Mosfet Booster as a piezo buffer with 200hz scoop for a cajon

[hoyager]

Hey there, just wondering if anyone could run their eye over this?

I want it to be able to plug straight into a standard line in on a mixer at unity gain.

Would there be enough gain in the booster to offset the loss through the eq section?

Most of it looks ok to me, just not exactly sure whether that last 10k to ground after the eq should be there or before the eq.

The zeners are there for overs coming in from the piezo.



Andy

[bool]

Imo it would be better design practice to move the t-filter from the output to the point between the mosfet drain and the bjt base.

[kingswayguitar]

Imo it would be better design practice to move the t-filter from the output to the point between the mosfet drain and the bjt base.

That would be a great idea to minimize the output impedence!  The original post has one nice design factor that I like and that is the direct dc coupling of the mosfet to the bjt.  What does this puppy sound like?

[GibsonGM]

LT Spice might offer some answers about the gain, etc....

[hoyager]

Yes, although I'm pretty sure I could build it much faster than learning spice...

What is the advantage of the dc coupling between transistors?

Would the ouput impedance be too high for a 10k load (mixer) if the eq is last, as it is?

Andy

[hoyager]

Here's another buffer after the eq/filter. Still unsure of gains, but haven't built it yet.



Does the 12k from 9v to the last BC547 control the output gain, or am I missing a few steps?

Also would anyone know whether the output would be protected from 48v from a desk?

http://www.diystompboxes.com/smfforum/index.php?topic=50531.msg378167#msg378167

http://www.diystompboxes.com/smfforum/index.php?topic=56030.0;

http://www.geofex.com/Article_Folders/EQs/paramet.htm#simple_eq

http://www.muzique.com/schem/mosfet.htm
http://www.muzique.com/lab/buffers.htm
http://www.muzique.com/lab/gyrator.htm

http://www.till.com/articles/GuitarPreamp/index.html

Andy

[GibsonGM]

The gain there is ~ 12k/4.7k= 2.5      Collector resistor /  emitter resistor
Emitter resistor is related to impedance, tho, so it's a balancing act....

I suggested you start in with a simulator because then you can read your insertion loss in dB from the filter, and correct for it with your gain stage...the other way is to make sure you have enough gain on tap, and use a volume control on the output!    Something like the last stage of a Big Muff Pi may be great for what you're doing - at least worth experimenting with.

The 10uF cap on the output protects the output from DC, yes.  You'd want to make sure it had a ~ 100V rating, IMHO, if you might hit it with 48V.       

[hoyager]

Cool, thanks very much. The output stage here is pretty similar to the bmpi, (actually from the bandpass output of a roland 121 vcf) same concept though.

Would that last 10uf need the positive facing out against 48v? or can it be bipolar and/or not matter?

What about, say, continuous protection from 48v, does that change anything ie. if the phantom is left on?

I have it built up, will update.

This site has some good basic explanations

http://www.kitrae.net/music/big_muff_guts.html

and this calculator

http://en.transistoramp.de/



Andy

[PRR]

> whether that last 10k to ground after the eq should be there or before the eq.

I dunno what that 10K is doing at all.

The mixer will be 10K-22K, so you may not need more loading.

If it's there to bleed output coupling cap DC leakage.... if your cap is any good, the bleeder could be a lot bigger.

The bridge-Tee network is 128K for bass but near-Zero (5nFd) for treble. Loading it with 10K will let full treble through but cause 10K/146K= _23_dB_ loss for bass. Loaded, it is a low-cut, not a notch.

Using a simple JFET would avoid all those diodes. Even with a MOSFET you probably don't need three diodes.

"Overs" from the source can clip naturally; in fact *will* clip somewhere in a 9V circuit before +/-5V clippers wake up.

The Gate trimmer plan you show does not cover all cases. 4.5V is not magic; could be very wrong for a low Vgs MOSFET. Your plan seems to supply higher voltages (to 9V!) but not lower.

You do not say how much gain you need. I know you don't know yet, but would be good to know if this "piezo" source is strong or weak (by musical instrument standards). Also if you expect needle-bending levels into the board, you can't do that with 9V. (You can however deliver 0.3V which any board will handle well with the input trim.)



Personally I'd lean to a TL072 solution. 10meg or 100Meg is fine for TL072. Two fancy-amps can isolate source, EQ, and load, whereas two discrete devices are always reflecting source and load impedances forward and back.

[hoyager]

I see what you're saying about the zeners vs power supply limit, thanks.

The 10M to Vr is straight from the mosfet booster, surely thats not a problem with the original design? I've just added a trimpot in place of the usual 62k, the clipping diodes and an extra cap over the inputs

As far as I understand (which isn't that far) the 10k to ground at the output set the output impedance to 10k... but I see how that changes the filter / eq.

Also AFAIHR (as far as i have read), piezo's can be quite hot, into the multiple volts.

What I am keen on is a bit of slight overdrive when the drum is getting hammered, I think that would even the dynamics out a bit and have a positive affect on the sound. This is just for a live music application so it doesn't need to be super clean.

I've just discovered all the jfet and buffer designs on AMZ, so lots of good stuff to go through, and lots of options that would work too..

http://www.muzique.com/tech/bipolar.htm
http://www.muzique.com/tech/bipolar-m.htm
http://www.muzique.com/tech/bipolar-j.htm

Also read that a piezo buffer into a tube amp (with high input impedance?) sounds really good, so am tempted to follow the mosfet booster (or other piezo buffer) with one of the jfet input stages attempting a tube amp emulation, then eq then buffer.... might be too complex, but also might sound really good.

Anyhow, I will know soon enough.
Cheers
Andy

[Gus]

Did you use "piezo" in the search at this forum?

[hoyager]

For sure. Although searching for "piezo mid scoop" only gets one hit

http://www.diystompboxes.com/smfforum/index.php?topic=61294.0

"piezo cajon" returns this thread...

Andy

[hoyager]

Well, it works well, based on the second schematic I sent.

Thanks for the pointers and help.

I'll upload a final version and sound clips after testing with a cajon and the right piezo.

Evidently piezo's don't have that much output (or at least the 50mm one I'm testing with), I had to adjust the gain in the last stage quite alot, and the gain in the mosfet stage is near max. Result is a bit of hiss, but I've got quite a few sockets on the board and trimpots which can be replace with short metal resistors, which may help.

Mid scoop works well, un-muddies the sound, and sounds much flatter in response than without.

Max output seems to be very near 0 (+4dbm) with unity gain on the mixer input (Mackie DBX12)

Andy