This seems to be musical instrument related and should therefore be in I&A where I will move it.
The key thing with piezo elements is that they must work into very high impedances.
Line implies not just signal strength of the order of one volt, but medium to low impedances.
Instrument level seems to mean different things to different people, but usually higher impedances than line. You might have some luck with some high-z guitar amp inputs, though not all.
Often, piezos have a preamp as close to the pickup as possible, to avoid capacitive losses of high frequencies in cabling, and interference pickup.
A simple fet source follower would be a good starting point.
As Kevin asks, what's it for?
Line implies not just signal strength of the order of one volt, but medium to low impedances.
Instrument level seems to mean different things to different people, but usually higher impedances than line. You might have some luck with some high-z guitar amp inputs, though not all.
Often, piezos have a preamp as close to the pickup as possible, to avoid capacitive losses of high frequencies in cabling, and interference pickup.
A simple fet source follower would be a good starting point.
As Kevin asks, what's it for?
Hi,
Typical piezo elements are instrument level and can work directly into high
impedance instrument inputs, as on guitar amp, but don't like long leads.
Battery powered preamps built into guitars allow EQ etc and the ability to
drive long leads but don't change the level, its still suitable for a amp input.
Best arrangement I found years ago was a shortish lead to an acoustic
guitar multi effects pedal and then on to the amplifier. Undersaddle pickup
with the socket in the endpin the best way of converting an acoustic.
rgds, sreten.
Typical piezo elements are instrument level and can work directly into high
impedance instrument inputs, as on guitar amp, but don't like long leads.
Battery powered preamps built into guitars allow EQ etc and the ability to
drive long leads but don't change the level, its still suitable for a amp input.
Best arrangement I found years ago was a shortish lead to an acoustic
guitar multi effects pedal and then on to the amplifier. Undersaddle pickup
with the socket in the endpin the best way of converting an acoustic.
rgds, sreten.
Except high impedance instrument inputs aren't really anywhere near high enough (only a few 100K usually), raw piezos need multi-megaohm inputs.
A simple FET buffer is all that's required, with perhaps a gain stage as well, if you need to bring the piezo up to a similar level as other instruments?.
1M is a common input impedance for (passive) instruments, but yeah, ideally they're looking at several megaohms of impedance, because you loose low-end otherwise.
The output of a piezo element can be pretty large, depending on how hard and how you compress/bend it (and the type of piezo element of course). You can expect several volts of output, so you need a preamp which has some overhead.
The output of a piezo element can be pretty large, depending on how hard and how you compress/bend it (and the type of piezo element of course). You can expect several volts of output, so you need a preamp which has some overhead.
This is a pickup in the bridge of a cello. I'm aware of the fact that the preamp should be close to the pickup, along with the problems with interference and signal degrade over long cables.
For preamps, I was looking at a modification of the Tillman preamp, which would offer about 14dB gain, or the Ritter preamp, which gives about 30dB gain. However, I'm not sure what sort of tone the Ritter preamp offers.
For preamps, I was looking at a modification of the Tillman preamp, which would offer about 14dB gain, or the Ritter preamp, which gives about 30dB gain. However, I'm not sure what sort of tone the Ritter preamp offers.
It shouldn't have any sort of 'tone' (which isn't really an electronic specification anyway), it should pass the piezo signal through unchanged, apart from it's size.
Back to the guitar amp thing, years back a couple of friends used to do discos, and they used a guitar amp fed directly from ceramic PU's - it souinded CRAP!!. I agreed to meet them one night at a disco, and I built a pair of FET source followers in a tobacco can, powered by a PP3 9V battery, and stuck it in my pocket. I went there straight from work, and the preamp changed the sound from 'crap' to 'absolutely amazing'
This still sticks in my mind, because it was a disco for my old school leavers party, and while they were eating we asked at the bar if they did bar snacks (seeing as I had come direct from work, no food!). They didn't, but asked if we were with the disco, and said they would sort something out. A few minutes later they took us through to the kitchens, where we were given bowls of chicken soup. The soup was absolutely beautiful - and Dave (the older of the two disco brothers) said "this must have been a good can". At this a chef in a tall white hat went bright red, grabbed a meat clever, and charged across the kitchen towards him. Four of five of the other staff jumped on him, pinned him to the floor, and took the clever off him.
