FET BOY - HAMPTONE - JFET PREAMP

[somorastik]

Hello,

The system sugested to start a new topic, since all the other ones I tried to reply to are old. So here goes.

I am building a FET BOY, I made a change to the original HAMPTONE schematic, adding a 2k Multi-turn trimpot to the drain of the 2N5457.

Can anyone explain to me, how to set the two trimpots in each stage?

I am very confused from the various replies, I found on this forum.

PDF in attachmetns

Thanks!

 

[pucho812]

why did you add the trim pot or was it already there in the schematic? Anyway read through the threads that although are old should have those details if you did not add the trim pot

 

[Matador]

Ostensibly they would allow one to set the Q point of each JFET independently, based on their individual IDSS and VGS off parameters.

A "maximum signal swing" test should work well here:

1) Dial both pots to maximum resistance
2) You inject a fixed amplitude signal into the gate: say for example, 0.1V PTP.
3) Look at the drain signal, you should see a clean waveform, centered somewhere around half of the positive supply (this is handwavey, and depends on the actual JFET, but gets the point across)
4) Increase the amplitude of the test signal and watch the drain signal:  at some point, one side of the wave will "flatten" as it nears one of the rails.  You then lower the resistance of the drain point until symmetry is restored.
5) Iterate between 3-4 until both the positive and negative peaks clip by the same amount.  This should be close to the optimum bias point for *this* JFET in *this* position.  Note the signal amplitude as this will define the input-headroom of this stage.

Can anyone explain the action of R7/R13?  These look close to regular CCS stages for Q2 and Q5, but lack the fixed base voltage.

 

[bruce0]

Roughly quoting Scott Hampton's original article on R7/R13...

"The output impedance of the voltage follower is a function of the 47 kilo-ohm resistor on the base of the ZTX653 (Q3/Q6).  If this resistor is too big, the output will not have enough drive and will be clipped on one 1/2 of the waveform.  On the opposite end of the spectrum, if you make it too small, more power will be dissipated than is needed in the output devices."

As far as your R19 pot I think it is adjusting the bias point of Q1, (looking at the Hampton tube pre article (basically same circuit) I think this would let you underbias it "reducing gain and headroom" or take it past the linear operating point and make it distort.).

I can't find a reference but I think R5 was supposed to interact with the caps to reduce third order harmonics or something but that is just a vague memory.

 

[PRR]

> 2k Multi-turn trimpot to the drain

I'm sure you mean "source".

Didn't Hamptone publish build-notes? Are they still on-line?

Watch the drain voltage (or Q2 emitter, same thing) as you trim R19. Voltage goes up and down. If you (can) set it to +24V, or to zero, it won't work at all well. In fact the maximum undistorted output requires this point to be mid-way between the rails, 10V-14V, not real critical. Maximum gain suggests a lower working point, but max undistort output falls-off faster than gain rises, so there's no point going way low. FWIW, a similar tube stage in guitar amps biases to 70%, or about 18V here. So try 12V, 6V, and 18V. You won't hear the difference in max undistort output unless you smack it pretty hard.

> explain the action of R7/R13?

R8 is small. The base of Q3 will be just a Volt or two. R7 has a near-constant 23V across it. Assume R7 is set to 92K. It passes 23V/92K= 0.25mA. Q7's hFE multiplies that by a factor of 40 to 250. Pencil "100". Then Q3 collector passes 0.25mA*100= 25mA. And will not vary (much) as Q3 collector bobs up and down. It's a near-constant current source.

My objection is that a fool like me may start R7 turned to zero ohms, "infinite" current will flow, something will burn-out. And a high hFE part can't be turned-down below 62mA, or 3/4W in both Q2 and Q3, which seems like a lot.

But I know Hamptone picked this topology and values with extensive ear-work, so we go with his expertise.

 

[somorastik]

Oh thank you all for replies!

I forgot for a second about this topic I started, but made progress on the preamp! And photos as well!

I set 4V across the 100Ohm resistor at Q3 emitter, that is 40mA for one stage. Then I used the 2k trimpot at the JFET source to set the Q2 emitter as close to the middle as I could, for one transistor it was around 16V for the second one around 17V. I could not get any lower, I tried adding 1k and 2k7 resistors to the source, but with no success, (wanted to set the Q2 emitter at 12V->that is half of the supply voltage)
I used a lab supply with 24V set exactly. And I used an Agilent generator and Tectronix digital scope, and set the 2k trim so that the signal at the output of each stage would start clipping symmetrically.
The 10k log pot set to max gives a lot of gain and the signal is clipped a lot. Otherwise it was ok.

Anyway I have a verry limited acces to 2N5457 fets ( I have maybe six of them) and stores here in europe dont sell them.

If anyone would be kind enough to pick two of those 2N5457 so they would be close to what I need, I would pay the price!

Thanks!

 

[rock soderstrom]

Hi somorastik,

have you seen this?

http://www.musikding.de/2N5457

or

http://uk-electronic.de/onlineshop/product_info.php?products_id=1339&osCsid=a62e37e60c3eb6d8cb5e3e4f4d110cf1

Have fun! 

 

[somorastik]

All working now.

Files are here
http://r-amps.g6.cz/fetboy-microphone-preamplifier/

 

[flaheu]

Thanks for sharing your work. 

 

[CheapChip]

Hi,
I downloaded these files some time ago, but forgot to thank you for your work...  :-[

Just a question:
How many preamp boards can be supplied by 1 PSU board?
I'd like to make an 8 channels device but don't know how many PSU I have to make.

Thanks again!