The search for overlap.

Ask for and share advice on using the PicoScope kit to fix vehicles here.

The search for overlap.

Postby Fat Freddy » Sat Mar 28, 2015 10:35 pm

PART 1

This is basically a continuation of hexibot43 excellent thread " topic10159.html " But decided it was did not fit totally within the question raised by hexibot43

It is based on a point/question Steve raised and many others have. - What can you see in the exhaust pulse, valve overlap etc.. And too be honest I have spent many hours trying to crack the code to decipher these wave forms.
Not being one to take for granted that something is gospel because Mr X the guru say's so, and also it's good to question.

This is not the most exciting post, but plenty of pictures of wiggly lines :D but I hope it gives a bit more insight to what an exhaust does and the possibility of seeing valve overlap and EGR etc.

Any way here's a couple of images of an Alfa Romeo 147 twin spark 2.0L. They are done using a WPS and a $5 home made FLS not the $400 real deal.
The first image is the normal running compression with the homemade FLS up the exhaust which pretty much all have seen. I have used this image as this is one we often see and can relate to - Plug disconnected, look what the FLS does and how it reacts to exhaust pressure rate of change.

Image 1.png
Alfa Romeo 2001 2L TS A - Waste number 4 to 1 - WPS C - BLS Exhaust D - Crank Running comp


For Steve II.psdata
(1.93 MiB) Downloaded 87 times


The second is with the WPS in cyl 1 plug hole but have the second plug for this twin spark system. This time we can what actually happens in the cylinder in relation to an unadulterated exhaust pulse. Here is where we should be able to see valve overlap.

Image 2.png
Alfa Romeo 2001 2L TS A - Waste number 4 to 1 - WPS C - BLS Exhaust D - Crank Firing comp


For Steve.psdata
(3.7 MiB) Downloaded 73 times


The following image was taken on a '98 Vectra 2.0L. Here the home made FLS is beside the WPS in the exhaust. I have also included Steve's Mazda as a comparison. Note the Vectra pulse is a lot smoother on both the FLS and the WPS compared to the Mazda or the Alfa. But we also get to see a similar reaction time to Steve's real FLS with a greater voltage generated. My personal thoughts are this shows the 5$ FLS is capable of recording sensitive pressure change. It can detect the change in fuel pressure regulator as injectors open as seen here - topic10429.html

Image 3.png




Mazda_6_20150326-0001.png


MAZDA WPS AND FIRST LOOK.psdata
(810.4 KiB) Downloaded 73 times



The next image is an indicator of reaction to a hard miss on this Vectra engine

Image 4.png




Next part I want to look at valve overlap indicators in another way. Can they be seen in the above files? :twisted:

Cheers
FF
Last edited by Fat Freddy on Wed Apr 01, 2015 8:39 am, edited 2 times in total.
User avatar
Fat Freddy
Zen Master
Zen Master
 
Posts: 267
Joined: Fri May 04, 2012 4:52 am

Re: The search for overlap.

Postby Steve Smith » Tue Mar 31, 2015 1:58 pm

Great work FF and thank you for taking the time out to post and share all this information.

I was intrigued by image 2 where you have the WPS connected in cylinder, running at idle with the remaining spark plug still installed and combustion taking place. Can you confirm this is correct? (Nice one by the way)

Could you post the psdata file for image 1 and 2 as image 2 is a rare look into the cylinder with combustion?

I am intrigued because of the complete absence of the expansion pocket at the base of the power stroke.

I do remember a number of queries surrounding the response time of the WPS when compared to FLS in the exhaust system.

I have looked into this and have some product information that will help us all, particularly when evaluating critical events in relation to crank and cam position such as overlap

Speaking to the designer of the WPS 500 pressure transducer, when using the device on Range 3 a low pass filter is applied (inside the WPS) which adds some integration to the signal to maintain accuracy of static and low speed qualities of low pressure pulsations.

This will enable these pressures to be displayed with excellent resolution and shape but at the expense of bandwidth. The low pass filter associated with Range 3 therefore adds a delay of ~1ms in response time which is approximately what users may have identified.

