Switchable onboard buffer for otherwise passive bass

[antonis]

I still need to do more research before deciding...

Decide about what exactly, Andy..?? 

Are you making a state-of-art research or you just wish to design an input buffer "quieter" than your guitar pickups..??
(for the last case, almost all the above suggestions should be considered quite satisfactory.. if not, a shielded breadborad & a noise analyzer could save you a lot of mind trouble..)

[merlinb]

> bootstrapped BJT Emitter follower

Mind giving a quick rundown of how it works?  Best I could come up with was "increased negative feedback at higher frequencies".  My transistor brain is getting atrophied.

It's an emitter follower using a Darlington pair. The input is biased with a potential divider (R87/R88) and then bootstrapped with a feedback cap C7. Dead simple, but it uses PNPs and is drawn 1969 style so everything is upside-down and inside out. Here's it is after being converted into non-Australian mode:

[Gus]

A bunch of stuff has been posted at this forum over the years.

A link to an over designed for guitar/bass buffer posted more for people to get ideas from

https://www.diystompboxes.com/smfforum/index.php?topic=105522.msg954811#msg954811

I would search for FET input rail to rail opamps that operate at 9VDC

Or use the sony textbook buffer from the above link.

[PRR]

> Just realized C7 value is 2000μF (2mF)..

No, just micro. There were not greek so didn't have the symbol. 2uFd is ample for pushing 75k around.

[Fancy Lime]

I still need to do more research before deciding...

Decide about what exactly, Andy..?? 

Are you making a state-of-art research or you just wish to design an input buffer "quieter" than your guitar pickups..??
(for the last case, almost all the above suggestions should be considered quite satisfactory.. if not, a shielded breadborad & a noise analyzer could save you a lot of mind trouble..)

Decide if the whole idea of a buffer is a good one to begin with (considering the unavoidable trade offs) or if I want to include a passive treble boost instead. The latter would mean that I need to tune the cable capacity such that the RLC lowpass filter of the pickup resistance and inductance with the cable capacity has its resonant  peak somewhere useful. That also means shelling out biggly for a decent LCR meter. Which shifts the decision to: What LCR meter?

Andy

[amptramp]

I use an AideTek DM4070 which is a direct reading LCR meter with ranges from 200 pF to 2000 µF, 200 µH to 20 H and 20 ohm to 20 megohms.  I also have a number of LCR bridges but they are not direct reading - I collect test equipment and certain pieces are better for different purposes.

[R.G.]

I still need to do more research before deciding...

Decide about what exactly, Andy..?? 

Decide if the whole idea of a buffer is a good one to begin with (considering the unavoidable trade offs) or if I want to include a passive treble boost instead. The latter would mean that I need to tune the cable capacity such that the RLC lowpass filter of the pickup resistance and inductance with the cable capacity has its resonant  peak somewhere useful. That also means shelling out biggly for a decent LCR meter. Which shifts the decision to: What LCR meter?

Unless you just enjoy the dithering, I think you're seriously overthinking (or at least over-typing) the issue. Your forum reading and typing time on this issue alone would probably have been enough to breadboard and try a couple of approaches. 
Not that there's anything wrong with overthinking, of course. 

[Digital Larry]

Probably due to "digital" operation of your transistor brain.. 
(just kidding, of course..)

No worries, thanks for trying!  I now recall my senior year at university when I began to study DSP.  I thought, "this seems much easier, you just tell the thing what you want it to do!"  Realizing that one still needs a well behaved high impedance preamp before one shreds the signal into screaming helpless ones and zeros.

[PRR]

https://xkcd.com/1952/

[antonis]

Realizing that one still needs a well behaved high impedance preamp before one shreds the signal into screaming helpless ones and zeros.

(couldn't figure out better description..)

[antonis]

> Just realized C7 value is 2000μF (2mF)..
No, just micro. There were not greek so didn't have the symbol. 2uFd is ample for pushing 75k around.

<off-topic on>

In such a case, apparent R86 value should be raised hardly X50 for open low E string and even lower (x10) for 20Hz signal..(supposing it's incorporated into an audio console..)
Can't see any reason, other than cost, for using such a low value capacitor..

