Mod input impedace for piezo
- Thread starter ELM
- Start date
Help Support GroupDIY Audio Forum:
What model box is it ? It's not a "standard" DI. Its input is for a standard low impedance balanced signal.
True, but it is using a FET input op-amp, so if it is only going to be used for unbalanced input signal, you could modify this:
That will change the input impedance for the tip connector of jack/pin2 of XLR to 1M. The input is not balanced with that mod, and the gain is about 6dB higher.
I’ts a palmer di, probably intended for keyboards. I had to sign papers to get the schematic, hope they dont sue me for publishing..
Glad to see that i Is so easy. I will definetly try this! Thanks!
Last edited:
Yes. And easy to change to unity gain.
Many people suggest higher Zin values for piezos but opinions vary afaik.
murrayatuptown
Well-known member
If you have the ability to measure the capacitance of the piezo device, the input impedance interaction with the capacitance (corner frequency of the resulting high pass filter) can be determined, instead of hoping for good results.
That is the reason 20k input Z is likely too low for many, if not most, piezos.
That is the reason 20k input Z is likely too low for many, if not most, piezos.
The model is Pan 08.
I plan to use this for my plate reverbs. It would be a great idea to create a 100Hz hpf at the input when I have the chance.
I ran the numbers thru a hpf calculator and got 200k Ohm for 100Hz.
Will this affect sound in the high end?
Att. piezo specs.
I plan to use this for my plate reverbs. It would be a great idea to create a 100Hz hpf at the input when I have the chance.
I ran the numbers thru a hpf calculator and got 200k Ohm for 100Hz.
Will this affect sound in the high end?
Att. piezo specs.
Attachments
Last edited:
murrayatuptown
Well-known member
0) Plate reverbs...like those mattress-sized things I saw in the basement of a recording studio on a tour of same?
1) Re: the 10k resistor to ground at the first op amp (+) input being changed to 1M. Will that affect anything else? Are the op amps powered with bipolar supplies? I ask because a lowish resistance 50% voltage divider there is sometimes part of biasing an op amp for single-supply use. You show a balanced input which is not balanced after 10k became 1M. There was also signal voltage division of 50% at (+) input that is no longer present with 1M.
HF end? The resonant frequency of the piezo disc being nominally 9kHz. Most ceramic disc piezos have an hf resonance that might be part of what people don't like about piezo sound. But that may be from people who made random sensor choices.
1) Re: the 10k resistor to ground at the first op amp (+) input being changed to 1M. Will that affect anything else? Are the op amps powered with bipolar supplies? I ask because a lowish resistance 50% voltage divider there is sometimes part of biasing an op amp for single-supply use. You show a balanced input which is not balanced after 10k became 1M. There was also signal voltage division of 50% at (+) input that is no longer present with 1M.
HF end? The resonant frequency of the piezo disc being nominally 9kHz. Most ceramic disc piezos have an hf resonance that might be part of what people don't like about piezo sound. But that may be from people who made random sensor choices.
murrayatuptown
Well-known member
Not sure if this USA-EN link works everywhere. 1-1002150-0-M : MEAS PIEZO FILM SENSORS WITH LEAD
This is the largest polymer film ribbon piezo I've ever used, 11000 pF, roughly 20 x 150 mm. ~15 USD or Euros in 2023. They are normally free of resonances. Some of the smaller (~350 pF) ones with a brass mass at one end sometimes have/had a resonance freq.listed. they are intended for other mechanical detection purposes and not audio, but I have carefully removed the brass cylinder (and got distracted by other things & didn't use).
9 years ago certain sizes were available with shielded molded cables at significantly higher cost. Most commonly they have factory-riveted connections to the conductive ink that forms a plate on one side of the treated PVDF piezo polymer. I don't remember if the opposite side is also silver ink...may be because they are very thin. Either riveted solder tabs (*) or twisted pair AWG 28.
(*) You CAN solder the leadless tabs but a heatsink is encouraged...the polymer (or connection to...not sure) will fail if overheated (70 C?) Conductive epoxy also works, but probably not any more appealing than the discouraging properties described so far. But they work quite well. I have shielded with Teflon then copper tape.
Double-sided foam tape works well for attaching, unless you want to reposition.
TE (acquired Measurement Specialties, Inc.) are a bit lazy in showing one illustration for an entire product line (Newark/Farn-something/Element 14 is also bad about this)
Murray
This is the largest polymer film ribbon piezo I've ever used, 11000 pF, roughly 20 x 150 mm. ~15 USD or Euros in 2023. They are normally free of resonances. Some of the smaller (~350 pF) ones with a brass mass at one end sometimes have/had a resonance freq.listed. they are intended for other mechanical detection purposes and not audio, but I have carefully removed the brass cylinder (and got distracted by other things & didn't use).
