Replies to This Discussion

Permalink Reply by Yellowbelly Flatt on November 13, 2010 at 8:08am

Well yes,
I have noticed that the piezzo pick ups that you mention do seem to vary from one instrument to another. I had always thought that this was due to the fact that each instrument that I have made has been different from every other that I've created. Different woods, different materials (wood/ply/tin/resonator et al), different sizes of instrument and different location of the piezzo in the construction will surely affect the tone and volume of the finished guitar, ukulele, fiddle, bass, mandolin, etc, etc.

Permalink Reply by bairfoot cajun on November 13, 2010 at 12:39pm

i have noticed this too,different charateristics are due sometimes i have noticed because of the different thicknesses of the ceramic coating on the metal disk.you can not see the difference but if you put the piezo inbetween your finger and thumb you can sometimes feel that one is thicker than the other. i did an experiment once to see if this was the case,i set up 2 different piezos with wires and input jack,one thicker than the other on the ceramic surface,and i could hear that the thicker one seemed to have more volume to it. i placed each one seperately on the top of a cigar box guitar in the same location and held it there. it might just be me but i think that this difference might explain some of the variables in each piezo that makes one different from another.

Permalink Reply by Mark on November 13, 2010 at 2:08pm

When you look at the specs for those discs the tolerances seem quite wide. According to Maplins, for the YU87U the numbers include:

Resonant frequency: 1.8kHz ±0.5kHz
Capacitance: 25nF ±30%

I honestly don't know what the implications of that are in terms of their behaviour as pickups - It might turn out that these sorts of variations in resonant frequency or capacitance make little difference to pickup tone. But it seems to me as if variation in resonant frequency might indicate variation in other frequency characteristics. And ±0.5kHz seems like quite a spread when the nominal resonant frequency is 1.8kHz.

So your theory seems credible to me.

Permalink Reply by smojo on November 20, 2010 at 7:52am

Physics lesson coming up - might help understand a bit. The discs are primarily designed as buzzers, they just happen to work microphonically as well. So I think the resonant frequency quoted is the one that it will 'buzz' or oscillate at when used that way. Using it as a mic is like putting it in reverse but the tonality will be more down to the construction/materials of the box but the piezo type may have a slight bearing.

Having messed about a bit with my graphic EQ to control feedback, I was surprised that the main frequencies of feedback were around the low end 300-500 Hz. That's the frequency that my guitar boxes are resonating at which is not the same as the one stated on the piezo when used as a buzzer. The normal audio range for humans is about 20-18000 Hz at best but from an instrument it is much lower. Top end more like about 6000Hz. So what I'm saying is there are various factors involved. In short though I have noticed different tonalities but not made a note of what type of disc was used. Probably is some slight difference but the individual resonances of the guitars and boxes make it hard to determine.

Permalink Reply by Mark on November 20, 2010 at 12:19pm

Your post prompted me to do a bit of reading on the subject. Although it's true that piezo disks are often marketed as cheap speakers, their use as pickups or sensors is also well established - and their frequency characteristics are of interest to people designing for such applications.

Wikipedia provided the following diagram of the general form of the sensitivity characteristics of piezo devices - specifically the sensitivity plotted against frequency. It seems that for good fidelity you want the frequencies you're trying to pick up to be in the flat region between the resonant frequency of the piezo and the "high pass" cutoff point. Or to be more practical you want to choose a piezo to match the sound frequency range you're working with. Unfortunately retailers don't seem to quote the "high pass" cutoff, just resonant frequency.

Of course, there's also a possible debate about whether you want a perfect flat frequency response curve for a guitar pickup, but I would have thought that if you were aiming for an "acoustic" sound then you probably want something that gives a reasonable reproduction of the physical sound. And that's a separate issue from the overall question of what makes a good guitar sound - which is something that depends on a combination of strings, neck, body, nut, and bridge and so on (including choice of materials, which all have different properties in terms of damping some tonal components and accentuating others) in addition to any tonal shaping from the pickups.

You also got me thinking with what you said about the frequencies at which you get feeback (which I recall you mentioning in West Brom). The 300-500 Hz range includes the fundamental frequencies of the higher strings on a guitar with regular tuning (top E = 329Hz, B = 246Hz). It suggests to me that the resonance of the strings might play a significant role in feedback - with the stings, the box and the neck acting together as a coupled oscillator - as opposed to feedback being something that's only a characteristic of the box. I think the fundamental will tend to be the harmonic component with the biggest amplitude, and it seems to me that would be the frequency which will reach self-sustaining resonance levels first (as you gradually increase the gain of the feedback route).

I guess these are things we can only know for sure with the benefit of more research.
smojo said:

Physics lesson coming up - might help understand a bit. The discs are primarily designed as buzzers, they just happen to work microphonically as well. So I think the resonant frequency quoted is the one that it will 'buzz' or oscillate at when used that way. Using it as a mic is like putting it in reverse but the tonality will be more down to the construction/materials of the box but the piezo type may have a slight bearing.

Having messed about a bit with my graphic EQ to control feedback, I was surprised that the main frequencies of feedback were around the low end 300-500 Hz. That's the frequency that my guitar boxes are resonating at which is not the same as the one stated on the piezo when used as a buzzer. The normal audio range for humans is about 20-18000 Hz at best but from an instrument it is much lower. Top end more like about 6000Hz. So what I'm saying is there are various factors involved. In short though I have noticed different tonalities but not made a note of what type of disc was used. Probably is some slight difference but the individual resonances of the guitars and boxes make it hard to determine.

Permalink Reply by smojo on November 23, 2010 at 6:21am

Wow - good work Mark. The YU87U has a quoted res. freq. of 1.8kHz whereas the QY13P has a res. higher freq of 4.2kHz. Taking your flat area as being below the resonant one, it would seem that they are quite well suited to the range of a guitar though maybe a little top endish. Be interesting to know the actual frequencies of that flat area. The QY13P will probably give a slightly brighter more trebley tonw

Just out of interest, from my BT days I know that the frequency band for 'intelligible' speech was quite narrow - only 300Hz - 3kHz which is all the telecomms equipment was designed to deliver. So again, the piezo makes quite a good mic for voice.

Permalink Reply by Mark C on December 31, 2010 at 12:32am

Thanks guys for the contributions, some very interesting stuff has come up, some of which I understood, some of which went over my head (no surprises there).  I think the only way I'll get any sort of conclusive answer to my question would be to make a test rig and record the sounds to something like Audacity software to compare the results.  Finding the time is the problem, though...but I think this is going to nag me so I'll put it on my project list.  Thanks once again gents.