Members | Sign In
ScannerDanner Forum > Diagnostic tools and techniques
avatar

Voltage relative compression

posted Apr 10, 2016 17:46:30 by Tyler
I've seen relative compression testing using voltage in trade magazines for awhile, but was reminded of it recently when Andy MacFadyen brought it up in another thread. I normally go straight to the amp probe for this test, but I can think of a few times when getting the probe in place was problematic. Why not experiment a bit with a new technique?

2000 Honda CR-V comes in with a flashing MIL and a noticeable miss at idle. Cranking the engine produces that noticeable 'skip' in the cranking cadence. Ruh roh, this won't end well.

Andy describes using voltage drop across the starter positive cable in his method, which is totally fine. In this example, I opted to try connecting the scope to the battery directly, then switching to AC coupling. I've seen enough starter voltage waveforms to know that I should see something here. Throw in an ignition trace on #1, crank it, and we get this:



Kinda a mess, I know, and I should have rescaled the ignition trace. Let's zoom in a bit:



Okay, now I can make something of this. The firing of #1 lines up too perfectly with the first positive voltage tower, so I'm associating higher voltage with more compression. Then we have a short tower, a strong one, and a meh one before starting over. I'm ready to say we have at least one weak cylinder.

The firing order is 1-3-4-2, which makes the weak cylinder #3. Going in cylinder backs up the relative compression results. Yay!



Nooooope, 70ish PSI does not cut it. Here's the same cylinder running:



The asymetrical tower, the low vacuum in the decompression pocket, and the exhaust valve opening happening too soon says leaking exhaust valve.

The static compression numbers were #1 at 140, #2 at 105, #3 at 75, and #4 at 140. All in all, the compression towers in the relative compression waveform corresponded nicely.

Thoughts/comments/complaints? Post 'em!
[Last edited Apr 11, 2016 01:00:09]
page   1
12 replies
avatar
Rob Longoria said Apr 11, 2016 01:47:17
where exactly are your test leads. I am assuming on the B+ at the battery to the b+ to the starter.
avatar
Tyler said Apr 11, 2016 02:58:18
I actually didn't go down to the starter for this one, just had one channel at battery positive, and an ignition sync. Scope ground is on battery negative.
avatar
Noah said Apr 12, 2016 12:22:39
You know, all the time I spend fiddling with my amp clamp, I could finish the whole test. I've seen this done before, and showed my friend how to do it when he bought a modis. I wonder if there's any more definition measuring amperage opposed to voltage. Probably not as they are both directly proportional.
What is the purpose of using AC coupling?
[Last edited Apr 12, 2016 12:23:22]
Massachusetts, USA
avatar
AndyMacFadyen said Apr 13, 2016 18:39:07
I just took a cranking volts compression capture connected via the cigar lighter socket -- the result a little noisey but it worked.



So I tried again with a low pass filter

[Last edited Apr 13, 2016 19:20:51]
"Rust never sleeps"
avatar
Tyler said Apr 14, 2016 00:17:28
You know, all the time I spend fiddling with my amp clamp, I could finish the whole test.


Right? Some cars, it's dead simple to get the clamp in place. Others, makes you reconsider your diagnostic direction!

I've seen this done before, and showed my friend how to do it when he bought a modis. I wonder if there's any more definition measuring amperage opposed to voltage. Probably not as they are both directly proportional.


I dunno, I should do a side-by-side comparison with an amp probe, see which one looks better. Honestly, after this diagnosis, I think I prefer the voltage method. Faster to setup, just as accurate.

I just took a cranking volts compression capture connected via the cigar lighter socket -- the result a little noisey but it worked.


Very nice, sir! Many thanks to you personally, otherwise I wouldn't have started experimenting with this method.
avatar
Tyler said Apr 14, 2016 00:19:44
Oh yeah, got the head off this one yesterday, found a big chunk missing out of one of the #3 exhaust valves. Someone's getting a valve job!
avatar
sam67 said Apr 14, 2016 03:42:40
Hey Andy ,I have the hantek scope plus low amp probe .You were right about the amp clamp not fitting round cable .Would like to try a voltage compression test to see if i can get a decent waveform .Any tips would be appreciated on where i can put my leads for a decent trace ?