Piezo pressure transducers utilise a piezo crystal that responds instantly to pressure transitions but returns rapidly to zero whilst also responding to noise which then generates spikes and can make the waveform look very active adding to its complexity. (Exhaust gas pulsations on a V8 for example)

If we therefore wish to view a pressure pulsation at full speed (10 kHz) we have either Range 1 or Range 2. (The in cylinder pressure measurement is not subject to the internal low pass filer and so users do not have to worry about delay in response time)

To look at exhaust gas pulsations is real time (no delay) Use Range 2 and AC couple using the PicoScope software (Do not use the zoom on the device)

If you wish to see the complete waveform (not losing the low frequency and DC component of the signal) use Range 2 DC coupled and use either the scaling feature or the Zoom features of PicoScope (Do not use the zoom on the device)

Now we have zoom, scaling, AC coupling, filtering and maths channels in PicoScope software all the extra modes of the WPS are for use with scopes without such valuable features .

I hope this helps

Take care……Steve
Steve Smith
Pico Staff Member
Pico Staff Member
 
Posts: 442
Joined: Sun Aug 25, 2013 7:22 am

Re: The search for overlap.

Postby Fat Freddy » Wed Apr 01, 2015 11:20 am

Thanks Steve.

I have added the files you asked for (liked the save option - in view. They were originally too large to post. Great feature)

I was intrigued by image 2 where you have the WPS connected in cylinder, running at idle with the remaining spark plug still installed and combustion taking place. Can you confirm this is correct? (Nice one by the way)


Yes, you are correct. Combustion is occurring. There are a few examples of this on Autonerdz. I think the first was done with a PV350 near 10 years ago IIRC. So can't take too much credit. :)

On a side note (and again IIRC) - Pico did comment on using the WPS in measuring running combustion (a diesel that particular time) gas temps can cause problems. So I wouldn't advise all to jump out and do this.

Here's another image you may like.

Combined.png


(Unfortunately the psdata file for the above image is over 12 MB. But what I did was save the cranking WPS and Crank sensor's as reference files and added them both to the firing compression. Using a delay on both of 550us I can nearly get them to line up. There is only 10RPM - about 1% or 7.5° over the cycle - difference between the two captures. A nice comparison to see valve movement)


Your comments actually fit in nicely with what I found in further testing. I shall explain more in part II once I have put it together.

HTH
FF
User avatar
Fat Freddy
Zen Master
Zen Master
 
Posts: 267
Joined: Fri May 04, 2012 4:52 am

Re: The search for overlap.

Postby hexibot43 » Wed Apr 08, 2015 7:06 am

FF,
I must say that I am humbled by everyone else here. I wish I had more time to keep up with everything.

How did you get all those wave forms up at once? You can set delays for individual wave forms? I counted six wave forms. Keep up the good work. I'm enjoying all of it.

MAB
User avatar
hexibot43
Advanced User
Advanced User
 
Posts: 87
Joined: Tue May 10, 2011 5:27 pm
Location: Redwood City, Californa, USA

Re: The search for overlap.

Postby Fat Freddy » Wed Apr 08, 2015 10:49 am

Hi

How did you get all those wave forms up at once? You can set delays for individual wave forms? I counted six wave forms. Keep up the good work. I'm enjoying all of it.


It's easier than you think. If using waveforms from the same engine under the same conditions the RPM should be fairly close (RPM was 10 out in the above capture even though one was the running comp, was misfiring). I use the crank sensor output as a 'clock'. It's not perfect over the cycle but does give a good indication of the events.


So on one capture I have crank and running compression and on a second capture I have crank and firing compression.

Next I make reference waveforms of the first captures crank and running compression. I then open the second capture and add the newly made reference waveforms.

Next I measure the difference between the crank waveform and the reference crank waveform. This figure is then added to the delay box of the reference waveforms and 'Hey Presto!" It should line up.
If the waveforms need moving back use the -'ve symbol.