So it's kind of stuck in my mind ever since, even though it was back in the 70's
You need to find out what signal level you're going to be getting out of your pickup.
It needs current gain, yes, so it can drive amp inputs and cables ok.
If you've got a scope and a 10x probe, attach that, play, discover output.
Without a scope and probe, try this circuit, three resistors, two caps, one n-channel j-fet.
It's a simple source follower. Vth of fets varies, so for max headroom select the source resistor R2 to put the source roughly half way between the power rails - ie about 4.5v. I suggest you start with 47k. If the voltage is lower than this, increase the resistance, if higher, decrease it, no lower than say 3k3.
Play it, see how loud it will go. If it isn't loud enough, then you need some voltage gain.
It needs current gain, yes, so it can drive amp inputs and cables ok.
If you've got a scope and a 10x probe, attach that, play, discover output.
Without a scope and probe, try this circuit, three resistors, two caps, one n-channel j-fet.
It's a simple source follower. Vth of fets varies, so for max headroom select the source resistor R2 to put the source roughly half way between the power rails - ie about 4.5v. I suggest you start with 47k. If the voltage is lower than this, increase the resistance, if higher, decrease it, no lower than say 3k3.
Play it, see how loud it will go. If it isn't loud enough, then you need some voltage gain.
That's essentially the source follower I was building back in the 70's, 47K would be FAR too high, I can't remember exactly what value I used, but it was only in the low single kilo-ohms, probably 1K or 2.2K
But as you say, select it for about 4.5V on the source - but no need to get too carried away, or you'd have to make it adjustable - just get a reasonable value (for a specific bunch of FET's), then stick to it.
There is something of a "how long is a piece of string" to all this.
I am surprised at the low values being suggested here for Rs - my schematic was for a source follower, with the assumption that it would be used to drive a hi-z instrument amp input - I wonder if there might perhaps be bit of confusion about exactly what type of circuit needs what value of Rs? What purpose is served by using a value as low as 1k or 2k here? The higher the standing current the shorter the battery life.
I've just breadboarded a couple of 2N3819s to see how they behaved in a source follower like the one I suggested.
Testing the first two out of the bag:
with Rs = 2k2, Vs = 1.21v & 1.33v
with Rs = 47k, Vs = 1.59v & 1.73v
Checking the datasheet for 2N3819, the range of Vgs for Id = 200uA is quite broad, -0.5v to -7.5v. The 2N5487/8/9 appears just as variable as the 2N3819, and my datasheet for the J201 again doesn't give Vgs range for a useful standing Ids, but it does give cut-off Vgs in the range -0.8v to -4.0v, with Idss in the range 0.2mA to 1.0mA
My recommendation of trying to get the source up to anything like 4.5v is infeasible for the 2N3819s in my current batch. I was wrong to have suggested it - fet threshold voltages generally are too variable from batch to batch. Without the luxury of selecting devices from multiple batches, specifying a target Vs for a source follower is unwise. Yes, ideally we'd like that much headroom, but we probably won't get it and might not need it anyway.
Two questions thus far unanswered are
"how much gain, if any, do you need, to drive the amp you're going to be using?"
and more fundamentally
"what sort of amplifier input do you want to drive?"
If you just want to cut to the chase, and assuming you want to drive a high z instrument amp input, which seems the most likely case, I'd recommend you go for the standard Till type preamp:
Discrete FET Guitar Preamp
It's not a source follower, but its output impedance is quite low enough for any reasonable cable and instrument amp input. It has modest gain, about 6.8/2.2 = about 3x, or 9.5dB. You can always think about the hot rod version later.
