News:

Happy New Year!


mr coffee

IDEAS: "a differential discrete jfet input stage"
« Reply #40 on: February 16, 2008, 04:58:42 PM »
Sorry it took a bit to get back. Busy-busy.

Here's the Alembic Tone Control. I traced it out 30+ years ago, but it took me a bit to pretty up my scratchings to be readable. It's just a Sallen and Key LPF with variable Q, but it sounds good with fairly flat pickups - you can tune it whatever pickup sound you want, more or less.

The rest of the Alembic circuit is a single-ended preamp for each pickup coil which has the signal from the humbar fed into the feedback loop to cancel out the hum. PM me if you want to see all that stuff.




Enjoy :green:


tv

IDEAS: "a differential discrete jfet input stage"
« Reply #41 on: February 17, 2008, 07:43:26 AM »
Interesting. Perhaps do you have more info or a diagram on how eactly did Alembic do their buffers and what are you reffering to as a "humbar"?

Do you think that the "humbar" technique is sufficiently general so that it would be doable with "shelf" pickups (i.e. putting something around bartolinis) or must everything be caculated/measured with painful precision?
If you sprinkle when you tinkle, please be neat and wipe the seat.

clintrubber

IDEAS: "a differential discrete jfet input stage"
« Reply #42 on: February 17, 2008, 01:34:49 PM »
Quote from: "tv"
Interesting. Perhaps do you have more info or a diagram on how eactly did Alembic do their buffers and what are you reffering to as a "humbar"?

Dutch magazine Music Maker did an article about this circuit long ago, complete with the buffer stuff etc. I could scan if still needed.
Much later on (after having perfboarded it back then) I noticed that the article-schematic had an error, will indicate that with the scan.

tv

IDEAS: "a differential discrete jfet input stage"
« Reply #43 on: February 17, 2008, 02:02:35 PM »
@Clintrubber: that would be very cool.

I'm especially interested in this "humbar" device MrCoffe mentioned. I suspect the essence is pictured in the diagram MrCoffe posted before the lowpass filter.


@Nikolay: did you simulate the circuit yet? Could you trace it from the layout comfortably?






I propose that this thread serves as a general thread for those who are fiddling with "buffered passive instruments" from now on.
If you sprinkle when you tinkle, please be neat and wipe the seat.

mr coffee

IDEAS: "a differential discrete jfet input stage"
« Reply #44 on: February 18, 2008, 10:12:05 AM »
@tv,
The humbar is just a "dummy" coil without a magnetised core, placed in between the single coil pickups, recessed flat with the top of the instrument (1/2" to 3/4" away from the strings).

It picks up stray electromagnetic fields (aka hum and crud) and very little of the string vibration. The signal is used to buck the hum from the pickups. Alembic calls it a humbar. I think there are some Gibson guitars out now that are using a less-sophisticated "dummy coil" to buck the hum from "soapbar" single-coils.

The Alembic is a pretty straight-forward design. Humbar signal is amplified, run through a trimpot attenuator for each pickup, and fed into the inverting input of the preamp for each pickup. Nothing critical. You can set the trimpot by ear sitting a few feet away from a hum source with the gain turned up high and the strings off or damped with a folded up cloth. Nulling out is always a compromise.

Single-coil Bartolini pickups?

@Peter,
If you have a half-pretty schematic you can scan and post I'd appreciate not having to redraw my faded hand-drawn schematics.

Here's a hand-drawn:



tv

IDEAS: "a differential discrete jfet input stage"
« Reply #45 on: February 18, 2008, 03:27:19 PM »
Eh dummycoil :oops: .... I was naively hoping the "humbar" to be the nulling technique as you described on previous page (with unity signal being fed into the "casing") ...

In that one particular bass I have older bart stacked j's. They are pretty sensitive for CRT and similar noise. But that bass' electronic is as well quite complicated and doesn't give much room for fiddling.

I was hoping to be able to f.e. wire a "special buffer" (maybe like those in alembic circuitry) and send a wire to the shielding in pu cavities - but that would require unproportional amount of work compared to my other axes.

Well, it looks like I'll have to make a stronger rf-filter. Do you think that chokes like this one would work sufficiently?
http://www.ferroxcube.com/appl/info/WBSM.pdf
I figure I could solder these directly on cables, use smt capacitors as well, spray the ditty with Urethane and finally put a thermo-shrink over it.


I was fiddling with one of my cheap basses and while I've got a surprisingly increased amount of resistance to CRT noise with simple changing of RC values @front of my buffer, I broke some innards ...