9 years ago certain sizes were available with shielded molded cables at significantly higher cost. Most commonly they have factory-riveted connections to the conductive ink that forms a plate on one side of the treated PVDF piezo polymer. I don't remember if the opposite side is also silver ink...may be because they are very thin. Either riveted solder tabs (*) or twisted pair AWG 28.
(*) You CAN solder the leadless tabs but a heatsink is encouraged...the polymer (or connection to...not sure) will fail if overheated (70 C?) Conductive epoxy also works, but probably not any more appealing than the discouraging properties described so far. But they work quite well. I have shielded with Teflon then copper tape.
Double-sided foam tape works well for attaching, unless you want to reposition.
TE (acquired Measurement Specialties, Inc.) are a bit lazy in showing one illustration for an entire product line (Newark/Farn-something/Element 14 is also bad about this)
Murray
Last edited:
Not to complicate your "easy" solution, but changing just one resistor will imbalance the differential amplifier.
Changing all 4 of the 10k resistors to 510k ohm will give you a roughly 1M input z, and maintain the differential balance. The higher impedance resistors will contribute some Johnson noise but should not be problematic (so does the one 1M).
I generally target higher than 1M for piezos, but 1M is better than 10k.
JR
I was discussing lately about the use of piezos with a buddy who a luthier , more specifically the K&K type , as opposed to under saddle .
Double sided foam tape isnt recomended to mount these up to a soundboard or other vibrating surface , your looking for direct contact for maximum output . Superglue gel is what he uses , if it ever needs to come off sliding the edge of a razor blade under it will lift it without any damage .
Determining the correct position for best sound is a matter of choice and is probably best done by temporarily mounting up the discs on the outside of the soundboard with blu-tak .
Ive tried soldering onto the brass disks and often they end up distorted and no longer flat , so definately try and heatsink and dont dwell to long with the iron .
Double sided foam tape isnt recomended to mount these up to a soundboard or other vibrating surface , your looking for direct contact for maximum output . Superglue gel is what he uses , if it ever needs to come off sliding the edge of a razor blade under it will lift it without any damage .
Determining the correct position for best sound is a matter of choice and is probably best done by temporarily mounting up the discs on the outside of the soundboard with blu-tak .
Ive tried soldering onto the brass disks and often they end up distorted and no longer flat , so definately try and heatsink and dont dwell to long with the iron .
Thanks for all your input, lots of great info there!
Yes, crazy big.
It's bipolar.
I will definitely consider trying these!
This is maybe a good idea. The input is becomes balanced when ground is lifted, right?Not to complicate your "easy" solution, but changing just one resistor will imbalance the differential amplifier.
Changing all 4 of the 10k resistors to 510k ohm will give you a roughly 1M input z, and maintain the differential balance. The higher impedance resistors will contribute some Johnson noise but should not be problematic (so does the one 1M).
I generally target higher than 1M for piezos, but 1M is better than 10k.
JR
Yes, the mounting is critical. Mine is superglued to the plate. Before glueing i painted the disc with layers of clear lacquer to avoid groundloops.
I generally target higher than 1M for piezos, but 1M is better than 10k.
for sure, much much better than 10k
I use from 5M to 10M for Piezos
I only use 1M for electric bass/guitar passive magnetic pickups
What would benenefit with higer impedance than 1M soundvice? A bass roll off is actualy a plus in my case. Using resistors i have lying around would give me roughly 1,3M btv.
I will also try changing the 10uf cap after first stage for xxnf to create a hpf.
I have no idea whats your specific case.
The Bass rool off is not a Steep High pass filter at 80hz in this case. Sometimes the bass roll off starts at around 500hz or 300hz, so it's not Sub that you loose, you loose Low Mids (Body) and also Mids, not that interesting to me personally.
I prefer to have the full spectrum from the instrument and then in case needed I use an EQ to dial precisely and cut what I really don't want. I want to make that decision myself and not have it decided by the impedance of source and load.
As an aside, the "plate reverb" could be home-made and not so massive, like this one from an Ikea metal shelving unit:
The reason I mention this youtube vid though, is their cheap focusrite interface already has a 1Mohm high-Z input. The guy in the vid asked someone he knew in the technical department.
It might be easier to buy a new interface for this project, the one in this vid (or if not in this vid, the previous vid of the same project.
The reason I mention this youtube vid though, is their cheap focusrite interface already has a 1Mohm high-Z input. The guy in the vid asked someone he knew in the technical department.
It might be easier to buy a new interface for this project, the one in this vid (or if not in this vid, the previous vid of the same project.
