Hi,

I haven't done the calculations or thought deeply about this, so I hope you won't take this as a flippant response. But it seems clear (not only by your unexpected result) that you are NOT measuring the BULK velocity of sound (either transverse or compressional polarizations) in the aluminum. In the first place, (you can look this up) a bar (as opposed to a bulk material) has about 5 different acoustic transmission modes. You are somewhat aware of that since you mention surface waves, and indeed that would be my suspicion if you had a large block of material and were propagaing from one edge to another (in which case I'd suggest going from the center of one face to the opposite face). But in this case I can see a greater problem. With your 100KHz frequency, you'd expect a wavelength for the bulk wave ~30-50mm. But the bar is much thinner than that! So what you have is not a bulk material but a WAVEGUIDE! Thus even if you have excited a compressional mode, the transmission isn't through a homogenous medium where you'd expect the velocity you mentioned, but a waveguide which has a totally different dispersion relationship depending on the dimensions in relation to the wavelength. Just like an optical fiber, for instance. Or it can be a simple bending mode (like a guitar string) which is excited LIKE a transverse mode but nevertheless is quite a different beast. All of these will have different propagation velocities (also depending on which transverse polarization, given that the crossection is rectangular).

Do you not think that this would explain the discrepancy? Or more to the point, can you get a piece of aluminum which is much much more than 1 wavelength in the transverse dimensions? (I realize it would be much too heavy to pick up! :-)

(If you'd like to respond then please email me directly so I will see it, thanks)

I was interested top read this as I have been trying to measure the speed of sound using a solenoid to tap a wooden bench and two piezo transducers connected to ch1 and ch2 of an oscilloscope. I also got strangely slow results, I was using the alt mode also and am now wondering if that was the problem, I will try again next week.

Alan

Is it possible that you were seeing a reflection from the wall behind your experiment instead of the wanted short-distance path? A suitable acoustic reflective material placed at 45 degree angle behind the experiment layout might help deflect overshoot signals from reflecting into your test setup. Acoustic absorption could also be attempted but it might be more difficult to obtain 100% signal loss.

If the desired response is very short, it may be getting masked by the initial transmitted pulse. This then leaves only unintended reflections as the viewable echo.

Tests using other media (water, oil, human tissue, etc.) might be interesting and may reveal some cause for earlier unexpected result.

One possible direction for this experimentation might be the use of electro-mechanical mountings for the transducer to make acoustic images from various objects, including human body parts or small animals.

Hmmm...maybe an acoustic SWR bridge? I have no idea how to build that, but it does sound interesting.

Arv

_._

I used some PC echosounding/sonar software to try to do distance measurement, and I found it strange that often a reflected echo of the sound (with a longer time of flight) would be the strongest signal received. Further, I found that I couldn't use the "earliest" signal either because (i guess) harmonics were coming into play and I would see times that could only be half the distance... Or maybe artifacts of digitization.

Dayle,

Indeed, I considered that. The result is the same, in chop or alternate (or single trace mode triggered by the TX burst or transient). I used chop-mode in the videos and pictures because the sweep rate is actually fairly low for these frequencies and the camera was having trouble with it. (You can see the herring-bone artefacts of the chopping in the still picture.)

None the less I was beginning to question my sanity and my test equipment... I checked the time-base against crystal oscillators to make sure it had not developed a fault.

I do wish I had a better scope. This Asian cheapy is getting quite old, but it has served me well, I've had it since I was 14, my Dad bought it for me and it has taught me a lot. The triggering is not especially good, and it is only a 20 MHz unit. I think my next scope will be a Tektronix. I have a PC-based DSO with more bandwidth, but its usability leaves a lot to be desired. The -ve triggering (mid-buffer) options of digital scopes are a dream to use though. I should probably use the DSO more for preparing articles for the web, screen shots are easier than photographing the analogue scope and you can extract the data digitally and manipulate it.

Unfortunately I won't have a lot of time in the immediate future to continue experiments. That bothers me, I really want a few days in the lab to work this one out. Meanwhile the list of other things I'd like to look into continues to grow without bound. Everything I do lately seems to open more areas to research and there are only so many hours in the day. If only I could work out a way to get paid to do this instead of a real job...

Regards,

Alan

Gidday. I have to say, I find your diverse range of experiments fascinating to say the least! I wonder here if you are caught with a scope problem that tripped me up recently when attempting accurate phase shift measurements at 455Kc. If the scope is used in alternate mode, errors abound, the best to use of course is the dual beam, rather than dual trace. Do they still manufacture these I wonder? regards