• TalkBass has been independent since 1998. Add your voice.
    Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
    Join free Log in
    Want zero display ads or expanded classifieds tools? Compare plans.

Double Bass How to make a piezo preamp

fdeck

fdeck

Supporting Member
Commercial User
Mar 20, 2004
18,674
21,418
Madison WI
sites.google.com
Disclosures
HPF Technology LLC
#1
Toneranger's pickup design piqued my interest. I had some of those piezo elements in my parts box, from some other project.

Note that the little vibration sensors in Toneranger's design are available from Mouser Electronics (Invalid Link Removed), part number 824-LDT0-028K/L, for $0.75 each. That's also where I get the parts for my preamp.

Speaking of which... here is yet another little preamp design. It's just about as simple as I could make it, and it seems to work quite well. For links to both of my preamps,

[Invalid or Expired Link Removed]

About my power amp project, it is officially dead. I decided that the draconian current limiting scheme used by the Tripath TA2022 switching power amp chip is a show stopper for a musical instrument amplifier. But it also occurs to me that there is very little to be gained from designing your own power amp, when there are lots of inexpensive offerings out there.
 
#2
fdeck said:
Toneranger's pickup design piqued my interest. I had some of those piezo elements in my parts box, from some other project.

Note that the little vibration sensors in Toneranger's design are available from Mouser Electronics (Invalid Link Removed), part number 824-LDT0-028K/L, for $0.75 each. That's also where I get the parts for my preamp.

Speaking of which... here is yet another little preamp design. It's just about as simple as I could make it, and it seems to work quite well. For links to both of my preamps,

[Invalid or Expired Link Removed]

About my power amp project, it is officially dead. I decided that the draconian current limiting scheme used by the Tripath TA2022 switching power amp chip is a show stopper for a musical instrument amplifier. But it also occurs to me that there is very little to be gained from designing your own power amp, when there are lots of inexpensive offerings out there.
Click to expand...

Beautiful, absolutely beautiful! Great design, explanation, execution, and description. Kudos.
 
#3
Nice one fdeck! That's definitely simpler than the ones I've tried previously and like you say, these piezos don't really need a gain boost. I'll give this one a try.
Have you tried adding a volume and passive tone control? If so, what value pots & capacitor would you suggest?
 
#5
elros,

My design is very much along the lines of Don Tilman's. His design has about 10 dB of gain.

ToneRanger, I would have to think about how to implement a tone control gracefully, but for a volume control, I would just hang a 10k audio taper potentiometer on the output (after the capacitor). It would be an asset to the design. Another thing that I think would be universally useful would be a mute switch, which you could do by simply shorting out the input. If you build this pre, let us all know how it sounds with your pickup design (while I scrounge in my parts bucket for those sensors, so I can try doing the same).

The reason why I bought some of those sensors was that I was helping a friend design a piece of industrial measurement gear, and we wanted a sensor that would indicate if the device was being subjected to excessive vibration. But I never thought of trying them on a bass.
 
#6
fdeck said:
elros,
My design is very much along the lines of Don Tilman's. His design has about 10 dB of gain.
Click to expand...

Don's notes say that with an input of 2.5v peak-to-peak, the output will be about 3.5v peak-to-peak. That's a facor of 1.4 or a gain of 2.9 dB. Did I miss something?

Speaking of gain, how could a modest amount be added to your design, if desired?

Thanks
 
#8
DRURB said:
Speaking of gain, how could a modest amount be added to your design, if desired?
Click to expand...


Fdeck's design is known as a source follower, the voltage at the gate (the terminal with the little arrow on it) is followed by the source (bottom terminal in this case). It can only ever have of gain of slightly less than 1.

Till's preamp is a common source amplifier, it distorts more and has a higher output impedance than a follower but can give gain. If using a high gm jfet and a high source resistor, gain is approximately -1*(Rdrain/Rsource).
 
#9
Tim__x and DRURB,

I assumed the -Rd/Rs equation for the Tilman circuit, but the source resistance is pretty low. The roughly 3 dB gain is for real. Now, how to increase the gain of my circuit? In a couple steps. First, turn it into Tilmans circuit with the addition of a resistor. With the 2N4393, equal source and drain resistances of 2.2k seem to model well in LTSpice. With the L201 transistor chosen by Tilman, his resistor values are appropriate.

Now, what to do about the gain? Increasing the drain resistance also increases the output impedance of the circuit. Not good. Lowering the source resistance disturbs the bias point, leading to degraded dynamic range.

Instead, take a 100-Ohm resistor in series with a 22 uF electrolytic capacitor, and wire this arrangement in parallel with the source resistor. It effectively lowers the source impedance, except at DC, so the biasing is preserved. The gain will be roughly 14 dB. Now you can also imagine making the extra resistor variable.

Clear as mud? Next chance I get, I will update my little web page with this information.

Too much gain is not a good thing, because a 9-V powered preamp will always have less headroom than the (typically) 30-V powered front end of a bass amp. Making it adjustable is probably worth the extra effort, so you can find the happy medium.
 