<off-topic off>

[sergiomr706]

https://www.talkbass.com/threads/onboard-tillman-inspired-jfet-preamp.1389669/ if you have previously done the variable cap and parallel series sw, i would try like the one at this post, maybe an active/inactive sw with the tillman is all you need

[Fancy Lime]

I still need to do more research before deciding...

Decide about what exactly, Andy..?? 

Decide if the whole idea of a buffer is a good one to begin with (considering the unavoidable trade offs) or if I want to include a passive treble boost instead. The latter would mean that I need to tune the cable capacity such that the RLC lowpass filter of the pickup resistance and inductance with the cable capacity has its resonant  peak somewhere useful. That also means shelling out biggly for a decent LCR meter. Which shifts the decision to: What LCR meter?

Unless you just enjoy the dithering, I think you're seriously overthinking (or at least over-typing) the issue. Your forum reading and typing time on this issue alone would probably have been enough to breadboard and try a couple of approaches. 
Not that there's anything wrong with overthinking, of course. 

Me, overthinking something? How dare y... uhmm ya that checks out, actually. As a kid I used to start lists of potential Christmas wishes in October to compare things I *might* want to want for Christmas. With tables comparing pros and cons. Other kids just wanted everything, I wanted to know what I really wanted much rather than actually wanting to have it. All absolutely moot because I knew full well that I never got what I wanted anyway. That realization did not stop me then, I'm afraid it still doesn't today. This forum seems to have become my year-round Santa but with a much better delivery rate. You are of course right, I should just build the obvious candidates and see what sticks. And as soon as I have my DIY bench back (having visitors at the moment), I will. But until the weekend, theorizing will have to do.

https://xkcd.com/1952/

I want royalties for being in that comic. Like I said, Christmas lists.

Andy

[Fancy Lime]

So I finally just did what I intended to do before starting to overthink it. Simple JFET source follower using a 2SK117GR with gate bias at V/2 and the source resistor tuned for the optimum balance between noise performance and current consumption as per the datasheet. The lowest quiescent source current that we want to use for that is about 750µA. With 9.2V from the battery, I measured 4.7V at the source of Q1, giving me 840µA across a 5k6 resistor. Close enough to not shed tears about 90µA wasted. Since the bias network has 2M across the rails, it does not contribute much more to the current consumption, so 840µA is also what I measured as the total current cunsumption. Should give me a few hundred hours of operation from an average 9V block; good enough for me. I used a "noiseless biasing" scheme because my ocd told me to and you guys did not stop me in time. Muhuhahaha. May not make a world of difference, though. The 10k series resistor on the input keeps the input resistance in the optimum range for best noise performance even when this thing is fed with a low impedance source. So this little thing does all I want it to as good or better than I expected it to. At only eleven parts, that was way easier than I (over)thought. And it does not audibly clip even at the loudest peaks from two Status Hyperactive pickups (which are passive despite the name but pretty hot) in series, so it should work fine with most pickup setups. Anyway, here's the layout (I did not bother drawing a schematic but that should be obvious to most readers and for the others it will be a good exercise to trace the schematic from the layout):



The layout is not verified at this point in time. [[[EDIT: I just found an error: the input resistor R1 is shorted by a trace. This trace should of course not be there. I'll fix it in the schematic as soon as I can.]]] I still have to build it and see if there are errors (only on the breadboard so far). I am not unhappy with the layout, there is only little wasted room for improving the compactness unless you use smaller footprint caps. One row could probably be scraped of by a more talented layout artist. An SMD version could probably be made to fit into most 1/4" plugs, so you could phantom power it through a stereo cable and thus make any passive bass or guitar active. I have seen this concept somewhere but I don't remember where. If someone knows, please add the original reference, I don't want credit for ideas that were not mine. If you want to power this from a wall wart instead of a battery, you can add a 100µF-470µF cap across the rails on the right edge of the schematic by making the board one row wider. Or make it two rows wider and use a 100µF-470µF electrolytic in parallel with a 100nF (or similar) film cap for really troublesome power supplies. In that case I would also change the diode to something with a bit more internal resistance, like a 1N5819 or even a 1N400x.