I hope that makes sense. Really quite easy. But looks cool. :D

Delay Referance.png


Hopefully when I get part two together soon it gives you more thought on that question you raised way back.
One capture I want to use is not as expected, repetitively. So I need to recheck with WPS. Unfortunately a heater tap letting go on the freeway has it on hold at the moment. :twisted:
But it will be done. Thanks for the feed back.

Cheers
FF
User avatar
Fat Freddy
Zen Master
Zen Master
 
Posts: 267
Joined: Fri May 04, 2012 4:52 am

Re: The search for overlap.

Postby Steve Smith » Mon Apr 13, 2015 10:20 am

Thanks again FF for posting here and for psdata files (Bug is now reported)

What has amazed me about the WPS running compression waveform (Combustion present) is the complete loss of the expansion pocket at the base of the power stroke.

If we see an event such as this during a running compression test (No combustion) we could assume the exhaust valve to be opening early (advanced)

We know in this case it cannot be relevant as the running compression (No combustion) has a perfect expansion pocket.

I wonder if this could be as a result of the increased pressure inside the cylinder due to combustion?

What is interesting is that we do not appear to have an increase in peak cylinder pressure after ignition as I would have imagined. (Diesel springs to mind)

Its all very good stuff FF as I have never been able to experiment like this before with such engines

Take care......Steve
Steve Smith
Pico Staff Member
Pico Staff Member
 
Posts: 442
Joined: Sun Aug 25, 2013 7:22 am

Re: The search for overlap.

Postby Fat Freddy » Wed Apr 15, 2015 11:32 am

What has amazed me about the WPS running compression waveform (Combustion present) is the complete loss of the expansion pocket at the base of the power stroke.

If we see an event such as this during a running compression test (No combustion) we could assume the exhaust valve to be opening early (advanced)

We know in this case it cannot be relevant as the running compression (No combustion) has a perfect expansion pocket.

I wonder if this could be as a result of the increased pressure inside the cylinder due to combustion?


It definitely is a cool waveform, an interesting waveform when you first see it. The lack of pocket is certainly caused by the combustion. My thought is, it's more by design. If you didn't burn enough fuel so that the gases didn't expand enough to eliminate the pocket, the engine would be running inefficiently by design on two fronts.

i - Since for a pocket to exist the pressure has to drop below exhaust pressure which is generally 0 psi (gauge) there fore for every psi below atmospheric the pocket went, you would have that force acting on the bottom of the piston, along with the already pumping losses going on in the crankcase anyway. Opposing the work the engine had just done. There fore more fuel would have to be burnt to pay for that loss.

ii - The second loss would occur when the exhaust valve opened. Exhaust gas would come rushing in to fill that pocket. So the engine would end up pumping out more exhaust gas than it created. Another inefficiency that would require more fuel to overcome.

Hope that makes sense.

What is interesting is that we do not appear to have an increase in peak cylinder pressure after ignition as I would have imagined. (Diesel springs to mind)


http://youtu.be/sEf8va1S7Sw

Definitely. I had idea's of a huge increase in pressure. But both the video and capture seem to show a longer more controlled burn.
I guess we couldn't expect to much of a massive increase in pressure as we only want to introduce enough energy to overcome frictional and pumping losses of idle speed.

What I find interesting is how much higher the non combustion peak pressure is compared to when combustion is occurring. This would be down to engine load I suspect. Inlet manifold pressure etc.
What I should of done was one with combustion occurring but introduced a miss on another cylinder. I would suspect the pressure's would be higher again.

Not trying to tell you how to suck eggs and that. Just my thoughts on the waveform.

(Side note - I think I have over lap in the bag. Just got to redo everything and some.
Is that "I told you so's) I hear. :) :?

HTH
FF
User avatar
Fat Freddy
Zen Master
Zen Master
 
Posts: 267
Joined: Fri May 04, 2012 4:52 am

Re: The search for overlap.

Postby Steve Smith » Thu Apr 16, 2015 8:10 am

Hi FF and thank you again.

Reading your explanation I can see this in my minds eye in relation to activity inside and outside the cylinder.

This is an incredible insight into activity within the cylinder (with combustion) and a future application note for sure.

Would you have any objections to a summary of this post (when complete) in the form of a topic for the newsletter?

