IDEAS: "a differential discrete jfet input stage"

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

tv

Well-known member
Joined
Feb 22, 2006
Messages
1,459
As in title.

Explanation: I'm looking for as-simple-as-possible DC-coupled jfet differential input stage idea that would work well with guitar pickups, connected in floating configuration. (sort-of making an "active pickup" from a vintage one)

Apart from the textbook 5(4)-resistor/2-fet circuit, does somebody know of some tricks employed by likes of Duncan, EMG, MEC ... in their "active pickup preamps", which are usually said to be "discrete floating differential". Meaning 2fets differential ampli imho, for a build the 2N3819 would be fine I guess.

Of course I'd prefer a "served" professional circuit that would work with dead batteries :) , but in reality, any insight would be ***fine***. Especially what do those builders do differently versus the textbook circuits ...

Or a schemo ***if exists***!
 
You can pretty much drop FETs in place of bipolars for the input LTP of most DOAs and roll your own. They will have lower transconductance so stability compensation is somewhat easier.

DC offset will be the harder part as even matched duals are not as good as bipolars.

JR
 
DC offset will be the harder part as even matched duals are not as good as bipolars.

That's what I'm interested in most: EMG and the likes must have developed a trick to devilishly self-bias the preamp to withstand working conditions from 6 - 18volts or so. What is that trick?

Soldering the pickup wires between two fet gates and putting a couple of pF cap there isn't much of a problem ...

Anybody got a schematic ... emg-duncan-mec style?

-------------------------------
WAIT A SECOND:
-------------------------------
"DC-coupled" in OP may be misleading to readers: what I meant is that the PICKUP is connected directly to both fet gates, in a floating manner, without any capacitors in series whatsoever.

Output from the pre would naturally be leeched via an elco or tantalum ...
 
The VANILLA diff-amp with 2n3819 is here:



Obviously, as I said, any performance-improving ideas or info on "how pro's do it" are WANTED.

I think my idea on shunting the input with small caps is a good one. Wires from the PU would have to be shielded with a grounded shield and braided, but otherwise I think I'm on the right track ???

P.s.: somebody test it before I blow up MY gear :razz:
 
Instrumentation 3-opamp configuration with TL074.

There has to be some DC return on the inputs. 10Meg to 1/2 Vcc ought to work.

I don't see any essential reason the inputs must be DC-coupled to the buffer. The DC output of a wound coil pickup is dead-nuts Zero. With 1Meg input resistors you only need 0.01uFd to reach below the lowest note of a string instrument.

I'm not real sure there is any great reason to be true Floating (which the common 4-R diff-amp isn't). Your leads are short. Unless purpose-designed, the pickup might be floating but is NOT Balanced: one lead is the outside of the coil, the other lead is the inside against the core/magnet, the two stray capacitances must be very different.

DC offset is a non-issue. Audio gain does not have to be high. DC gain can be less with a cap or two in the loop. You admit the output is cap-coupled. The only real effect of DC offset at the output is to reduce audio output swing. And the usual output from "active pickups" is much less than your proposed 6V supply.

> what do those builders do differently versus the textbook circuits ...

I don't know the specific products. But what often happens in Product Design is you write the advertising bumph first, and then toss-up a few parts which mostly don't disagree with the ads.

> sort-of making an "active pickup" from a vintage one

That's different from what Duncan and EMG are selling: coils and amps designed and sold together.

And FWIW, I can't find any bumph on their sites saying DC coupled or floating.... do you have a clue?

From the specs, they could be one low-Idss FET, a resistor, and a cap.

> prefer a "served" professional circuit that would work with dead batteries :)

The battery drains claimed on the few active pickups I found will be very hard to meet with generic RF-JFET parts.
 
> The VANILLA diff-amp with 2n3819 is here:

Yeah, that's what I mighta posted if not so lazy.

R2 is probably not needed but costs little and does no harm.

Build it.