Managed to get some good injector and coil waveforms ..Just trying to get to grips with the scope.
Thanks
avatar
AndyMacFadyen said Apr 14, 2016 07:23:11
The Hantek 1008 is great value for money but the software is not great and the hardware is cheap and has limtitations. One of the bigest hardware limitations is it has no AC coupling which is handy for this particular test. I built an external AC coupling filter with just a couple of capacitors to block the dc component.
It goes without say to turn any channels you are not using off.
For some unkown reason I found the best way to get a decent looking trace on the Hantek is to load one of the pre-programed tests and modify the time base and voltage too suit.
[Last edited Apr 14, 2016 17:58:10]
"Rust never sleeps"
avatar
sam67 said Apr 15, 2016 23:04:44
Cheers Andy,yeah i use the pre loaded software and just play with the settings .I have all the pico waveforms saved as pdfs in a folder on my laptop and phone and use some of the settings from the pico setups as a guide for a justing the hantek and it gives me a better guide doing it this way.
avatar
juergen scholl said May 13, 2016 14:02:03
Tyler,

the stress on the battery while a cylinder is on compression stroke will LOWER the voltage. Therfor the compression peak will be a downward peak on the voltage trace, iaw it located in the valley of the signal, not on top.
Voltage ist just opposite to amperage in this case.When amperage goes high => voltage will go down. To interprete your voltage measurement the same way you do with your amp reading you have to invert the picture....

To proof this concept you can just take out a spark plug and take a voltage reading. Even better, hook up two channels and read amps and volts at the same time....
avatar
Tyler said May 14, 2016 21:50:59
Hey Juergen, thanks for your comments!

the stress on the battery while a cylinder is on compression stroke will LOWER the voltage.


I agree with this 100%.

Therfor the compression peak will be a downward peak on the voltage trace, iaw it located in the valley of the signal, not on top.
Voltage ist just opposite to amperage in this case.When amperage goes high => voltage will go down. To interprete your voltage measurement the same way you do with your amp reading you have to invert the picture....


This is true! It we wanted correctly relate voltage and current during this test, then the voltage trace SHOULD be inverted. Then, downward peaks would represent cylinder compression.

My counterpoint would be that I don't believe it would change the results of the test. I decided to take your advice and test this on my vehicle, with a spark plug out of one cylinder. FYI, this is a coil-on plug design as opposed to a distributor in the original captures, so the sync trace is the #2 coil trigger signal.





As you can see, the low cylinder is clear either way, so I think it just comes down to personal preference. I'm familiar with amperage relative compression testing, so it's natural for me to associate compression with positive peaks. Inverting the trace may make more sense to you, and that works just as well.
[Last edited May 14, 2016 21:51:29]
avatar
juergen scholl said May 15, 2016 19:40:05
Hi Tyler,

thank you for your time and work...

I totally agree with you that it comes down to personal preferences. As you demonstrated the fault condition - low cylinder pressure -shows up evidently in both ways to view the voltage trace, either inverted or not...

The reason I made my first comment is the following:

You stated in your original post that you connected the scope to the battery directly, then switched to AC coupling. The #1 firing lined up perfectly with the first positive voltage tower. But this means you had to invert or flip over the original voltage trace, otherwise the #1 firing was to line up with the downward spike, iaw the valley. This being true can clearly be seen in the two pictures in your last post, observing where the trigger signal is taking place with regards to peaks and valleys in the original and inverted picture.

Now, I didn't read in your o.p. about inverting the signal (and to me it looks like you inverted it) and thought people starting with this technique of using the voltage trace may not be aware of it. This might cause some confusion when they see their timing event at a valley - low voltage level - whereas your example shows a different picture.

This was all my comment was about...

Juergen
Login below to reply: