Variable Resonance Pickups - Fun with Buffers

[charrich56]

Charlie,

No worries, you're welcome, I'm curious to see how the OPA188 works out as well. No big deal if it doesn't since there are so many choices for a single in a SO-8 package.

The slew rate might be a concern but it should work out just fine for guitar where we are really staying below 1V peak to peak a lot of the time. We're a long way from the 741 design and all the nasties it did on overvoltage, latchup, and phase reversal. For an 18 volt bass design you are running a lot hotter signal I think so I would have more concerns there.

The low power low noise , 9VDC+ choices Joel gave me were:

http://www.ti.com/lit/ds/symlink/opa188.pdf 8.8 nV/√Hz 510 μA (max)
http://www.ti.com/lit/ds/symlink/opa180.pdf 10 nV/√Hz 525 μA (max)
http://www.ti.com/lit/ds/symlink/opa171-q1.pdf 14 nV/√Hz 475 µA
http://www.ti.com/lit/ds/symlink/lmv641.pdf 14 nV/√Hz 138 µA

[Passinwind]

The low power low noise , 9VDC+ choices Joel gave me were:

http://www.ti.com/lit/ds/symlink/opa188.pdf 8.8 nV/√Hz 510 μA (max)
http://www.ti.com/lit/ds/symlink/opa180.pdf 10 nV/√Hz 525 μA (max)
http://www.ti.com/lit/ds/symlink/opa171-q1.pdf 14 nV/√Hz 475 µA
http://www.ti.com/lit/ds/symlink/lmv641.pdf 14 nV/√Hz 138 µA

Effectively spec'ing noise often requires going a lot deeper than voltage noise density, especially for relatively high impedance inputs. I assume you're familiar with this doc: http://www.analog.com/media/en/train...als/MT-047.pdf ?

[charrich56]

Thanks Charlie, that's a very good resource paper and I'm glad you tipped me off to it. I think I had run across it a while back but it's still one of the best concise articles on op-amp noise. I'll make sure I look at the current noise data as well on making a final pick. But for now the voltage noise density is a good quick selection yardstick. Discrete JFET may still possibly win on this one, but it's going to be fun to find out

[salvarsan]

The MT-047 design note looked familiar, reminded me of Walt Jung's old Op Amp Cookbook. It references two design books at the end, both costly as hardcopy, but available as free PDFs if you know where to look. Note: "costly" means ~$90 new. AYFKM?! Seems Elsevier bought the hardcopy distribution rights, applied academic/institutional price models, and they went from $30 to $90 overnight.

download FREE at:

The Linear Circuit Design Handbook PDF is in the Analog Devices educational material section.

Walt Jung's last iteration of Op Amp Applications Handbook, 2005 is there, too.

(edit) General noise considerations are in chapter 1.
An instructive worked example of total noise starts in Chapter 6 page 150.

Not so free:

E-book Analog Circuit Design is still reasonably priced at $20/download on the SysCompDesign site.

-drh

[Passinwind]

Just a quick note that I am following through on releasing an open source design for my variable resonance swept frequency bass preamp: https://www.talkbass.com/threads/the...reamp.1259692/

That thread has generated well upwards of 500 views already, so I think there will be some real interest in the other Charlie's approach as well.

[charrich56]

Charlie,

Thanks for releasing open source, and thanks for all your time put into the design.

I'm going to forward the link over to a couple of forums and FB groups and I'm sure there's going to be significant interest.

I have PCB's for all three variants of the buffer in hand from OSHPark and will be assembling and testing pending a few parts coming in.

BTW Charlie, thanks to your suggestion, SMT hand soldering is not quite as bad as I had imagined. I'm getting better at it. I really don't have do do anything smaller than an 0805 thank goodness.

-Charlie

[Passinwind]

Charlie,

Thanks for releasing open source, and thanks for all your time put into the design.

I'm going to forward the link over to a couple of forums and FB groups and I'm sure there's going to be significant interest.

I have PCB's for all three variants of the buffer in hand from OSHPark and will be assembling and testing pending a few parts coming in.

BTW Charlie, thanks to your suggestion, SMT hand soldering is not quite as bad as I had imagined. I'm getting better at it. I really don't have do do anything smaller than an 0805 thank goodness.

603s are about the same as 805s as far as degree of difficulty for hand soldering. Handling them is a little more nerve frazzling though, especially without a good magnifier rig.

My preamp design has garnered over 2500 views on Talkbass so far, so I guess there is substantial interest in this general concept. It's been getting quite a bit of pre-NAMM buzz too, and I'm really looking forward to seeing how that goes.

[Passinwind]

Charlie,

Thanks for releasing open source, and thanks for all your time put into the design.

I'm going to forward the link over to a couple of forums and FB groups and I'm sure there's going to be significant interest.

Just wanted to check back in and mention that my open source project is finally now live: https://www.talkbass.com/wiki/pw3b-l...nboard-preamp/

Feedback gathered at NAMM and elsewhere led me to try a few different filter alignments to make the effect of turning the LPF knobs more obvious to the end user, and also to allow me to spec tighter tolerances for the LPF caps. It's a tweaky circuit by nature and using 2.5% caps makes fine tuning a much more predictable exercise than going with 5 or 10 percenters. I also adjusted some of the component footprints a bit to make board rework more practical, a facility I have definitely needed to take advantage of a few times while dialing in a legit v1.0 spec.

[David Schwab]

Sounds great! I hadn't heard of the CAL Redeemer buffer so I had to look it up...

Creation Audio Labs

Kinda pricey especially when you figure in the $16.87 shipping charge for UPS ground...

A few years back I had one of these to play with. What I noticed was it didn't sound like a buffer. I'm used to buffers extending the high frequencies, and tightening the low end. The Redeemer sounds like the guitar does passively, but I suppose it retains that sound no matter what you have it plugged into.

Here's a bass through the Redeemer, and passive. You will hear a beep when I switch. Can you tell the difference?

http://www.sgd-lutherie.com/media/passive-redeemer.mp3

So I recorded some samples with the Redeemer, an EMG BTC, and a JFET buffer, based on the Alembic Stratoblaster. I labeled them A, B, and C. I posted them online (at TalkBass) and most people preferred the JFET buffer! That included myself.

One thing about the Redeemer is I believe it works even if you disconnect the battery. The circuit is kind of elaborate, and it's sort of in parallel with the signal from the guitar. This was based on looking at the patent application.

Here's the test links. One is the Redeemer, one is the JFET, and the other is an EMG BTC set flat.

http://www.sgd-lutherie.com/media/redemmer_test_A.mp3

http://www.sgd-lutherie.com/media/redemmer_test_B.mp3

http://www.sgd-lutherie.com/media/redemmer_test_C.mp3

Here's the bass passive:

http://www.sgd-lutherie.com/media/re...st_passive.mp3

[charrich56]

David,

Thanks for your info on the Redeemer. In OEM /dealer quantity (5) just the module can be had significantly cheaper but probably not under $30.00 or so.

You are right, it does have the ability to pass signal when the battery is drained or disconnected. The big deal on these is current draw, so expect about 100-200 hours for a battery. For professional use I would expect that you would want to change the battery out proactively, before major gigs or sessions, weekly, etc. So not a major deal breaker for a single buffer system, but I'm forecasting that I will want two variable rez tone controls on a 2-pickup instrument so need a lower current draw solution.

My current favorite buffer is my own design 2-JFET follower and I have built a couple of these and put them in demo instruments. The input Z is 10 Megohms (different from shown schematic) so this would also work for a piezo pickup, albeit just as a buffer.



I'm using the LSK389A dual low noise JFET on these instead of 2 J201's, and am getting less than 700 microamps current draw for a buffer that has a greater than 6V peak to peak output swing. Given the low noise floor, this is way better on undistorted (or darn near undistorted) dynamic range than a single JFET follower running on 9V for the same current draw. The Redeemer is probably "better" signal-wise on these measurements but the current draw is significantly more.

Because of the dynamic range, IMHO my buffer circuit could work very well for bass pre-active EQ as is. I'd really look at Passinwind's new open source bass preamp design for the ultimate in flexibility, though.

Audibly my buffer sounds great to me, but I haven't put one of my boards on a distortion analyzer, or tested one with RightMark yet. (but soon.)

I have a OSHPark SMT design for this buffer and it's about 1" by 1.1". The connector is a JST XH series 8 position. Here's a pic:



Here's a pic of the board (sorry, not a very good one) with the 8 pin DIP header beside it. The header allows soldering in of resistors and caps , to make a plug-in module to tune the resonant frequency range and resistive loading, to allow tweaking of the system to the characteristics of the pickups being used and desired tonality:



Anecdotal evidence from lots of players and pedal designers suggests that discrete JFET preamps/buffers (whether Tillman, Stratoblaster, etc.) add a very desirable tonal quality to the mag pickup signal. I'm still looking at newer op amps right now, but JFET buffers seem to be thought of as musical and quiet, in line with what you were saying about your testing.


The only drawback on my buffer is that the dual JFET is $6.00 each and I am using completely top quality parts (high quality Cornell-Dubilier acrylic film caps in the audio signal path, for example), so that I can't get the price of the buffer as low as I would like. Most likely it will be in the same price range as the Redeemer. I hope I can do something about that shipping though; it seems excessive on the Redeemer.


-Charlie

[MisterBzr]

Charlie,

I would like to experiment with your buffer.
I have a couple of questions.

1. What made you switch to the LSK389A instead of the J201 JFETs?
I've plenty of J201s so I would love to use these.
2. I've been using a Till preamp with the 9v powersupply stompbox, so I can feed the preamp through my jack cable.
If I add a cap. on node 3 on your schematic and connect that in series with the jack cable and also make a connection at that point to C3, would that work?
3. Do I have to match the J201s?