#10
Well, I finally got around to building fdeck's little pre-amp. I use a Rev. Solo pickup, an iAmp200, and a Wizzy cab. I tried the pre-amp both on a very short cable near the pickup-end and at the amp-end. The first thing I did was plug the output of the pre-amp into the "active" input of the iAmp. That was a total bust as there was not nearly enough output. It just has to go into the "passive" input of the iAmp. That is, it must go into the same input that the pickup does when it is used alone. Perhaps fdeck can comment on whether that poses some undesirable impedance match but, based on sound alone, that input is the only reasonable choice.

So what about the sound? Well, if anything, I was biased in favor of hearing some phenomenal difference. Actually, I was surprised at how subtle the difference was. With the pre-amp in the circuit, things do sound a bit "warmer." That is, the overall balance is tipped more toward the lower end and that makes sense. I listened very carefully to whether the very low frequencies were more intense with the pre-amp in the circuit than with it out of the circuit, as might be expected if the pre-amp were effecting a lower high-pass cutoff of the piezo's response. For the life of me, I could not hear any substantial difference. This suggests that the particular combination of the Rev. Solo and the iAmp comes pretty close to being optimal.

I'll continue to listen and play around but, for now, I can say that with my rig (and to my ears), the pre-amp provides a slight improvement. With other rigs, the effects may be more dramatic. Whatever the case, you sure can't go wrong given the low, low cost to build fdeck's little jewel.

Thanks again, fdeck!
 
#12
fdeck said:
Which input impedance did you choose? 1M or 10M?

Glad to hear the preamp worked for you. I am not disappointed that the improvement was modest, as it indicates that the iAmp is treating you well.

F
Click to expand...

I used the 10M as the iAmp input is 1M. I wanted the best chance to achieve a difference. As you may have read in my other posts, I was and am happy with the Rev. Solo and the iAmp to begin with. I just wanted to see if I could be even happier.
 
#14
Do you have the URL of a preferred Australian supplier? Then I could perhaps check out what they have against my design model. I ran tests on all of the standard JFET models in the LTSpice program, and found:

2N4416 is good
2N5432 is good
2N5484 is tolerable -- less dynamic range
2N5485 is OK
2N5486 is great
U309 should be OK

All of these might require you to use "PN" instead of 2N.

Best of luck. If you use one of these alternatives, please share your results with me.
 
#15
fdeck said:
Do you have the URL of a preferred Australian supplier? Then I could perhaps check out what they have against my design model. I ran tests on all of the standard JFET models in the LTSpice program, and found:

2N4416 is good
2N5432 is good
2N5484 is tolerable -- less dynamic range
2N5485 is OK
2N5486 is great
U309 should be OK
Click to expand...

So why does your design call for 4393? What am I missing?
 
#16
Ah, you caught me. I noticed the 4393 one day in a piezo preamp design at the Jensen Transformer website. It is a good design to look at if you don't mind a bit more complexity. You can use it without the transformer. The input noise is extremely low, owing to the relatively high drain current and high transconductance of the 4393.

To compound the issue, I was looking at the 4393 for something I was doing at my day job, utterly unrelated to bass or even audio.

And then I bought some 4393's for yet another project for a friend. Thus, when I decided to prototype the 4393, it was the "devil known," and I had some on hand.

Looking at different JFETs, each has a different source voltage when the gate is grounded in my circuit. You want this voltage to be such that the output covers a fairly wide range before clipping. But the 2N5486 has an even wider range, and it might be my choice if I were starting from scratch. The 4393 handles +/- 1.2 V input, but the 5486 handles almost +/- 4 V.

The lesser input voltage range of the 4393 is not a problem in the Jensen circuit, as it has 12 dB of gain, so the output would clip before the input does.
 
#18
ToneRanger said:
Thanks for the quick reply. I can source the 5486 locally - now I just hope the supplier hasn't closed down for a month like so many local businesses do at Christmas!
Click to expand...
I would not mind knowing how the 5486 performs. Don't forget to verify the pinout, because I don't know if JFETs are standardized.
 
#19
fdeck said:
Ah, you caught me. I noticed the 4393 one day in a piezo preamp design at the Jensen Transformer website. It is a good design to look at if you don't mind a bit more complexity. You can use it without the transformer. The input noise is extremely low, owing to the relatively high drain current and high transconductance of the 4393.

To compound the issue, I was looking at the 4393 for something I was doing at my day job, utterly unrelated to bass or even audio.

And then I bought some 4393's for yet another project for a friend. Thus, when I decided to prototype the 4393, it was the "devil known," and I had some on hand.

Looking at different JFETs, each has a different source voltage when the gate is grounded in my circuit. You want this voltage to be such that the output covers a fairly wide range before clipping. But the 2N5486 has an even wider range, and it might be my choice if I were starting from scratch. The 4393 handles +/- 1.2 V input, but the 5486 handles almost +/- 4 V.

The lesser input voltage range of the 4393 is not a problem in the Jensen circuit, as it has 12 dB of gain, so the output would clip before the input does.
Click to expand...



Thanks for the reply. Does this mean that I could sub a 5486 right into your simple piezo design (making sure of the pinout)?
I assume it's available from Mouser.
 
#20
fdeck said:
I would not mind knowing how the 5486 performs. Don't forget to verify the pinout, because I don't know if JFETs are standardized.
Click to expand...


Pinout for 2N5486 is the same as PN4393 called for in your design. See data sheet at:

Invalid Link Removed
 

Latest posts

Top Bottom
Back