Hello folks,

I have a bit of problem, just a bit though  ;)

I can't figure out which sensor(module) would suit my application better.

We are going to be measuring pressure"waves" in the exhaust manifold to optimize manifold design and scavenging effects...

nothing to worry about the temperature... we'll be cooling the gases down and correcting for the temp mathematically.

But... would the FLS (FirstLookSensor) be better suited or the WPS500???

Please help... we want to order the system ASAP...

The local (netherlands) distributors aren't up to the job answering this seemingly easy question...

Thanks you all.

Rudolph

The firstlook reacts to pulse waves or differences in pressure.

The WPS is a pressure transducer that measures the actual pressure and is scaled.Do you need a scale to work to ?

I find the firstlook more sensitive.

Rudolph,

The FLS is more sensitive but the WPS also does exhaust puses well along with many other things like in cylinder pressure testing.

The WPS is a pressure transducer the FLS is not.

The FLS is a pulse sensor. Any steady state of positive or negative pressure results in a zero output.  However, it is very responsive to changes in pressure.  The output being proportional to the amount of change and the rate of change.

If all you want to do is exhaust pulses, the FLS would do well as a lower cost choice.  If you think you might want ability to do in cylinder pressure testing get the WPS.

buy the WPS. Then make the BLS - in the project forums (costs 5 bucks to make) and then you can have the best of both worlds. :)

BLS???


Thanks you all for your answer...

But if "all it does", the FLS, is record pulses and therefor being more sensitive... Shouldn't I be able to "just" make up a little algorithm in MATLAB to determine the rate of change?

And in doing so approximating the results I should've gotten directly from a FLS...???

And yes it (being the FLS or WPS) will be used to measure more than just pulses...

EPS wrote on Sep 13^th, 2011 at 12:43pm:

BLS???

BLS is just a piezo element sealed in a chamber where vacuum creates a voltage we can read on the scope. It is a DIY version and it is proven to work extremely well.

To calculate rate of change you would have to accumulate a lot of data and use calculus to arrive at the answers. As far as I know, Pico does not support calculus equations (yet) in the custom probes, only linear equations.
So I think your best bet in your situation is going to be the WPS.
Then buy a piezo element and make a rate of change/piezo sensor.

If you do happen to figure out the calculus to get scaling data from the piezo/FLS sensors, please let us know. I would be very interested to see the formula.

I used to be in engineering school - but it just wasnt for me. I would rather fix the broken machines that engineers designed. :)

EPS wrote on Sep 13^th, 2011 at 12:43pm:

And yes it (being the FLS or WPS) will be used to measure more than just pulses...

Hang the expense and go the WPS.Far more versatile tool.
It has three ranges and you can customize settings within the ranges.
Great for in cylinder testing and allows you to check mechanical condtion of an engine, compression pressures, check valve timing, turbo boost...

I have both and use both.Two slightly different tools  :-/

Randy H. wrote on Sep 13^th, 2011 at 1:22pm:

To calculate rate of change you would have to accumulate a lot of data and use calculus to arrive at the answers. As far as I know, Pico does not support calculus equations (yet) in the custom probes, only linear equations. So I think your best bet in your situation is going to be the WPS.

I will be using MATLAB/Simulink, check www.mathworks.com, to do calculations. It is a data-analysis program with a math engine behind it.



Thanks for all the help folks.

The FLS does not measure pressure at all. It measures pulses (immediate changes in pressure), both positive and negative. In fact, there is no way to convert any steady pressures with the FLS to any kind of value.

James wrote on Sep 22^nd, 2011 at 8:14am:

In fact, there is no way

I wouldnt say that EXACTLY. I would say that its not a linear relationship, and you could only measure a delta pressure value, no steady state value. I have collected a few values on the very similar BLS, and it appears to be an logarithmic function. However I didnt get into it very deep.

If someone wanted to do the research and did some clever math and collected the data, it would be possible to get a custom probe (delta pressure).

But i see what you are saying also James.

I wish to share with you some gathered data.

