When is a bad cat and loss of power not a bad cat?
A couple of months ago I had this Jetta come to me from another shop. This is the information that I had gotten on it as written on the ro, sorry about the caps, but it's how it is in our system! "CUSTOMERS CONCERN IS THAT THE CAR IS NOT RUNNING WELL AND HAS CHECK ENGINE AND EPC LIGHTS ON, THIS ISSUE STARTED 3 MONTHS AGO AND THE SHOP THAT HAD THE CAR SAID IT NEEDED A NEW PEDAL AND WAS REPLACED WITH A USED ONE AND THEY WERE CERTAIN THAT WAS THE PROBLEM... THEY ALSO REPLACED THE MAF AND ECU AFTER THE CAR STILL WOULD NOT RUN RIGHT. HE SAID SOMETHING ABOUT THE ECU HAVING A DIFFERENT NUMBER THEN THE ORIGINAL TOO ."
Well I found the APP was stuck at 99% when actually at idle, and they had already fired the parts cannon at it. However I know they didn't try a new ECM because these require the IMMO to be adapted to get it to start, and if they would have had the tools to do so, they would have fixed the car! Long story short, I fixed the APP issue with an ECM and it still had a lack of power, and that's where this story starts.
The car sounded like it had a bad cat. You know that deep tinny rumble coupled with the lack of engine breathing type sound. So I did a compression test with my WPS, and then I did a running compression snap throttle test on the same cylinder.
200 psi normal compression, and then 260 psi during the snap test with about 25 psi backpressure, engine can't breathe. I had a p0420 along with other faults and positive 10% on my multiplicative fuel trim.
The STFT was pegging lean during acceleration, but the fuel pump tested text book perfect. This should have been a slam dunk bad cat.
But it wasn't a bad cat... They came back today to make sure it was a bad cat before they spent the money (that nearly equaled the value of the car) on a new cat. So I lifted it up and it had an aftermarket cat welded in place with the rear O2 sensor pointing straight up and damaged because it was bottoming out on the road. No way it couldn't be a bad cat...
I was going to test backpressure manually with my analog gauge before and after the cat just in case there was something clogging it post cat. So I pull the rear O2 sensor and the cat is empty... Well, there's my p0420! Now where's all that backpressure I saw? I reinstall the rear O2 and pull the front and do my test. No backpressure at all. Nothing.
So I'm thinking in my head what else causes these symptoms, and bad cam timing will do it, but there were no correlation faults. It's an easy check anyway, so I pull the cam covers and find this.
Three teeth advanced on the camshaft! Yeah that will do it, but no cam correlation faults! I've had these cars set correlation faults half a tooth out before... So I reset the timing and retested snap accel with the WPS, all good. Test drive said it was just like any other 2.0 that I've driven.
I never heard when the timing belt was last done, or if they messed with it at the other shop the three months before I first got the car. It sounded to me like the loss of power just happened (like a clogged cat), and wasn't always there. But for the $220 or so he paid this time, I'm sure he'll be happy with the significant increase in power. Hopefully he'll spend what he saved on front end work because it rattled on every bump!
Thanks for sharing, and good job on going the extra steps
Thank you for taking the time to diagnose this properly and for taking the time to write this up and sharing with us. You had logical thoughts and worked through them instead of replacing parts based off assumptions and statistics.
I wouldn't expect anything different from you but wanted to share my appreciation publicly.
What is weird is when you did in cylinder pressuere transducer and spark, it is on tdc? I would expect to show degrees retarded, probably if you had keep the spark channel during snap may ha e catch it there .
I'm glad you were able to solve the problem but I think you could have saved yourself some time by putting cursors on your idle waveform and measuring the valve events.
A cam that's off by 3 teeth will be very easy to see when cursors are applied. Instructors like Bernie Thompson and John Thornton will tell you to always use cursors or overlays when trying to decipher a pressure waveform.
In reading over your well written post, a few things jump out at me.
1...I had a p0420 along with other faults and positive 10% on my multiplicative fuel trim.
A restricted exhaust will not cause a shift in fuel trims, since the are flowing through the engine is being metered by the MAF.
2.. The STFT was pegging lean during acceleration, but the fuel pump tested text book perfect. This should have been a slam dunk bad cat.
here again, the trims are not pointing towards a restricted exhaust, they are pointing elsewhere.
3. the incylinder waveform you posted, with the little pressure spike at the end of the exhaust stroke,,,,,,, that little pressure spike is pointing directly to a cam timing problem. Now, if you had put some cursors on the waveform, life would have been a lot easier.
Good job on the diag!!
Thought you might like seeing your running capture measured.
BTW. I wanted to mention, the most common way for the cam timing to get off leaving the camshaft advanced of the crankshaft is for someone to rotate the engine by hand, backwards.
Yup, I definitely could have done it different and had a better conclusion the first time. I made some assumptions the first time the car was here, and have a really hard time getting any real information from my service advisors. Most of the time they just say "here's this car that runs bad, what's wrong?"! I'm still learning in cylinder waveforms and haven't been fortunate enough to have any classes on it. I'm taking some at Vision next year though! It's extremely hard to get any training in Utah, especially when I'm paying for it and the tools myself.
I appreciate the feedback and am learning from all of it!
It would sure e nice if we all were perfect, but in the mean time, we learn by our mistakes, well, some of us do. I once heard, to error is human, but when the erasure on the pencil wears out before the pencil, we are erroring too much :) Thanks for putting your thoughts up so the rest of us can learn.
I bought a class on a flash drive from AES Wave that helped me a lot, but I’m still a baby in it aswell
With the tools and techniques you're currently using you are already far ahead of the average tech out there. It's a constant learning process and the path is not always straight and narrow. The main point I wanted to add to this discussion is the value of using cursors and overlays on your pressure and vacuum waveforms. You need those to really make sense out of the data you have captured. Sometimes there are things that stick out without them but most times the overlays will help to accurately decipher the waveforms.
Keep up the great work!
what does "The STFT was pegging lean " mean?. Engines can run lean. O2's can say the engine is lean. Fuel trims are positive or negative.
I have a good reference on where these engines run with regards to fuel trim and I know that bad fuel pumps are common. When I test drove it, the O2 sensors showed positive 25% during acceleration. The multiplicative fuel trim (part throttle learning) was positive 10%. On this car, I know that's bad. Either the engine isn't getting enough fuel, or it's not seeing the correct amount of air pass through the MAF; both issues are pretty common actually. In hindsight, I could have added a VE test as well, but I could have got skewed numbers due to the cam timing. But I did check the pump and it was perfect waveform wise, which has been an accurate health indicator for these fuel pumps. I have plenty of captures of bad fuel pumps to reference!
My main concern about what I went through, is this car didn't set a cam/crank correlation fault. It's the first one that I've ever see NOT set that fault and have the timing off. If I had seen that fault in my list, timing would have been checked visually before I did anything else. I've pretty much exclusively worked on VW/Audi since this car was built, and it's the first one I've seen not report a correlation issue.
Great case study. Thanks for sharing!
Interesting that the relative compression with sync didn't show the timing as being off.