I need to change pots anway, now I'll have to do it much sooner.


Ah well ...


Thanks for the schematics btw.!
If you sprinkle when you tinkle, please be neat and wipe the seat.

clintrubber

IDEAS: "a differential discrete jfet input stage"
« Reply #46 on: February 18, 2008, 04:29:11 PM »
Quote from: "clintrubber"
Dutch magazine Music Maker did an article about this circuit long ago, complete with the buffer stuff etc. I could scan if still needed.
Much later on (after having perfboarded it back then) I noticed that the article-schematic had an error, will indicate that with the scan.


Here we are:
http://www.twin-x.com/groupdiy/albums/userpics/RealBassesArePassiveAndHaveFourStrings.zip
(763kb)

NOTE:
#1 the drawings ar probably redrawn by the magazine, no idea about accuracy w.r.t. originals
#2 article-error: caps 6n8 // 39p should attach one resistor-node to the east for best results
#3 for those able to read Dutch, feel free to ignore the suggested opamp-type and the tant-caps

Enjoy,

  Peter

tv

IDEAS: "a differential discrete jfet input stage"
« Reply #47 on: February 18, 2008, 06:02:14 PM »
Here's the 'touched preamp section (w/o the bloody "humbar" section) - to me it looks "right":





Clearly, this is the "father" of the tubescreamer. It says "4558" and it has the filter. Imagine Mr.Clarke pushin it hard.

This spells "PURE MOJO". The numerologists would say simply "4558".
If you sprinkle when you tinkle, please be neat and wipe the seat.

clintrubber

IDEAS: "a differential discrete jfet input stage"
« Reply #48 on: February 18, 2008, 06:13:38 PM »
Quote from: "tv"
Clearly, this is the "father" of the tubescreamer. It says "4558" and it has the filter.

:shock:
My vacuum cleaner has a filter as well, but that doesn't make it a TS I'd say.

In other words: the Alembic circuit hasn't any TS-like clipping stuff, but feel free of course to add...

Also note that the text says that in the original the FET & IC-type aren't indicated; the mentioned types are just suggestions from the author of the article.

Bye,

  Peter

tv

IDEAS: "a differential discrete jfet input stage"
« Reply #49 on: February 18, 2008, 06:19:10 PM »
Well - that was meant as a mojo-joke.
If you sprinkle when you tinkle, please be neat and wipe the seat.


clintrubber

IDEAS: "a differential discrete jfet input stage"
« Reply #50 on: February 18, 2008, 06:25:37 PM »
Quote from: "tv"
Well - that was meant as a mojo-joke.

Hmmm, you're sure ? But OK then...  :evil:  
But imho it's quite a sick idea to make fun of active basses, they're sad enough already  :wink:  :green:

Why not add envelope follower control to that LPF ? On-board autowah  :twisted:

mr coffee

IDEAS: "a differential discrete jfet input stage"
« Reply #51 on: February 18, 2008, 06:57:26 PM »
tv,
Quote
I was naively hoping the "humbar" to be the nulling technique as you described on previous page (with unity signal being fed into the "casing") ...


It's not really a "nulling" technique. By driving the shield from a low impedance node, you get the best junk reduction you can from a high impedance series-connected humbucking pickup. The reason the drive is in phase with the input is so it doesn't load the humbucker coils with more capacitance. The signal bounces up and down with the desired input signal, so the effective capacitance is roughly zero. Probably one of the luminaries around here can explain it better (probably in terms of Miller capacitance or some such). Since the pickup shield is attached to a low impedance node, it works a tad better at keeping the EMI\RFI out and shunting it away from your preamp.

Quote
In that one particular bass I have older bart stacked j's. They are pretty sensitive for CRT and similar noise.


I doubt you can improve on that very much. CRT noise is pretty intense if you are close to one. Unless you replace the barolinis. Stacked pickups reject hum OK but they can't do much to reject that kind of noise source. Maybe Bartolini makes a side by side "hotrail"-type design.

I wouldn't order a special smt filter do-dad. Just run some magnet wire through a ferrite bead a few times, maybe do the differential trick. Put it in series with the input and hang some capacitors on both sides. It will work as well as and probably better.

It's just not going to get much better with CRT noise and high impedance.

@ Peter,

Thanks for posting the article.

The op amps in the real deal Alembic I worked on were those old-timey bipolar op amps with the adjustable bias terminal - I forgot the number. TI thought it was a great idea for a while. Get this, they were in flat-pack packages - remember those? (Sorry, tv, no 4558 mojo.) LOL. The whole instrument was pretty darn quiet for those days by virtue of the JFET inputs.

