I have a Hantek 8 channel. It works.

Interestingly, the crank channel looked pretty good on initial start up but deteriorated within a couple minutes. Might be looking at a pending thermal failure. Also interestingly, when my original sensor failed, I had at least 6 total shutdowns before I ever received a code. Getting to be an OBDII reader is about as obsolete as an 8-track player.

I'll get a Mopar sensor and generate some normal waveforms. Might as well get another AZ sensor since it's free and see how it performs out of the box.

Don't mean to jack your thread but I'm sure 97G is anxious for another chapter to read.

 

No sir, your not jacking the thread at all, we are imploring an issue and I appreciate every bit of your contributions. Thank you

After all the threads I've read, sites, forums..... this is the first that you and awg have covered some much needed food for thought, ideas, because not every diagnosis is cut and dry. Please by all means post your newfound wave forms. This is a good collection of knowledge and I find it highly useful.



You humbly say 8 channel, that makes me chuckle as 8 channels and it's capable to say the least. Lol

Awg strongly recommends having a trained doagnostic technician scan my jeep, however with both your knowledge I haven't heard weather you bekieve I can achieve enough information with a 2 channel cheaper scope, or is this a waste of money?

 

I really can't speak to that. Labor cost around here is $102 an hour, there's no guarantee they'll find anything, and let's face it, this is a problem that from a practical point of view, doesn't need to be fixed. You'd have a better shot if you found somebody with old Mopar experience, as XJ CPS are the weirdest thing I've ever seen in my life (and getting weirder). It's like they have their own personality.

IMO, one's diagnostic scanner needs to have at least number of cylinders of the vehicle you're working on plus two. I've had horrible gas mileage and had toyed with the idea of replacing injectors (one is physically broken, but hey, the thing starts and gets me where I'm going)(and I'm "frugal")(and no I didn't give my XJ a woman's name)(it's like, "The MF Jeep") but now I may wire up the injectors and see what they're doing while this is going on. The CPS connector is in a vulnerable place so maybe I just have a bad contact someplace so need to jiggle a few more wires (but that wouldn't carry the same amount of irony - you with a code and no problem, me with a problem and no code).

I did comment on that:

cause that one you were looking at didn't seem to have any leads helpful for vehicle testing.

 

Interesting you have a Hantek, I am thinking of getting one, it seems minimum entry level for an automotive scope

did you find the learning curve difficult?..(which is why I suggested the OP find a tech, instead of buying a scope and learning himself)

As you have kindly provided that info & snapshot to help him move forward, if I was the OP, I would probably bite the bullet for a Hantek

which if you want to go to "next level" fault finding, you cant really buy a toy, it needs to be a fit for purpose tool

think the full Hantek kit is under $US200 ?

 

Did a voltage check on all the injectors, and they're fine. Representative waveform:



Seems pretty clear that CPS is not solely responsible for their activity.

Need to look at timing stability.

 

As noted, IMO for channels you need cylinders plus two. For the rare instance you want to look at all cylinders simultaneously. Although since you're usually looking for a dead something, in practice you just go down the line and probe. I can just see 8 cables caught up in the fan blade.

I'll never own another 8 cylinder, so I'm good.

The biggest complaint I've heard about Hantek is getting the software installed. Apparently some people don't know how drivers work.

No but I got skilz.

Yeah, I'm not going to push it. There's money and time involved, and if the other diagnostic people knew what they were doing the OP may not find anything out that he doesn't already know.

That said, I'm looking at my CPS and saying "Just WTF is it that you actually DO!?"

And probably need to be able to use it more than once.

Yup.

 

97G has his own problems now, but yeah, that CKPS signal looks a bit hosed. Thanks for posting.

 

Timing looks a little choppy from ~4 - 11 degrees. We'll see what a new CPS does to it.

 

have you read this one I just stumbled across ?

https://www.pirate4x4.com/forum/jeep...-hate-xjs.html

note the picture of the flexplate with slight ding
I do not recall you reporting back on my proposed flexplate examination,


not sure myself how easy it is to do, but being one of very few common parts, there is really no option imo but to exclude this possibility

(though it was not the issue in the case study shown)

 

Well, 2 things to close it out.

I got nice waveforms with the new CPS:



And I am a big stoopid.

 

I did see daves Hantek for 100+, might be worth my looking into pending software cost and I need a computer.

 

I'm gonna stop short of saying "Boy you really need to out and buy this!!" tho. It has a few shortcomings (like it's difficult to measure stuff easily, and all those missing spikes were a software or acquisition problem which I need to explore); you have to ask yourself how many times are you actually going to use it so can the cost be justified; and it'll take a bit to get the hang of understanding stuff (I've got a lot of background in waveform analysis so I don't really have a "learning curve"). OTOH, it's cheap, and in the above example it can tell you everything you could possibly need to analyze signals. And I guess a lot of the time you're interested in basically if you got a waveform or not.

 

Good stuff.
How do you manage to get those screen shots though, without showing your thumbs or background in the picture?
just curious

 

The Hantek "oscilloscope" is just a signal generator. The software is Windows based. Hantek connects to the laptop via USB. Hit printsceen, edit with Irfanview, post.

 

Well, turns out there's no "problem", that's simply the limitations of the device.

If you look at the specifications of the device:

2.4 mega-samples per second should easily and reliably collect a signal of 200Hz. By my calculations, a CPS running at idle is pulsing at about 225Hz so this thing should have put up a quality waveform of the CPS easily.

Welllllll, there's that little disclaimer there that says "Single Channel". Turns out low end oscilloscopes often utilize a pile of converters to generate a signal. One converter will collect for a bit, then another one will collect while the first processes data, and so on ("interleaving"). If you're only running a single channel you get to use all the converters, but if you add more channels, less converters are available, and if you're trying to collect a very busy channel some things get distorted or are missing altogether ("aliasing").

Testing this on the 1008C, I put up 2 CPS waveforms on a 50ms scale and got



However, the act of simply adding another channel caused aliasing to appear:



There are several ways to correct aliasing, but it appears the only way to do that on the 1008C is by reducing the timescale (NBD)(which is what was done on the initially-posted waveforms)(which is what I should have done before getting a new CPS and changing it out)(which is why I am a big stoopid).

Parts cannon fires again...