question for RG? where to connect RC network from transformer secondary

[tempus]

Hey all;

I've been re-thinking my last schem for a balanced out to PA circuit for my piezo. I think it should look like this:



The other option I was wondering about was connecting the ends of the RC network to the 2ndary centre tap. Does anyone know which is correct?

Thanks

[R.G.]

I don't know that either one is more correct. If the tolerances of the Rs, Cs and the balance on the secondary winding were all perfect, there would be no voltage between the center tap and the "ground" symbols anyway. So whether the CT is connected to ground or not is theoretically immaterial. In real practice, if the parts are unbalanced or the CT is unbalanced by any significant amount, it would shove the winding to one electrical side or the other, increasing offset at the PA input, which is designed to reject that anyway. The differences should be very small - I think.

However, is your piezo pickup buffered inside your guitar? That setup will present a load of around 10K to your guitar, and ordinary pickups will lose treble from the loading. A raw piezo will lose a huge amount of bass. But if the piezo pickup is electronically buffered, all may be well.

[tempus]

Thanks for the clarification RG.

So in this situation, I would leave the 2ndary CT unconnected. I am a little unclear about the RC network, though. I realized that I shouldn't connect this to the guitar ground, so I figured the place to connect it would be the PA ground (which of course is not connected to the guitar ground), but only by default. I still don't really see how it will do anything, since the 2ndary signal is not connected to the PA ground in the first place. I trust you though 

Oh and yes, the signal will be buffered at the guitar with this:



Except the MPF 102 has been replaced by a J201.

Thanks again

[R.G.]

So in this situation, I would leave the 2ndary CT unconnected. I am a little unclear about the RC network, though. I realized that I shouldn't connect this to the guitar ground, so I figured the place to connect it would be the PA ground (which of course is not connected to the guitar ground), but only by default. I still don't really see how it will do anything, since the 2ndary signal is not connected to the PA ground in the first place. I trust you though 

Actually, the secondary signal is connected to the PA ground, just through the internal impedances of the PA inputs if nothing else. If you don't connect the Rs and Cs to the PA ground, then you could theoretically get more hum, although chances are it will be minimal. As I said, I don't think it matters much. If you leave the Rs and Cs connected to each other and just quit calling it "ground" where they connect to one another, that works too. In that case, it's purely a load across the secondary winding. It's really splitting hairs for differences on this one.

However, you do know that your guitar signal is going to be reduced dramatically, don't you? That's probably a 10K to 600 ohm transformer. That means the signal will be reduced by the square root of the impedance ratio. In this case it's a factor of four smaller as measured in volts.

Oh and yes, the signal will be buffered at the guitar with this:

Except the MPF 102 has been replaced by a J201.

Good. That will help. The 1.5K drain resistor will do fine. But I have a question - have you already tried that? It looks to me like the gate of the JFET has no DC path to ground, so it will float to the same potential as the source, and the working current will be set by Idss. That may cause quite a bit of distortion if you drive it from a high output piezo. I haven't dug too deep, but that's kind of what I'd worry about.

[tempus]

However, you do know that your guitar signal is going to be reduced dramatically, don't you? That's probably a 10K to 600 ohm transformer. That means the signal will be reduced by the square root of the impedance ratio. In this case it's a factor of four smaller as measured in volts.

The tranny I'm using is here:
http://www.edcorusa.com/Products/ShowProduct.aspx?ID=304

The plan was to send a balanced out to a mic input on the PA, so I think the voltage reduction should be acceptable (necessary even - although hopefully it's not too low).

But I have a question - have you already tried that? It looks to me like the gate of the JFET has no DC path to ground, so it will float to the same potential as the source, and the working current will be set by Idss. That may cause quite a bit of distortion if you drive it from a high output piezo. I haven't dug too deep, but that's kind of what I'd worry about.

I have tried this circuit, and it works fine. There is definitely some distortion with a magnetic pickup, but I can't hear any distortion at all from my piezo, no matter how hard I pound the strings, so I'm assuming the piezo has a lower output. I don't have anything to actually measure with, but it sounds fine to my ears. I arrived at this circuit by paring down different versions of buffers I'd seen. I was planning to mount it inside a phone plug (although I may just try to get it inside the guitar somewhere), and I was hoping to get away with a minimum amount of parts because of size limitations. After reading your comment, though, it may be worthwhile trying to drop a resistor to ground at the gate, although it'd have to be pretty big (10 meg or so) because I'm pretty sure the output Z of my pickup is 1.3M at 100 Hz. Come to think of it, that may be why I dropped that resistor in the first place.

Thanks again RG

[R.G.]

The plan was to send a balanced out to a mic input on the PA, so I think the voltage reduction should be acceptable (necessary even - although hopefully it's not too low).

Cool - if the lowered voltage is OK, great.

I have tried this circuit, and it works fine. There is definitely some distortion with a magnetic pickup, but I can't hear any distortion at all from my piezo, no matter how hard I pound the strings, so I'm assuming the piezo has a lower output. I don't have anything to actually measure with, but it sounds fine to my ears.

Again, great. If it works, it works!