Do I need a buffer or a DI or a preamp?

[Mbira]

Call me stupid.  ;D

I play an instrument using piezo pickups.  I usually play into my old fender super reverb, so the sound is good to me.  My understanding is these tube preamp stages can handle the high input impedance that the piezos crave.

I've been getting more inclined to just want to carry around a DI or a buffer instead of lugging around my amp-especially if the places we are playing already have a PA.

So...what are the pros and cons of :

a fet buffer vs a tube preamp vs a DI?  

I have the parts coming for an FET buffer and they should be here tomorrow, so I'll whip one together for tomorrow nights show...

Thanks!
Joel

 

[MagnetoSound]

So...what are the pros and cons of : a fet buffer vs a tube preamp vs a DI? 

The FET buffer offers very high input impedance, and a simple buffer with an unbalanced output could be fed to any subsequent amp or DI box with minimal fuss. It can also be made very small and light, and it's power requirement is very low so your 9v battery will last for ages ...

The tube preamp offers the possibility of euphonic colouration, much greater tone manipulation and signal balancing but none of the advantages outlined above ...

The DI box is primarily a balancing device. A passive (transformer) DI would not be likely to have the very high input impedance of the the FET buffer but will offer some colouration, whereas a good quality active DI might offer all the advantages of the FET buffer plus superior CMRR.

 

[abbey road d enfer]

A passive (transformer) DI would not be likely to have the very high input impedance of the the FET buffer but will offer some colouration, whereas a good quality active DI might offer all the advantages of the FET buffer plus superior CMRR.

Yes. Whatever Jensen or Radial claim, stay away from passive DI's

 

[tv]

Build a "two-batteries" (18V) powered jfet buffer with the cheap 2n3819 (has low "input" capacitance -- 8pF says datasheet). It's best to build this buffer right into your instrument, so that the pickup cables are as short as possible AND all shielding is done to minimize buzz/hum etc. Such a buffer will give you a "canned" good sound when executed properly - pretty much the same regardless the next stage's "input" impedance or cable length...

Experiment with the buffer input impedance to fit your pickups for "the best tone".


To connect to the PA, you will probably need a DI box anyway.

 

[helterbelter]

I once bought a Radial Engineering active DI, excellenty built, excellent tone, but that´s also what I expect when buying a DI of 200 euros....
But, I also have a JLM DI kit. In a direct comparison I did, the sound was simply identical. Since we´re on a DIY forum, I can really advise the JLM kit. I must add, I have the old version DI kit, the new one uses FETs. BTW, You´ll need to add an outputtransformer when powering it with Phantompower.

http://www.jlmaudio.com/JLMDI.htm

 

[AMZ-FX]

Joel,

There is a schematic for a simple jfet buffer for piezo pickups in my blog:

http://www.muzique.com/news/piezo-pickups/

Let me know if you decide to try it and need any help.

regards, Jack

 

[Mbira]

Thanks guys,
The buffer I'll slap together is based on Scott's: http://scotthelmke.com/Mint-box-buffer.html

 

[emrr]

Isn't there a schematic for hooking up the JLM DI via phantom power, as a stand alone box?  That's hard to beat. 

Ye olde passive transformer DI suffers from inherently low input Z.  They seem great for many other jobs, but not the normal job description.  If you need a transformer, you will be better off with an active unit driving a low ratio winding. 

 

[helterbelter]

Isn't there a schematic for hooking up the JLM DI via phantom power, as a stand alone box?  That's hard to beat.   

Yes, there is one for the older version DI kit, but it will be necessary to swap the 2604 opamp with a TLO72, and add a few resistors :

http://www.jlmaudio.com/JLM%20DI%20kit%2048v%20Simple.pdf

This was also covered on Kev's site, IIRC....

 

[AMZ-FX]

Thanks guys,
The buffer I'll slap together is based on Scott's: http://scotthelmke.com/Mint-box-buffer.html

A piezo pickup can put out a larger signal than a regular guitar pickup: up to 2v.  For a signal larger than about 0.65v, that buffer will distort as the gate-source junction becomes forward biased.  Even with the switch set to divide the input, there will be distortion on the peaks. Many players find jfet distortion pleasing, so it may not be not a problem, but the signal will not be as clean as if you used the buffer I linked previously, which can pass the 2v signal without significant distortion.

If you decide to go with the circuit that you linked, forget the 1pf... the gate-source capacitance is larger than that.  Also, I would change the 10k build out resistor to a lower value.... a value between 100R and 1k would be good.

Show us some pix when you get it built!

Best regards, Jack

 

[alexc]

I'm doing a diy hi-Z buffer+pre out of spare parts, using 2x 9V batteries

I want to do a single jfet buffer and a basic 4 transistor line amp to drive a cheap but surprisingly good sounding 1:1 traffo
into the line inputs of my desk.

Just starting to veroboard the circuit - all the mechanical stuff is done.

It's like a cheap, battery powered API pre!

 

[abbey road d enfer]

The buffer I'll slap together is based on Scott's: http://scotthelmke.com/Mint-box-buffer.html

A piezo pickup can put out a larger signal than a regular guitar pickup: up to 2v.  For a signal larger than about 0.65v, that buffer will distort as the gate-source junction becomes forward biased.  Even with the switch set to divide the input, there will be distortion on the peaks. Many players find jfet distortion pleasing, so it may not be not a problem, but the signal will not be as clean as if you used the buffer I linked previously, which can pass the 2v signal without significant distortion.
Best regards, Jack

I'm sorry, Jack, your analysis is flawed.
Signals above 0.65v will not turn the gs junction on, because the source closely follows the variations of the gate (that's what this circuit is designed for). That would be true only if the circuit was not powered, though.
The real issue is the following:
Depending on the pinch voltage of the FET, its source will stand  from 1 to 6V above its gate. For optimum output capability, it should be somewhat above 1/2 Vbatt. The mint box buffer, with a FET with 5-6V pinch voltage would fit this description. Your version has a voltage divider providing 1/2Vbatt bias to the gate of the FET; if you have an FET with low pinch voltage, the source will stand 1-2V above 1/2Vbatt. But if you use an FET with high pinch voltage, the source will be too positive, which may clip the positive half wave of the output, just the same as the negative half wave may be clipped if the FET has low pinch voltage and no positive bias on the gate. The proper implementation  should be with the voltage divider, but having the proper values to properly bias the FET at hand.
And I must say the mint box also has its flaws: I don't particularly like the 220k source resistor, which makes the effects of loading quite sensitive. And the variations regarding the type of FET seem to be devoid of serious analysis.

 

[AMZ-FX]

I breadboarded the circuit and tested it before I posted, so I'm only reporting my results. Maybe I had it laid out wrong, or it was a variation in my transistor, but I definitely got clipping at signals below 1v.

I have noticed a wide variation in gate voltages of 2N5457 transistors so it will be necessary to test several devices to get one in the range that passes the pickup signal without clipping.

regards, Jack

 

[skipwave]

Bo Hansen has a nice solution to adjust bias for a given FET:

http://web.telia.com/~u31617586/#simple FET impedance converter