R4 is simply a bleed resistor to keep the output side plate of C1 at ground potential, so that plugging into an amp doesn't produce a nasty crack. 51k is fine, as is 220k. Its value is not related to that of R3.
I am surprised at the low values being suggested here for Rs - my schematic was for a source follower, with the assumption that it would be used to drive a hi-z instrument amp input - I wonder if there might perhaps be bit of confusion about exactly what type of circuit needs what value of Rs? What purpose is served by using a value as low as 1k or 2k here? The higher the standing current the shorter the battery life.
I've just breadboarded a couple of 2N3819s to see how they behaved in a source follower like the one I suggested.
Testing the first two out of the bag:
with Rs = 2k2, Vs = 1.21v & 1.33v
with Rs = 47k, Vs = 1.59v & 1.73v
Checking the datasheet for 2N3819, the range of Vgs for Id = 200uA is quite broad, -0.5v to -7.5v. The 2N5487/8/9 appears just as variable as the 2N3819, and my datasheet for the J201 again doesn't give Vgs range for a useful standing Ids, but it does give cut-off Vgs in the range -0.8v to -4.0v, with Idss in the range 0.2mA to 1.0mA
My recommendation of trying to get the source up to anything like 4.5v is infeasible for the 2N3819s in my current batch. I was wrong to have suggested it - fet threshold voltages generally are too variable from batch to batch. Without the luxury of selecting devices from multiple batches, specifying a target Vs for a source follower is unwise. Yes, ideally we'd like that much headroom, but we probably won't get it and might not need it anyway.
Two questions thus far unanswered are
"how much gain, if any, do you need, to drive the amp you're going to be using?"
and more fundamentally
"what sort of amplifier input do you want to drive?"
If you just want to cut to the chase, and assuming you want to drive a high z instrument amp input, which seems the most likely case, I'd recommend you go for the standard Till type preamp:
Discrete FET Guitar Preamp
It's not a source follower, but its output impedance is quite low enough for any reasonable cable and instrument amp input. It has modest gain, about 6.8/2.2 = about 3x, or 9.5dB. You can always think about the hot rod version later.
R4 is simply a bleed resistor to keep the output side plate of C1 at ground potential, so that plugging into an amp doesn't produce a nasty crack. 51k is fine, as is 220k. Its value is not related to that of R3.
Hi JGAN, no worries.
I've had vague ideas about using piezo pickups on various musical oddities for ages now, and your thread gave me a push to actually do a little bit of thinking, as opposed to day dreaming, so thank you.
I think in future I'm going to add a couple of jfets every time I order components, so as to try to get a selection from different batches, though I bet I'll end up with loads from the ends of the Vth spread and none from the middle...
3M3 will be easier to get hold of than the 3M0 specified and should be fine. If the bass frequencies show some drop off, bump that up to 4M7, it'll depend on the capacitance of the piezo element.
If you do select devices for the standard circuit, try to end up with roughly, half the supply voltage dropped across the fet.
If you get stuck or want to ask about something, feel free to do so, here, pm, or email. Be interested to hear how you get on.
I've had vague ideas about using piezo pickups on various musical oddities for ages now, and your thread gave me a push to actually do a little bit of thinking, as opposed to day dreaming, so thank you.
I think in future I'm going to add a couple of jfets every time I order components, so as to try to get a selection from different batches, though I bet I'll end up with loads from the ends of the Vth spread and none from the middle...
3M3 will be easier to get hold of than the 3M0 specified and should be fine. If the bass frequencies show some drop off, bump that up to 4M7, it'll depend on the capacitance of the piezo element.
If you do select devices for the standard circuit, try to end up with roughly, half the supply voltage dropped across the fet.
If you get stuck or want to ask about something, feel free to do so, here, pm, or email. Be interested to hear how you get on.
