Can you play guitar through a smoke alarm?

[anotherjim]

I was just about to mention the "mic" function. Because of the high impedance, the piezo does make a better mic, or indeed, a trigger pulse generator - they can do both. Other things more speaker-like make better "surface exciters". Old headphones/earbuds etc.

3 wire piezo is, as already said, meant to form an oscillator running at its resonant frequency. The more usual and useful kind is 2 wire. The back/substrate normally used as 0v/ground and the middle coated active side as the hot/signal connection. This last is tricky to solder - so keep the wires it already has.

Piezo isn't a complex load like the inductive coil of a loudspeaker, so the simplest 386 circuit may do very well.

[moid]

Put volume  potentiometer on the input. See Fig. 10 of this document http://www.ti.com/lit/ds/symlink/lm386.pdf
Note the 0.05uF cap + 10R resistor combo from output to ground to suppress oscillation. LM386 tends to oscillate without it, but maybe that's what you might have fun with.

Thanks very much for the link - I haven't vanished, but had to do a lot of DIY work for my mother in law and it was far too hot to break out a soldering iron! (It was too hot to be putting up shelves and curtain rails too, but that's another story). The weather finally cooled down today so I got all my tools and parts out for the first time in a year, managed to desolder the piezo elements and the capacitor, looked at the schematic you linked me too and realised I've forgotten so much since last year that it is rather sad (and alarming!). I've decided to use the schematic as a chance to re-learn some things so I have chosen the option to have a bass boost and will now draw it on vero so I can build it. I'll post it soon (I'm drawing it at the moment).



This is the schematic I'll use thanks - I suspect that the piezo elements will be somewhat bass light to put it mildly so anything that can push more bass might be a good idea. Another thought came to me - this diagram ends in a speaker. If the piezo doesn't work, I do have a small 2" speaker that reads 4 ohms on the back - could I connect that instead? Also could this circuit be used as a booster circuit effects pedal for guitar? Or is it for outputting sound to a piezo or speaker only?

[moid]

One thing that I can't see mentioned anywhere is that a piezo "sounder" can be used in reverse as a piezo *mic*. They make excellent drum triggers. You attach them (glue) to the surface of your drum and then when you whack it, you get a big signal at a fairly consistent frequency and at a volume dependent on how hard you hit it. You can then use your favourite envelope follower design to produce a volume envelope, or use a trigger circuit to trigger a drum sound generator, or some mixture of both.

Given that it isn't going to be putting Celestion out of business any time soon, perhaps that's a better use for it.

T.

Thanks for that - something new I didn't know! I don't have any drums... hmmm some toy ones perhaps... but I'll keep them for this purpose as a back up if they don't work!

[moid]

I was just about to mention the "mic" function. Because of the high impedance, the piezo does make a better mic, or indeed, a trigger pulse generator - they can do both. Other things more speaker-like make better "surface exciters". Old headphones/earbuds etc.

3 wire piezo is, as already said, meant to form an oscillator running at its resonant frequency. The more usual and useful kind is 2 wire. The back/substrate normally used as 0v/ground and the middle coated active side as the hot/signal connection. This last is tricky to solder - so keep the wires it already has.

Piezo isn't a complex load like the inductive coil of a loudspeaker, so the simplest 386 circuit may do very well.

Thanks anotherjim - I may well try the old 2" speaker I found instead.

[thermionix]

(and alarming!)

[moid]

(and alarming!)

I do hope I heard a groan for that pun My Dad joke genes were rather pleased with it!

Onto electrical things - I've made a drawing (with colours too!). It might even look like a vero layout of the lm386 amplifier. Or maybe it's abstract art? Your choice. If anyone electrically minded could look over this before I start building it that would be lovely If anyone wants to assess it on a scale of aesthetic importance and provide some art criticism, that too will be duly noted. Thanks!

[duck_arse]

moids - your electro caps are all arse-about. and, why so wide? izzat a pre-size board or summink? I'd move that 100nF across to the left 3 rows, and maybe fit the link under the IC. then I'd move the 100nF/100uF caps top connections down 3 strips, and move the both caps left to near below the IC. then I'd while away the hours shuffling all those right hand end parts across to wherever there was space on their left.

but that's just me. you probably want a log pot on the Volume, you might need an anti-log "C" taper on the Gain pot, I'm not sure.

[moid]

moids - your electro caps are all arse-about.

You know, I was wondering about that... I ummed and aahhhed about it for a while... and then I erred (on the side of failure). Thanks for spotting that one!

and, why so wide? izzat a pre-size board or summink?

Because I like big booty and I cannot lie? Oh whoops, wrong forum... ahem... is incompetence and ignorance a valid reason?