> my idea on shunting the input with small caps is a good one. Wires from the PU would have to be shielded with a grounded shield

Well, 18 inches of shielded cable gives about the same 47pFd you show, likely a better balance than small-pFd caps, similar or lower part-cost and 4 less blobs to trim and solder. Cable does add hot-hot capacitance which 2 caps don't, but the coil is full of end-end C already.
 
http://www.schecterguitars.com/schematic/hb105.htm

http://www.zzounds.com/item--SEMAHB1

(with the blah-blah)
They use unbalanced inputs in a differential preamp. The problem is, an unbalanced differential preamp is not very effective at canceling hum. Seymour Duncan engineers figured out how to capture the tone that players want in an active design, but using balanced inputs.

The result is that not only do these Seymour Duncan Blackouts active humbuckers have less hum than the competition, but they also have more lows, more highs, and more output. Simply put, the Blackouts have more tone than other active pickups. Players will notice the benefits of the reduced hum when they record, especially if they’re around computer monitors or fluorescent lighting. As for the benefits of having more tone, well those will be apparent everywhere.

From my pickup-buffering experience, there is some difference in tone if a capacitor is IN SERIES with the input device even a superhigh-impendance mosfet.

Plus it is cleaner, simpler and in my jfet config on 0-VOLT potential meaning FAR LESS CHANCE for shit to happen when not wanted.

Basically I want ALL THE TONE AND SUSTAIN from the pickup and kill all RF, hum and other electronic noise while at it.

Yeah, that's what I mighta posted if not so lazy.
Wow I'm good.
Build it.
WEELLLLLLLL not so fast :wink: While I think it JUST might work, I'll sleep over it and wait for some more input.

Especially regarding autobiasing (if exists) or if somebody actually has the info on how the Big Boys do it.


AHA - and there I was thinking on how to drop a BJT follower @ the output. Or something smarter that would have LOW output impendance while keeping the circuit below 1mA, ideally @ 250uA (I should raise the 10K resistors @ 47k ... ).


-------------------- IMPORTANT --------------------
The "not really balanced pickup" has bothered me before! Now I don't know if this can be skipped or how big impact could it have, but I think that the basic concern is to minimize the buzz and other shit that pickup may collect. This shit is even more pronounced if pickup goes straight into a very-hi-Z input. So there ...

the OTHER important thing which bothered me is HOW to do the balanced trick with singlecoils? ...
 
Yet another permutation of the basic circuit.



Changes:
- drain and source resistors enlarged to draw less battery juice (hmmm ..)
- added T3 as a emitter follower buffer
- T1 drain resistor removed (objectionable ???)
- T2 drain resistor bootstrapped for more even load.

While the first version was somewhat simpler than a balanced fet-opamp version would be, I'm not so sure about the "badass" version ... it should kick ass versus an opamp version in order to build it.

Although the major bonus is that pickup is @ 0 VOLTS with the discrete version NO MATTER WHAT.



Can somebody throw this into a SIM or just plain review? Values were ***eyeballed***(TM) in the drawing process.
 
I would go this way:

sympickupamp.gif
 
Cable does add hot-hot capacitance which 2 caps don't
There's two 47pf caps going hot-ground @ each lead and a 330pf cap going hot-hot (between fet gates). I have seen some opamp differential pickup-pre's (Carvin use them in some models, f.e.) and in most cases, they have said 330pF cap per pickup (***but not sure if this is for humbuckers or singlecoils).

Still, if somebody is in posession of duncan-emg... active pickup differential fet preamp schematic, GIVE'R AWAY !!!
 
With the current drain specs, I doubt seriously there is anything in an EMG like a differential amp using JFETs.

I heard that EMG used lower-impedance coils (less turns, tad bigger wire) than your standard passive pickup. Whole deal may be BJT-based.

As a wild guess, I'd figure a CE BJT stage into a 2-pole BJT LPF to mimic the high-Q bump of passive pickups. With lower impedance pickup coil and a humbucking "dummy" coil (with non-magnetic core) in series. Two magnetic core coils to mimic sound of the wide aperture humbuckers. If the coil impedance is above 2-3K, maybe a common source JFET input.

Tweak the LPF cutoff frequency and Q to mimic your favorite passive pickup.

Hunt up Bartolini's patents and old (1970s) sales flyer for some good info on pickup resonance and stuff.

