Measuring RPM From Video

[Chris-PA]

Recently I finished a squish reduced 42cc Poulan/Craftsman that I've been very happy with. I also have a 46cc Poulan 2775 that was the first saw I ported, and I have usually been quite impressed with it. They are two different designs, although both clamshell homeowner saws. The 46cc runs a 20" bar with 0.325" Oregon 20BPX, while the 42cc has a 19" GB bar running 3/8 lo pro Oregon 91VXL (so it has a narrower kerf). I wondered how they compared, and so I tried cutting cookies off a white oak log.

This tree has varied a lot, with some of it being quite soft and some quite hard - and this chunk was hard. The 46cc saw has had a difficult time in the past on very hard wood, and here it was no exception - it got crushed by the 42cc saw (14sec vs. 24sec). It didn't feel like it was struggling, and so in an attempt to figure out what's going on I decided to try to measure the rpm from the video (audio really).

First, here are the two videos:




I run Linux, so I'm using an application called Audacity - there may be Windows programs that can do similar things. I selected a portion of the track where the rpm and amplitude seemed fairly constant, and then selected Plot Spectrum from the Analyze menu:


The plot can be full screened, and the cursor will tell you what the frequency is at the cursor (this is the 46cc):

The first big spike is the fundamental frequency of the engine, in this case 161Hz. Multiply by 60 and you get 9660rpm. Now for the 42cc:

It shows basically the same thing, which tells me the issue here is probably mostly the chain (which was sharp). Anyway, this can be done for most any of the saw recordings we show on AS, although of course the result does depend on what part of the recording you select.

 

[johnny5ny]

I know you addressed it already but the comparison vids sure do look like dull chain vs. sharp chain...

 

[Chris-PA]

I know you addressed it already but the comparison vids sure do look like dull chain vs. sharp chain...

There was another video I took first but didn't post - because it was slow I had filed it right before the video shown. So it was sharp but of course there could also be issues with depth gauges and angles, which I will be looking into. In fact that chain is wonderful in other wood.

The saw was running at a decent rpm, but that doesn't tell you how well it held it or if it would have bogged down if I had pushed harder. My impression is that the 42cc maintains rpm a little better with loading, but then the wider chain on the 46cc will be more of a load.

Mostly I wanted to show how to measure the rpm from the recording.

 

[Raganr]

Ha. Chain police are always the first in.

Clever way to measure rpms. You are pretty inovative with your saw mods and now this. Keep up the good work.

 

[SAWMIKAZE]

It makes me nervous how smart some of you guys are.

But thats pretty cool.

 

[angelo c]

It makes me nervous how smart some of you guys are.

But thats pretty cool.

You're right Ryan....I hate when these young kids with their fancy computer thingees....especially when they know what to do with 'em....and they do good.

Well done Chris....even fer a youngster like you .

 

[Chris-PA]

even fer a youngster like you

I don't seem to get called that much any more!

 

[one.man.band]

when it's time to steamboat. you steam. - clemens, make sure you steam in the proper direction?

look past the over-ported 46 comparison, to the right hand side of the graph. something more interesting happening there. when the light bulb turns on......wish you well in explaining it here. did not have much luck trying a few years ago. dust off the calculator.

 

[Chris-PA]

I thought this was interesting, and maybe an illustration of how this could be useful. Here's a video of the 46cc from last year - same saw, same bar, same chain noodling a big round from the same tree:

Here it was full bar and cutting pretty fast for a 46cc saw. Mostly it's running around 9300rpm, so it's definitely loaded more.

I'll be looking at that chain in the next few days.

look past the over-ported 46 comparison, to the right hand side of the graph. something more interesting happening there. when the light bulb turns on......wish you well in explaining it here. did not have much luck trying a few years ago. dust off the calculator.

The strong lines to the right of the fundamental are harmonics - multiples of the fundamental frequency. All systems will have harmonic resonances like that, and I would not assign a lot of meaning to them. The rest of the hash up there at higher frequencies is likely to be from things like air rush (intake and exhaust), and other high frequency resonances and background noise. Here's the spectrum of the 46cc at idle from the beginning of the video:



These plots might be more directly useful for looking at muffler mods and noise levels.

 

[one.man.band]

"The strong lines to the right of the fundamental are harmonics - multiples of the fundamental frequency. All systems will have harmonic resonances like that, and I would not assign a lot of meaning to them. The rest of the hash up there at higher frequencies is likely to be from things like air rush (intake and exhaust), and other high frequency resonances and background noise. Here's the spectrum of the 46cc at idle from the beginning of the video:"

.......so your implying there are resonances. how can this be? ..... in a can muff, no less. thought science did not apply in this black hole.

my point exactly.

what happens around the freq point of harmonic motion? how can the freq be changed? how could this be beneficial?

just wanted to get you thinking a bit is all, chris.