Please bear in mind this data actually isn't valid, because the engine (Chevy small block) is running without external resistance of any kind at 3000 rpm.

the four graphs are done with MATLAB

Parameters:
Shorty headers
Passengers bank of the V8
pressure measured in the collector (not exactly inside, but branched off)
only variable is the mufflers
   - chambered muffler (we call it a reflection muffler)
   - cherrybomb
   - 3 inch sidepipe (open)
   - some weird home grown tiny bore sidepipe

The reason for doing this is to get to know the equipment and how to get good data/measurements before going to an enginedyno

But despite that fact, I think the difference is significant. Which is why I want to share.

http://www.classiccarauto.com/impala/how_to/images/distributor.png

according to the above link you can see the firing order, the ignition signal you see has been taken off the 2nd cilinder.

Hope you like it.

---

My question to you is this:
DO you think these differences are caused by slight differences in backpressure, resonance, or a combination of both?

I have noticed that on cars without a cat resonance is a big player - and can make exhaust waveforms nearly unreadable.

Can you do these tests again during cranking? This would be interesting. The cranking test, to me, should always be done first before comprehending the running waveforms. Its just something I practice that Mick taught me. Im not shooting down your test, just offering some friendly advice.

For some reason the BLS (very similar to FLS) is very much affected by resonance, but if I use an old MAP sensor it is not nearly affected as much, although it (resonance) is still present to some degree. You really have to be familiar with the tool you are using thats for sure. Engines with lots of valve overlap can really show some strange patterns, and be nearly unreadable when running. But they become a lot more useful during crank with fuel disabled.

Thanks for sharing this EPS, and look forward to some more of your findings, and your input here on the forums. ;) :)

If you look solely at the waveamplitude you can see a significant difference between the first plot, which has an OEM style silencer/muffler/dampener sirectly on the exhaust manifold, and the second plus third.

But between the second and third plot, in terms of amplitude, there isn't any noticeable difference.
I think this is mainly cause by them being similar in style. "straight-through".

Another thing is. This is a V8 measured on only one bank with VERY short headers!!! As mentioned before. The plots above are mainly to get to know the hard- and software and how to extract the data.
I just noticed something which could be linked to a variable, the: silencers/muffler/dampener in this case.

Coming up are plots with headers at least twice the length. and I will be measuring the pressure wave on cil 2, to trace it.
If this works I'll be doing the same thing to all the runners, so I can overlay and retrace what's happening to the pressure.

But to answer your questions, I hope.

resonance, which do you mean?

Pressure wave resonance can be useful as well as a nuisance. What I am trying to accomplish is optimizing headers to increase, or infuence, the scavenging effect. In order to do this I first need to know, or basically understand the pressure wave behavior inside exhaust runners, collectors, etc.

I'll try to do the cranking test.
(Need to figure out the trigger function)

PS: what I want to do is recreate what has been done in some literature.
look-up the following ISBN code for some interesting stuff about exhaust and intake systems.
9 780837 603094

http://www.amazon.com/Scientific-Exhaust-Systems-Engineering-Performance/dp/0837603099

PS2: I'm using the WPS500

EPS wrote on Oct 1^st, 2011 at 2:36am:

This is a V8 measured on only one bank with VERY short headers!!!

Can you post a picture of your set up?


EPS wrote on Oct 1^st, 2011 at 2:36am:

But between the second and third plot, in terms of amplitude, there isn't any noticeable difference. I think this is mainly cause by them being similar in style. "straight-through".

Are these 4 plots taken at different times and different cylinders on one bank but synced of cylinder 3 or 4?

EPS wrote on Sep 13^th, 2011 at 4:48am:

Hello folks, I have a bit of problem, just a bit though  ;) I can't figure out which sensor(module) would suit my application better. We are going to be measuring pressure"waves" in the exhaust manifold to optimize manifold design and scavenging effects... nothing to worry about the temperature... we'll be cooling the gases down and correcting for the temp mathematically. But... would the FLS (FirstLookSensor) be better suited or the WPS500??? Please help... we want to order the system ASAP... The local (netherlands) distributors aren't up to the job answering this seemingly easy question... Thanks you all. Rudolph

No dilemma... Get'em both.   8-)

Ok!!! new graph, I hope this one will be a bit better

Please bear in mind. we are still starting up and checking how to get useable data.

This is, if I say so, one of the better graphs.

the green line is a performance header.

The elapsed time is 4 ign-cycles taken from cil.2 (so it starts at cil.2 and end at cil.2)


the graph posted is wrong

i mixed up some data... sorry guys...