If you just want a preamp in the guitar for a regular passive pickup, use a JFET and BJT follower buffer with a series resistor of 10K on the output. You will prolly find you want to hang a good sized capacitor in parallel with the pickup to "tune" the resonant frequency hump since it expects a guitar cord-sized capacitance, and they tend to be 15 feet long, not 8 inches.

May I ask what you are trying to accomplish?

Best wishes on your project and let us know what you come up with!





:thumb:
 
Yes, emg advertises low-impendance winding. Not so sure about duncan, although duncan advertises symetrical preamps - see above. Carvin uses differential preamp (1/2 of mc33178) for pickups in some of their axes.

As I said, I need as hum-noise-buzz-free output with the sound of existing pickup as possible. I have been using "standard", i.e. non-symetric, non-differential preamps for some 15years now, they are good, but I need as noise-resistant as it gets and I don't want to swap pickups.

The inputs as I drawn on a layout have 330pF hot-hot, and 47pF hot-ground each, the 330pF is a "good enough" compromise to my ears, and often seen in commercial designs as well. Have been experimenting with values in years past :) ...

The inputs may benefit with added R-L circuit, f.e. 12k/ferrite bead for each branch in front of the existing circuit.

Possibly the key would be to lower input impendance and use an opamp (for better cmrr).

I'd like to get a "test" guitar to test the circuitry before I mod my working instrument.


I'll look into bartolini patents one of these days.
 
[quote author="tv"]As in title.

Explanation: I'm looking for as-simple-as-possible DC-coupled jfet differential input stage idea that would work well with guitar pickups, connected in floating configuration.[/quote]
You could have a look at the Neve 33609, the section after the diode bridge. It's like a deluxe-version ('CFP') of the circuit Wavebourn posted.

http://www.ams-neve.com/downloads_all/files/Outboard/33609J.pdf

page 18, TR16,17 (JFET) & TR3,4 (BJT)

Bye,

Peter
 
I don't think this would be a good idea for an onboard circuit. In any case, it would need an aaditional opamp debalancer for the appropriate output.

As I said, added benefit of an all-fet circuitry is that pickup can be DC-coupled and stays @ 0volt DC-potential at all times (no serial caps in signal path).
 
Hi tv,

but I need as noise-resistant as it gets and I don't want to swap pickups.

High impedance pickups are inherently more prone to noise pickup from the usual suspects (florescent lights, TV stage lighting, triac/scr dimmers, etc.).

So if your goal is really less noise pickup, audition a lower impedance model whose acoustic signature matches your favorite pickup, EMGs really sound good. If your goal is more to keep your current pickup, try greater shielding if the coil is exposed. The whole pickup can be wrapped in copper, which will cut the electrostatic pickup.

If you want less noise AND keeping your current pickup, well, I'm afraid you may be out of luck.
 
I already have a working and over a decade-and-a-half field tested solution (my own design, single-ended). I don't want to swap pickups (no way), but I need to improve on what I already have, that's the whole point.
 
[quote author="tv"]I don't think this would be a good idea for an onboard circuit. In any case, it would need an aaditional opamp debalancer for the appropriate output.

As I said, added benefit of an all-fet circuitry is that pickup can be DC-coupled and stays @ 0volt DC-potential at all times (no serial caps in signal path).[/quote]
OK, FWIW, no caps, can be done as well. You didn't say anything about not using a small TX...
Use two of these: http://www.jensen-transformers.com/as/as098.pdf, add an offset-adjust and DC-couple the TX-prim. Check for stability of the offset-adjust.
 
Hehe jensen take on TL/H/6791-24 from an32 pg.8 :)

Such circuit is a contender for a single-ended "ultramojo" onboard preamp.

courtesy of me:


(slight mistake slipped in: R7 goes to ground)
 
I already have a working and over a decade-and-a-half field tested solution (my own design, single-ended). I don't want to swap pickups (no way), but I need to improve on what I already have, that's the whole point.

Improve WHAT to accomplish WHAT?

Do you have RFI filtering on the preamp input?

Shielding?

Lots of us have gotten sucked in to trying to keep up with ad copy. Perhaps that is the case here?
 

Latest posts

Back
Top