I'd move that 100nF across to the left 3 rows, and maybe fit the link under the IC. then I'd move the 100nF/100uF caps top connections down 3 strips, and move the both caps left to near below the IC. then I'd while away the hours shuffling all those right hand end parts across to wherever there was space on their left.

but that's just me. you probably want a log pot on the Volume, you might need an anti-log "C" taper on the Gain pot, I'm not sure.

Thanks very much - I think I got all of that (please see below image if you don't mind), I don't have any anti-log pots, but in the miraculous event that this circuit actually works I will try log and lin pots on the volume to see which works best.

Will this circuit work OK with a 4ohm Speaker? I've never wired speakers to a circuit before and I know they have different ratings for impedance, but perhaps this only matters if I was putting serious volume through them?

How's things in Downside Up land? We finally have rain here! I'm not actually sweating 24 hours a day

[GibsonGM]

A 4 ohm speaker should do fine with the 386, Moid!  Highly doubt the 386 can toast it, and I know others have used them too.

[moid]

Thanks GibsonGM, I'll start cutting/soldering/cursing then

[moid]

Some updates

Well first of all I chickened out on using the smoke alarm piezos - I went with the speaker I had instead (which turned out to be 8 ohm and not 4; this might be why I have a small issue below...that's what she said anyway*). Sound comes through the circuit, plays out of the speaker, which is currently taped to the inside of a thin walled metal tin (tall rectangular thing that had some Italian Amaretti biscuits in it years ago) and I have a contact mic with three piezo discs (for acoustic instruments) attached to different parts of the tin and these pick up the sound very well. Depending on the frequency of the music (and volume) there is a very interesting distortion / muffling effect that also manages to to odd things to the pitch in places; chords and bass notes cause the most distortion. There isn't any reverb effect that I can noticeably hear, but the tin isn't very large and I assume to get some reverb there needs to be a larger physical distance between the speaker and the contact mics. At the moment the sound is reminiscent of the Death By Audio Reverberation Machine in terms of slightly random distortion, but it needs more reverb. I will go on the hunt for a larger tin box tomorrow, and also I'll try an old metal baking tray that is hiding in the garage to see whether that will work better.

*problem (not devastating). When I turn the bass boost pot towards one end of its sweep the sound is cut from the speaker and there's no audio. Is this because there is too much bass frequency going into a tiny 2" speaker? Or is it because the speaker is 8ohm and not the 4 ohm I thought it was? Or something else?

I'm already thinking the circuit needs a blend between the clean output and the distortion the contact microphones pick up. If I was to have just the signal coming from the circuit (not the speakers) as one end of a the sweep of a pot with the other end being just the contact mics, would the clean output cause issues to an amp? I'm thinking this circuit works quite well as a clean boost, is there any reasons why I shouldn't use it in this way?

Thanks! I'll make some recordings once I find the right container for the device.

[anotherjim]

8ohm just can't develop as much power as a 4ohm can with the same circuit.
The cut out with bass boost could mean the power supply isn't happy.

I would wait until happy with the basic rig before building a more specific circuit.

Many reverb devices are simple. One in and one out. They are commonly used in a send/return system. Typically a mixer auxiliary send feeds the reverb and an auxiliary return input adjust the "Wet" reverberated signal back into the mix. The original "Dry" signal goes into the mix via the channels normal fader path.
That arrangement can be built into one box if you want. Input pre-amp > pre eq > power amp > driver (speaker) > reciever > preamp > post eq > wet/dry mix > output.

[moid]

Thanks AnotherJim. I will look for a 4 ohm speaker instead. I discovered I could remove the bass cut out part of the sweep by adjusting the setting on the gain control... so something is odd there. I'll try a different speaker instead once I've got one. I also need to audition some different metal tins... the current one isn't giving me the right tone so far Not metal enough (ba dum tssss).... actually when it distorts the high notes they go out of pitch badly, and it's hard to get any variation in the fuzz effect; it's either full on or off depending on how hard I play, so I will try different surfaces to reverberate through. I looked up real plate reverbs and they need a 1m x2m sheet of 0.4mm stainless steel and I thought hmmm I wonder what they cost...apparently about £250 - £300 in the UK. Ouch, metal hurts.

Anyway all testing is out for a couple of days because I've just done my back in moving a shed in my mother in law's garden

[duck_arse]

you get junk like this there in blighty, yes? must be worth a try, what with all the surface area to resonate. springs across, or from lid to base, away you go.

you'll need to remove all the internal damping material before use, tho.

[edit :] I forgot to add - sweet.

[anotherjim]

Well, stainless because rust sounds awful. Galvanised/painted steel sheet can be found. Tin sheds are made of it. Corrugations can be rolled flat -  or left as is.