Hope that helps people who tend to run circles inside their own brains, like I do.
Cheers,
Andy

[Ben N]

An SMD version could probably be made to fit into most 1/4" plugs, so you could phantom power it through a stereo cable and thus make any passive bass or guitar active. I have seen this concept somewhere but I don't remember where. If someone knows, please add the original reference, I don't want credit for ideas that were not mine.

http://www.till.com/articles/PreampCable/

[PRR]

> The 10k series resistor on the input keeps the input resistance in the optimum range for best noise performance even when this thing is fed with a low impedance source.

Never add series resistance "to suit OSI". The resistor is not The Source.

Yes, in stage work a 10k is good protection and probably a non-issue for hiss.

[Fancy Lime]

An SMD version could probably be made to fit into most 1/4" plugs, so you could phantom power it through a stereo cable and thus make any passive bass or guitar active. I have seen this concept somewhere but I don't remember where. If someone knows, please add the original reference, I don't want credit for ideas that were not mine.

http://www.till.com/articles/PreampCable/

Yes, that's the one. I remembered incorrectly that it was a buffer, though, instead of a booster. The concept can be adapted for a buffer, I think.

> The 10k series resistor on the input keeps the input resistance in the optimum range for best noise performance even when this thing is fed with a low impedance source.

Never add series resistance "to suit OSI". The resistor is not The Source.

Yes, in stage work a 10k is good protection and probably a non-issue for hiss.

Paul, you are one of those people here whose word I'm willing to take without an explanation if I have to. Nevertheless, I would greatly appreciate if you could expand a little on that topic, since it still baffles me. Why is a series resistor bad in terms of noise/hiss? There is no DC voltage across it, so it does not give of thermal noise, right? Or does it? Any noise contribution that is only present when there is an AC signal would certainly matter if we were to build an input for a super accurate scientific measuring probe but for audio, especially a guitar signal, I don't think it matters because human ears would not be able to tell the difference (as long as we're not going completely outside of reasonable design territory).
And what is the "source resistance" anyway? The signal comes from a guitar pickup, so the reactance of the inductor swamps any ohmic resistance for audio band AC signals. What does that mean in relation to the (usually) terse info given in datasheets? Some tell me the optimum source resistance at 1kHz (https://www.onsemi.com/pub/Collateral/2N5457-D.PDF), but most datasheets do not specify the influence of source resistance at all, let alone the relationship that has with frequency. And when this sort of information is given it is mostly "typical" curves, when "best and worst" curves would be so much more useful. Very frustrating not to be give this information which I assume the manufacturers must at least partially have.
Also: what do I protect the gate from with the series resistor? Gate-Source breakdown voltage being exceeded by electrostatic discharge? If so, how does the resistor help?

Thanks and Cheers,
Andy

[PRR]

> how does the resistor help?

Limits current flow.

[merlinb]

There is no DC voltage across it, so it does not give of thermal noise, right? Or does it?

Yes it does, it's called Johnson or thermal noise.

And what is the "source resistance" anyway?

The intrinsic internal resistance of the signal source (e.g. the pickup coil).

the reactance of the inductor swamps any ohmic resistance for audio band AC signals.

Reactance does not generate noise, resistance does. (However, impedance does provide a way for noise current coming out of the amplifying device to produce a noise voltage which is then amplified.)

And when this sort of information is given it is mostly "typical" curves, when "best and worst" curves would be so much more useful. Very frustrating not to be give this information which I assume the manufacturers must at least partially have.

That information applies to the impedance-matched world of RF (power signals), not AF (voltage signals). The optimum source impedance for audio is "the internal impedance you're stuck with after removing all unnecessary source impedance!"

[Fancy Lime]

Thanks Merlin, that clears things up, to a degree. And as it is the habit of any good answer, it generates more questions

Am I right in assuming that the "impedance noise" is coupled to the current that flows into the gate, then? In other words, probably negligible for a JFET but potentially troublesome inf form of a base current when using a BJT instead?

That information applies to the impedance-matched world of RF (power signals), not AF (voltage signals). The optimum source impedance for audio is "the internal impedance you're stuck with after removing all unnecessary source impedance

"Don't worry  bout it", gotcha.

> how does the resistor help?

Limits current flow.

What current, though? Should the current into a JFET gate not be very limited already by the fact that is a JFET gate?


Thanks guys, I would be even more lost without y'all,
Andy