Not sure when, but I think the information and theories here are invaluable and should be shared.

With reference to

What I find interesting is how much higher the non combustion peak pressure is compared to when combustion is occurring.


Based on the explanations above, the video and the fact the demand on intake air will be increased during combustion I can see how how supply cannot be met with demand given the increase in speed of the piston during combustion.

In this scenario its not surprising the peak pressure is slightly lower during combustion that the running compression test (no combustion)

When I first saw this I was amazed at just how low the peak pressure was during idle speed when compared to cranking and WOT. If we were to measure the compression of our engine and found the peak to be 4 bar, we would condemn that cylinder.

In reality 4 bar is normal at idle speed due to the demand of the cylinder for intake air, but the restriction to airflow presented by the throttle butterfly prevents the engine from breathing.
We can only compress the volume air arriving in the cylinder during the intake stroke, hence, low volume of air in = lower compression.

Open the throttle and "happy days". Maximum volume of air in = maximum compression

During cranking, although the throttle will be partially closed, we do see a typical cranking peak compression value of around 15 bar due to cranking speed (250 rpm approx)
Here the reduction in piston speed means air intake demand can be met during the intake stroke (even with a partaily closed throttle).

However as soon as the engine starts, we see the demand for air increase from 250 to 850+ rpm, the throttle partially closed and peak compression decrease.

I have include an image below that I hope helps with the explanations above. I was not aware of the behaviour of peak cylinder pressures through the different engine run stages until using the WPS, what a revelation.

WPS IN CYLINDER.jpg
WPS IN CYLINDER


Thanks again FF I know (like me) there are many out there looking forward to your next post.

All the best and thank you.

Take care.......Steve
Steve Smith
Pico Staff Member
Pico Staff Member
 
Posts: 442
Joined: Sun Aug 25, 2013 7:22 am

Re: The search for overlap.

Postby hexibot43 » Sun Apr 19, 2015 12:30 am

Ditto!
User avatar
hexibot43
Advanced User
Advanced User
 
Posts: 87
Joined: Tue May 10, 2011 5:27 pm
Location: Redwood City, Californa, USA

Re: The search for overlap.

Postby Fat Freddy » Sat Apr 25, 2015 5:02 am

Would you have any objections to a summary of this post (when complete) in the form of a topic for the newsletter?

Not sure when, but I think the information and theories here are invaluable and should be shared.


I have no problem with that. It would be good to get someone with editing skills make something of it.

Based on the explanations above, the video and the fact the demand on intake air will be increased during combustion I can see how how supply cannot be met with demand given the increase in speed of the piston during combustion.

In this scenario its not surprising the peak pressure is slightly lower during combustion that the running compression test (no combustion)


That is true mainly on a carburetted engine. The running compression (no combustion) would effectively be a miss. Therefore engine RPM would slow giving more time for air to fill the inlet manifold and more time to fill a cylinder. That increase in fill time would improve the volumetric efficiency of the cylinder there fore increasing the compression pressure.
But an EFI engine generally want's to achieve a target idle RPM. So when the running compression (no combustion) test is done, again creating a misfire. The number of cylinders has dropped to 3 (obviously) but the load has increased. The only way to deal with added load and to maintain desired RPM is to convert more energy. Burn fuel. More fuel needs more air. So the engine has to open up the throttle. So now more air gets in during the same time. Once more increasing volumetric efficiency.
This is why it's easier on a EFI engine to match events occurring with the crank/clock and WPS. Although not perfect. I hope to show that in the next couple of days. (problem with saving - see forum feedback).

I thought I'd better just mention that increase in compression is due mainly to a combination of throttle opening and and crank speed. Which your image shows well. This is why I mentioned earlier it would of been nice to get a running compression (with combustion) and disabling another cylinder to see how much more load it did put on a firing cylinder. It'll just have to wait for another day I guess.

Cheers
FF
User avatar
Fat Freddy
Zen Master
Zen Master
 
Posts: 267
Joined: Fri May 04, 2012 4:52 am

Next

Return to Diagnostic discussion

Who is online

Users browsing this forum: No registered users and 0 guests