 

[Chris-PA]

.......so your implying there are resonances. how can this be? ..... in a can muff, no less. thought science did not apply in this black hole.

my point exactly.

what happens around the freq point of harmonic motion? how can the freq be changed? how could this be beneficial?

There are always harmonic resonances when real systems are driven at a frequency, although I was not saying they were from the muffler. The whole crazy thing is vibrating in response to the firings and crankshaft rotation at the engine rpm - cooling fins, various parts, what have you. The exhaust pulses themselves will have harmonics in some curve, and some of these harmonics will be close to resonant frequencies of various parts so they may accentuate a given harmonic. That could include acoustic resonances in the muffler, but the volume is pretty small and it's intentionally designed to attenuate these (that's its purpose).

Automobile designers do analysis much like this on new designs, where they vibrate the car at various frequencies and look for parts that resonate (trim pieces that buzz, etc.), in order to make them quieter.

 

[wde_1978]

Interesting!
Although I don't quite understand how frequency can be used to calculate rpm's - but then again I am no mathematician either nor do I usually bother with stuff like this.

 

[one.man.band]

take another look see, at your first post 42cc graph.

..i will point closer.

just east the 2nd harm, and anti-node........notice the amplitude spikes. they are spiking even higher than the 1st order harm.

what is causing this change? could it be the can?

 

[one.man.band]

is there a way to get rid of the 'log' graph and change display to a regular incremental graph?

 

[one.man.band]

for the love of hockey!

chris, you are gleaming over some interesting things. a few folks have already looked at freq analyzing here in the past.

what you and they have not seen in doing this is that it is actually test bed. a sound dyno for can muffs.

jam a 10mm from your 1/4 inch socket into your muff exit......do the sound sample. figure the freq. and rpm at load.

many openings could be tested in a few minutes.

look at the graphs......then you will know the significance at what i was pointing at.

 

[Chris-PA]

take another look see, at your first post 42cc graph.

..i will point closer.

just east the 2nd harm, and anti-node........notice the amplitude spikes. they are spiking even higher than the 1st order harm.

what is causing this change? could it be the can?

I'm not sure what you mean by anti-node - those areas between the harmonics are just the background sound level at those frequencies, it's not driven lower by being between the spikes. It looks to me like the 2nd harmonic is some 18dB down from the fundamental, and none of the others is higher either - and in this context I see no real significance to it anyway. I'm not sure what you think they mean? They just look like a typical harmonic response of a system driven at a frequency.

is there a way to get rid of the 'log' graph and change display to a regular incremental graph?

Yes, but it just makes it harder to read, which is why a log scale is normally used for such plots.

 

[Chris-PA]

Interesting!
Although I don't quite understand how frequency can be used to calculate rpm's - but then again I am no mathematician either nor do I usually bother with stuff like this.

Frequency is measured in Hertz (Hz), which has units of cycles per second. Multiply by 60 and you get cycles per minute. A crankshaft revolution is a cycle, so cycles per minute = revolutions per minute, or rpm.

 

[one.man.band]

Yes, but it just makes it harder to read, which is why a log scale is normally used for such plots.

......more difficult to read log if you are searching for the can res freq.?

my interest is in other areas at the moment.

hope it helps, chris

 

[malk315]

I'm a linux nerd too -- using it right now on a desktop to post this... been using linux since 1994 when nobody knew what it was! Embedded linux is what the products I create and work on run -- it's fantastic. Linux is everywhere these days -- android phones run a linux kernel and application code that makeup the Android OS.

Anyway -- being the die hard linux guy that I am I can confirm Audacity runs just fine on Windoze too -- I muck w/ that OS too like most people! I use it to grab audio from youtube and then save as an mp3 file using the "stereo mix" most sound card drivers on windows support but don't enable by default. Here's a howto which mentions audacity for this: http://www.computerhope.com/issues/ch001137.htm

As far as measuring RPMs with audio once you have how many HZ the saw firing is happening at you can just multiply by 60 to get RPM. That is because HZ is cycles per second -- i.e. how many times per second the cylinder is firing. The cylinder fires once every revolution so the spike on the graph the first frequency peak shown you have how many revolutions per second the saw is firing. 60 seconds in a minute you multiply by 60 and you get revolutions per minute or RPM. Further above one of you was asking I'm hoping this helps explain it more.

With the saw in the wood firing every revolution and not "4 cycling" you can get a decent measure of how many times it fires per second in HZ with just an audio recording. Pretty sweet!

Cheers Gentlemen.

 

[wde_1978]

This is a graph out of Nero WaveEditor featuring the frequency levels of my PS-7900 cookie cutting pear (video featured HERE) , but I can't figure out where to read the hertz from:



Audacity won't run on my desktop PC (and I know why) , so I'll just stick to what my tach claims!