'll se the fault code P1341 (for bank 1). I have this and ordered seals kit from Hartech in UK - those guys actually put the seals kit together. I'll do the actuator seals and few other items to my 996 in January when I have time to pull the engine out. The kit is not expensive at all in comparison for purchasing a new actuators. Connect with Lee Jenkins at Hartech and he will sort you with a seals kit and instructions. Hartech video to change the seals is here: You Tube

Some scanners can read the "actual cam angle" this changes when the engine revs over 3k(ish) forget the actual number. U can graph it and see if its lazy but since u r row u cant compare the 2 sides. The orings chewing themselves up is a very common issue and u may not have done anything wrong. I had this issue with bank 1 2k miles after i did pads. I suspect pulling and messing with the actuator during the pad replacent increases the probability of an issue. No way to really test as it all sits under the valve cover. If u pull the sump pan i suspect u will find green oring bits at the bottom.

Quotes source post: 20368604

Sounds likely, I have definetly fiddled with the actuator. Didn't have the proper tool, so had to press and lock it manually. Might have upset something g. Wasn't aware that it's full of o rings, otherwise I'd have rebuilt it when I was in there.... Well, you live and learn. Just hate having to go so deep in the engine without a 100% chance of Succes. I have a icarsoft v3 which does track the actual timing and can do logging. I'll check it it also has a 'desired cam angle' or something, a delta between the two should be a good indication of things being wrong. If that doesn't work I think I can also actuate the Solenoid with the icarsoft. See what happens when I do that, hope it gets the odds in my favor. ​ Ill fiddle around with that later and will report back.

Once you go in deep, it's not economical to cut corners. This is why when I rebuilt a 5-chain engine for a neighbor, I bought two brand new actuators from Porsche. It was not cheap, but cheaper than doing it twice.

Test drove today and graphed the actual timing. From what I understood a sticky actuator would lead to sluggish behavior in this graph, but the response seems pretty direct? I will first remove the plugs and check the timing. My current hypothesis is still that the sprocket is moving a bit. Could be wrong, but it's the easiest thing to check before going deeper.

The time scale of that graph is too long. Try a scanner like durametric where you can actuate the Variocam just once, and then watch the cam variation response.

Took the plug off and took a picture. Timing definetly seems off. @yelcab, I had the device in view while driving. Response was instant.