Last edited:
No problem Simon,
I actually forgot to answer your questions posted earlier Now, to preface this, I'm not a rock/electric musician of any sort, so this is my first project with pickups, preamps, mixers, amps, etc. I'm just going to put it in layman's terms, because that's about as good as I can get...
I'm not driving a power amp, just one of those regular guitar amps.
Not exactly sure what that means, but 1/4" TS? If you're talking impedances, then I have no clue.
For impedances, I read that ideally the preamp input impedance should be 10x the pickup impedance. So if I do the Tillman preamp, could I use a multimeter to measure the impedance of the pickup, multiply by ten, round up, and then change R1 to that value?
And one last question
I would like to add a power-on LED + appropriate (220 ohm) resistor to the circuit. Should I put these between the +9 volts and ground, parallel to the C2 cap?
I actually forgot to answer your questions posted earlier
Now, to preface this, I'm not a rock/electric musician of any sort, so this is my first project with pickups, preamps, mixers, amps, etc. I'm just going to put it in layman's terms, because that's about as good as I can get...I'm not driving a power amp, just one of those regular guitar amps.
Not exactly sure what that means, but 1/4" TS? If you're talking impedances, then I have no clue.
For impedances, I read that ideally the preamp input impedance should be 10x the pickup impedance. So if I do the Tillman preamp, could I use a multimeter to measure the impedance of the pickup, multiply by ten, round up, and then change R1 to that value?
And one last question
I would like to add a power-on LED + appropriate (220 ohm) resistor to the circuit. Should I put these between the +9 volts and ground, parallel to the C2 cap?
Last edited:
A regular guitar amp will have a high impedance (z) input, the standard till circuit should drive it fine.
Instrument pickups usually, though not always, work best when they're lightly loaded - the thing they're driving is higher z than they are.
Your piezo element should give a reading of infinity when checked at dc with a meter - simplifying a bit, its impedance is like a capacitor - it's a complex impedance - not just R, but C too. Its value depends on the frequecy you're measuring it at. The basic formula for the reactance of a capacitor is:
Xc = 1 / (2 x pi x f x C)
C is in Farads, with f = 0, Xc is infinite
You could add a power on led, but I really really wouldn't! Why? Because a reasonable drive level for an led is about 20mA, ie at least 40 times the amount of current the preamp will draw!
Btw, your 220R resistor will give about 30mA with 9v supply, very bright, may well fail quickly.
If you end up with say a standing Id of 0.5mA, a PP3 battery should last about 1200 hours, ie 50 days. Put and led in at 20mA and you're down to 30 hours....up to you though!
Instrument pickups usually, though not always, work best when they're lightly loaded - the thing they're driving is higher z than they are.
Your piezo element should give a reading of infinity when checked at dc with a meter - simplifying a bit, its impedance is like a capacitor - it's a complex impedance - not just R, but C too. Its value depends on the frequecy you're measuring it at. The basic formula for the reactance of a capacitor is:
Xc = 1 / (2 x pi x f x C)
C is in Farads, with f = 0, Xc is infinite
You could add a power on led, but I really really wouldn't! Why? Because a reasonable drive level for an led is about 20mA, ie at least 40 times the amount of current the preamp will draw!
Btw, your 220R resistor will give about 30mA with 9v supply, very bright, may well fail quickly.
If you end up with say a standing Id of 0.5mA, a PP3 battery should last about 1200 hours, ie 50 days. Put and led in at 20mA and you're down to 30 hours....up to you though!
How do I calculate C then? Or should I just stick with 3M ohms? I feel like I should use an oscilloscope for this... haha
I understand that using a LED will greatly diminish battery life, but tbh I don't really mind. And the bulb does have a current of about 30mA, so yes it will drain even faster (12 hour lifespan).
I understand that using a LED will greatly diminish battery life, but tbh I don't really mind. And the bulb does have a current of about 30mA, so yes it will drain even faster (12 hour lifespan).
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.