Have you a Wickes nearby?
https://www.wickes.co.uk/Wickes-Metal-Sheet-Galvanised-Steel-250-x-500mm/p/116718

More useful for casing/chassis, but...
https://www.wickes.co.uk/Wickes-Metal-Sheet-Plain-Uncoated-Aluminium---120-x-0-8mm-x-1m/p/116700

[PRR]

The velocity of sound, in air or in metal, is so high you need a LOT of it to get enough delay for the ear to say "reverb". Hammond coiled up six+ feet (2m) of hard wire in the reverb tanks. A garden-hose air-delay wants to be 30 feet (10m) or more.

I think you'll have to eat a lot more candy.

[moid]

you get junk like this there in blighty, yes? must be worth a try, what with all the surface area to resonate. springs across, or from lid to base, away you go.

you'll need to remove all the internal damping material before use, tho.

[edit :] I forgot to add - sweet.

You read my mind Duck I'm now using a large old biscuit tin (octagonal for booteek style of course) It distorts gently, but no reverb. I did try one of my wife's large cake tins, but apparently she doesn't have the same opinion as me when it comes to audio experimentation so that one's out of the question (it didn't sound that good anyway). The octagonal biscuit tin was better, but no reverb. I think you may be onto something regarding springs though. I ordered some springs (an Amazon cheap box of assorted sizes but they aren't very large... I wonder if I can link multiple springs together?) I think I'll have to try. Do you think the springs should be under tension? Maybe not? In the images of spring reverbs I've found they seem to be glued to the centre dome of the speaker and I doubt that material (whatever it is) will take a lot of stress if the spring is stretched across to the other side of the tin... Hmmm I've got two types of spring in this box, compressor and extender - I'm guessing I should use compressors because they won't add much tension to the speaker cone.

For the sake of audio experimentation I attached the circuit I'd made to a radiator (well it's hollow and made of metal, so why not) and played guitar through the radiator. It's not a great sound and I couldn't get any reverb alas - it filtered out frequencies around 3Khz, but massively boosted ones around 4Khz... so its sounds like a rather unpleasant %^&*ed wah with a light distortion. I can't recommend radiators for their tonal properties - just in case anyone else is genuinely interested in future, I don't think it's worth the effort.

[moid]

Well, stainless because rust sounds awful. Galvanised/painted steel sheet can be found. Tin sheds are made of it. Corrugations can be rolled flat -  or left as is.

Have you a Wickes nearby?
https://www.wickes.co.uk/Wickes-Metal-Sheet-Galvanised-Steel-250-x-500mm/p/116718

More useful for casing/chassis, but...
https://www.wickes.co.uk/Wickes-Metal-Sheet-Plain-Uncoated-Aluminium---120-x-0-8mm-x-1m/p/116700

Thanks AnotherJim - that's a good point, I never realised that Wickes sold sheet metal. Yes I live near one

This entry might be even better (It's larger)
https://www.wickes.co.uk/Wickes-Metal-Sheet-Raw-Steel---600mm-x-1m/p/116705

now all I have to do is figure out where to store such a thing... time for a conversation with my wife

[moid]

The velocity of sound, in air or in metal, is so high you need a LOT of it to get enough delay for the ear to say "reverb". Hammond coiled up six+ feet (2m) of hard wire in the reverb tanks. A garden-hose delay wants to be 30 feet (10m) or more.

I think you'l have to eat a lot more candy.

Now don't tempt me Thanks for the info about the wire, that's really helpful. 2m sounds like a lot but it's not that large once it's all coiled. My main problem is now that I've bought some springs the longest is 70mm long and it has a 4mm diameter. I counted the amount of turns in the spring (105) which gives me 1.31m, so if I link two of these springs together that would give me more spring than the Hammond tank. The only issue is that these are extender springs so I don't know if that's the right thing to use or not - I suspect the coils shouldn't be touching, but I might be wrong?

Regarding the velocity of sound in materials, the velocity of sound in aluminium is far faster than it is in steel... now I'm wondering what metal those biscuit / sweet tins are made from?

[PRR]

> the velocity of sound in aluminium is far faster than it is in steel...

Not "far".

stuff               Metres/second
Aluminium    6300
Beryllium        12900
Copper          4700
Glass(crown)    5300
Iron                  5900
Nylon, 6.6         2600
Oil (SAE 30)   1700
Plexiglas         1700
Steel, Mild    5920
Steel, Stainless    5800
http://www.classltd.com/sound_velocity_table.html

6300 vs 5900.... the available or convenient sizes will make more difference than that.

BTW: for a long-time small-size delay you want *slow* speed of sound.

> now I'm wondering what metal those biscuit / sweet tins are made from?

The only likely stuffs are plastics, Aluminum, and steel. ("Tin can" is 99.9% steel.) A magnet will sort all common steels from anything else. (Stainless Steel and Nickel are very weakly magnetic but too expensive for candy.)