Thanks in advance,

Hans

[charrich56]

Charlie,

I would like to experiment with your buffer.
I have a couple of questions.

1. What made you switch to the LSK389A instead of the J201 JFETs?
I've plenty of J201s so I would love to use these.

J201's in general should be OK.

I'm using the LSK389A's because of ultra low noise, good consistency on Vgs and Idss parameters, and also the cost of the dual JFET ($6.00 USD) is less than two good quality JFETS ($4.00 each.) I don't want to have to hand tweak every board, or hand-select JFETS when I get to selling the commercial version. But I have no problem with adjusting and tweaking a circuit for DIY; in fact, that's part of the fun.

If you have some good Fairchild or Siliconix J201's, you can use those. Some of the recent overseas parts sellers on Ebay from what I hear, are not going to give you a good quality part. Linear Systems also sells good quality J201's, and also LSK170A's.

2. I've been using a Till preamp with the 9v powersupply stompbox, so I can feed the preamp through my jack cable.
If I add a cap. on node 3 on your schematic and connect that in series with the jack cable and also make a connection at that point to C3, would that work?

I think you're asking if the buffer can be phantom powered, as in the Tillman preamp cable. I don't think that would work, unfortunately. If you don't want to use an onboard battery, you can use a stereo 1/4" jack, 1/4" cable, and use one of the conductors in the cable for power from the power supply box.

3. Do I have to match the J201s?

The two JFETS perform different roles where the bottom JFET is a current source for the upper JFET source follower. Technically they don't have to be matched but they end up operating in almost a push-pull mode. Practically speaking, you would get the best dynamic range if the JFETS were relatively matched for Vgs.

Even with unmatched JFETS (or loosely matched within 10-15% on Vgs) , as long as you get the current through the pair OK, the circuit will have more than enough undistorted voltage swing to work fine for guitar.

When you build the circuit, try to get around 500 microamps minimum through the JFETs as a starting point. R8 will be the best place to experiment. If you can't get the current low enough when you go to 0 ohms for R8. the J201's probably won't work in this circuit and you have to hand select some different ones. If you want to go up to 1 mA or so, that's fine, but stay at or around 600 to 700 uA current draw (for the whole circuit including the voltage divider reference) for best onboard battery life.

Here's the Circuitlab URL for the circuit where I think you can do limited simulations without having to subscribe to the service.

https://www.circuitlab.com/circuit/v...fer-j201-v1_1/

I do have a SMT PCB layout, done in DipTrace, for the discrete JFET version of the buffer.

-Charlie

[MisterBzr]

Thanks for your reply, that's sound advice!
It will take some before I have all the stuff.

Hans

[David Schwab]

David,

Thanks for your info on the Redeemer. In OEM /dealer quantity (5) just the module can be had significantly cheaper but probably not under $30.00 or so.

You can make the JFET buffer I used in the sound clips for under $5 with quality parts. It's a dirt simple. I used the Stratoblaster circuit with a few part changes. My complaint about the Redeemer was I felt the input impedance was too low. That's why it sounds like the bass does passive. Then what's the point in having a buffer? It's not really isolating the pickups from the other loads, as far as the tone is involved since it's dulling the top end.

Your preamp looks interesting. Personally I'd increase the input impedance. When I get a chance I'm going to build one.

[charrich56]

You can make the JFET buffer I used in the sound clips for under $5 with quality parts. It's a dirt simple. I used the Stratoblaster circuit with a few part changes. My complaint about the Redeemer was I felt the input impedance was too low. That's why it sounds like the bass does passive. Then what's the point in having a buffer? It's not really isolating the pickups from the other loads, as far as the tone is involved since it's dulling the top end.

Your preamp looks interesting. Personally I'd increase the input impedance. When I get a chance I'm going to build one.

David,

Would you want over 10 Megohm input impedance? I was thinking that 10 Megs was enough for even a piezo pickup. But you could go higher if you wanted.

The surprise is that the Redeemer info says it is 20 Megohms input Z. I know that CAL have changed the Redeemer design at least once so don't know if you might have tested an earlier revision, or whatever else would explain the dulling of highs you heard on the test.

I've built several Tillman and Stratoblaster style JFET preamps/buffers over the years, usually on perf board because as you say, they are dirt simple. They work well but usually require tweaking of the source and/or drain resistor values. Usually that wasn't a problem for me because I like doing the DIY thing.

Please let me know via forum or PM if I can help you in any way if you're going to build a version. If you're going to use the circuit just as a buffer, you don't need anything that is connected to the TC1 and TC2 terminals on the board, and you don't need the extra output capacitor C8.

At some point this year I will order a small production run of the SMT version and make them available. But if there's sufficient interest I could do a through-hole PCB as well. I am somewhat tempted because it would be possible to have more parts choice on things like the film capacitors, etc., and it might be a cheaper board.

-Charlie