The magazine schematic looks about right for the tone control except the Alembic took the feedback from the op amp output directly to the 6800pf cap, not through the output capacitor. Plus of course the previous mentioned snafu tv corrected in his follow-up diagram.

tv

IDEAS: "a differential discrete jfet input stage"
« Reply #52 on: February 18, 2008, 07:21:48 PM »
Quote
I wouldn't order a special smt filter do-dad.
The reason for smt bead should be so that I could simply solder two 0603-size caps on it and the whole thing would be both very small and sturdy, ready to be dipped in resin so it will all be practically seamless from mechanical p-o-v.

In my cheapos I have had a "nest" of point-to-point things, but later moved all to new main pcb so it was all clean and fast to repair (always had a spare pre for each axe).

Can't do that in the bass with barts; I'll have to cut cables and insert the filters there, so it will all have to be very compact, neat and sturdy.
If you sprinkle when you tinkle, please be neat and wipe the seat.

Nikolay

IDEAS: "a differential discrete jfet input stage"
« Reply #53 on: February 20, 2008, 09:54:52 AM »
hi there again ;)
I found a picture inside one alembic:


more pictures here:
http://www.thelowend.net/gallery/viewtopic.php?t=1700

Probably the small black transistors are these fets (only 3, not 4), but I can't find any operational amplifier.
Also there are stereo pot, mono pot and small dpdt switch - looks like the schematics from mr. clintrubber ;)

Cool maybe we guess the alembic mystery :)
http://n-audio.net/
https://www.facebook.com/naudiocustom
Handbuilt boutique stompboxes and custom switching systems

Nikolay

IDEAS: "a differential discrete jfet input stage"
« Reply #54 on: February 21, 2008, 03:49:14 AM »
Hi again.
I see the schematics and PCB pictures that I post
I see that this small black transistors are maybe 2 fets in one package
Also there are SMD opamp solder to bottom side

Here a the same pictures but with zoom ;)
http://n-audio.net/
https://www.facebook.com/naudiocustom
Handbuilt boutique stompboxes and custom switching systems

mr coffee

IDEAS: "a differential discrete jfet input stage"
« Reply #55 on: February 21, 2008, 10:23:12 AM »
@ Nikolay,

Yes, the JFETs are duals (two matched jfets in a single package).

Yes, you found the op amps- they do "surface mount", but they are a predecessor of modern smt packaging from the seventies called flat-pack packages. The leads come out of the sides of the package and are about half an inch long! You solder them down to long pads that are just like regular pcb traces. High tech for it's day.

@tv,

Quote
The reason for smt bead should be so that I could simply solder two 0603-size caps on it and the whole thing would be both very small and sturdy, ready to be dipped in resin so it will all be practically seamless from mechanical p-o-v.


It should be pretty simple to do a tiny pcb - you could just draw it right on pcb laminate with a sharpie marker.

I'd suggest trying the differential-wound inductor with multiple turns for each side, followed by a plain pi-filter in series with the hot side. Mount it as close to the preamp input as humanly possible. If you want to pot it, don't encapsulate it in epoxy. Use RTV instead. Ferrite beads just don't seem to like epoxy. Or put a thin layer of RTV on it and then pot it in epoxy.

Are the Bartolinis at least two leads and a shield? Is the shield connected to the coil in the pickup? If the shield isn't connected to the coil, tap out of the preamp circuit and drive the shield. If the preamp is one of those Bartolini molded preamps with Active Bass-Treble tone controls, tap off the input side of the Treble pot (the lug) with a high impedance buffer. Check the phase, and use a simple attenuator to drive the shield.

Maybe a dumb question, but are the strings grounded?

Nikolay

IDEAS: "a differential discrete jfet input stage"
« Reply #56 on: February 21, 2008, 12:02:47 PM »
Thanks mr coffee!  :thumb:
I will try alembic schem when I have more time.

About the bartolini, I have a chance to install a few electronics and different types of bartolini pickups. They use only one shilelded wire, and classic schematics - shield to GND and hot to volume pot.

Also this is a bartolini active schematics again found over the net ;)
http://www.picvalley.net/u/4/950_838.JPG

And also you can see wiring instructions from bartolini sites.

I never see a differential wired pickup using externel electronics at this moment.
http://n-audio.net/
https://www.facebook.com/naudiocustom
Handbuilt boutique stompboxes and custom switching systems