SBT pickup: Dummy experiment

[Cuki79]

On this video (without sound) I am tapping everywhere on the guitar top, sides, bridge and I also hit the strings muted...

I always see a huge boost at 3 KHz... (Right part of the spectrum)

...This thing must be investigated.



Is it the pickup or the position?

Anyway, I pulled out the pickup and I can insure the double tape bond is very strong.

One sensor did not make it...

[Cuki79]

Did some more experiments... I taped 3 sensors on the spruce plate: a naked disc, a K&K style pickup, and the DIY triangle. I did two experiments, changing the position of the K&K and DIY pickup.

The spruce plate is place on the left and recorded with a Focusrite audio interface. The excitation was done by a speaker on the right feeded with white noise.

The conclusions is the 3 sensors measure the same thing.

Swapping the pickup position (remove tape, retape and clamp overnight) had less impact than actually flip the spruce plate up-side down.

If the 3 sensors are equally exposed the result is similar or identical. If one sensor is placed at the bottom (flipping up-side down the spruce plate), the sensor that ends of the bottom of the plate will have its spectrum altered...

I am not sure what to conclude... except this way of exciting the top plate is not reliable.

A part from that failed experiment, what I think I "understand for now" is:

* Single sensor position can favour certain frequencides with respect to the top mode resonances. By multiplying the number of sensors one can have an even sound across the 6 strings but also mean the contribution of all those resonances and flatten the response in the high mids.

However all sensors will be sensitive to the main low-frequency mode. The only way to compensate is EQ, IR or Ultra-tonic pickup idea.

* If you decide to go with 2 sensors, sensor position counts. However, testing the sensor position without glue or tape is meaningless since the tone changes completely. The most sensible answer is the epoxy. However, you still need to "test". Using one string is tough because you need to know what you are looking for... This is still unclear somehow to me.

I tested my TA Amulet and apart from the exaggerated treble I can't spot a clear 3-4 Khz as I did with the DIY Pickup v8... So the Amulet problem is probably more a tape bond problem while the Pickup v8 might be position...

... to be continued.

[jonfields45]

Hi Cuki,

Not sure if the is relevant, but I've installed some more JJB 220's (2 x 20 mm) lately and I'd testify, though not controlled or scientific in any way and in inexpensive guitars, that I like the tone and balance better than a 3 smaller head K&K.

Jon

[Methos1979]

In the very near future I think we'll be reading through posts in this subforum that go something like this:

"So I just picked up my new Froggy Bottom Model C today and I looking into what pickup system to put into it. I was thinking maybe a Trance Amulet or even maybe the Dazzo but I'm really leaning towards the Cuki79 system that everyone is raving about!"

We'll all be saying that we knew him when he was just one of us!

[SpruceTop]

Cuki: Try taping a single piezo disk to the underside of the top of your Martin HD-28, *just off the rear of the bridge plate and on a centerline between the 3rd and 4th strings. This is the position where I used to superglue a single Shadow piezo round transducer, the output of which I then ran into an MXR MicroAmp Pedal as a buffer preamp, and then into a rack-mounted Furman Sound Parametric EQ. I was able to get a well-balanced, warm, and woody tone from the two or three HD-28 guitars I did this installation in back in the early 1980s. The positioning of the piezo disk may have less gain before feedback compared to a bridge plate under-saddle-line mounting but also may sound darn good to your ears. In most pubs I played in I was able to get a good tone and volume without feedback. Your use of a tape installation in this experiment won't sound as good as the usage of superglue but may give you an indication of how good it could sound if permanently mounted with superglue. The use of a single piezo disk also eliminates any out-of-phase issues that may arise compared to systems using two or three piezo disks. I don't expect anyone to actually superglue the single disk in the position I've indicated because most players aren't as crazy and cavalier as I was back then.

*Why off the rear of the bridgeplate? Because there wasn't (still isn't on the newer HD-28?) enough room on the brideplate behind the bridge-pin holes to mount the 1-inch diameter Shadow Piezo disk.

[Cuki79]

Originally Posted by SpruceTop

Cuki: Try taping a single piezo disk to the underside of the top of your Martin HD-28, *just off the rear of the bridge plate and on a centerline between the 3rd and 4th strings.

Hi Ken,

Thanks for the advice. I know it will work.

There are two SBT install that leads to VERY natural sound:

* For single sensor: The position you just advise me. There are lots of feedback about that position in the internet... Although it is known to be feedback prone.

* For dual sensors: In the PUTW #54 position (exactly like my DIY Pickup v5)

There is a good reason why I am not investigating those:

I am not going for mic-like sound this time. I found out that although pickup v5 sounds WAY more mic-like than other pickups. I have a more usable plug-in sound with Amulet +IR.

Why? Because when "pushed" the good warm & woody tone becomes mud. Once you EQ the mud out, you realize that the highs are either harsh or not cutting enough.

Conclusion: Mic-like tone is not really usable.

Also, since the pickup v5, I made an effort improving my IR process (Match files) to be more "competitive" against Tonedexter. Right now, I can really improve a SBT under the saddle line with IR.

So this project is about making the best SBT pickup I can do: under the saddle line in order to have a solid string tone (less wood, less body) and less feedback

My goal is the following:
* Passive SBT pickup placed under the saddle-line
* Reasonable output
* Reasonable impedance friendly
* Balanced tone without EQ/IR

The idea is:
1) If I have my pedalboard: Get the tone I want from IR and EQ power (Nux NSS-5 + EAE stompmix, High pass + PEQ)
2) If I don't have a pedalboard: Get something decent with just a box in my guitar case.

Note that Pickup v8.0 was already close to tick all the box... But I am not satisfied with sacrifying the high end with the tone pot. I am a bit greedy, but I want to get closer to Dazzo installed by Teddy or Amulet installed by Maury... From what I see right now it's mostly a problem of pickup install.

Note2: Although, I have strong feeling that Amulet uses PVDF sensing elements as Pickup v5 did... I can't go this road because I wanto to go passive.

Anyway, If I don't suceed, I will have accumulated enough experience to be able to install a Dazzo pickup myself with epoxy... I don't mind buying already made pickups. But I don't like the idea of buying a pickup and not being able to install it properly...

I am complicated person...

[Cuki79]

Originally Posted by Methos1979

In the very near future...

I am pretty sure it is not going to be near...

I would tell the guy:

1) Go to Doug Young's page: Choose the pickup you think sounds best & meet your loudness/usage requirements

2) Buy a Tonedexter/Voiceprint DI or make IR youreself.

3) Play and enjoy the music

That's clearly the most sensible thing to do!

If the guy is still not satisfied:
2b) Add a sunnaudio MS-2 + internal mic

[SpruceTop]

Originally Posted by Cuki79

Anyway, If I don't suceed, I will have accumulated enough experience to be able to install a Dazzo pickup myself with epoxy... I don't mind buying already made pickups. But I don't like the idea of buying a pickup and not being able to install it properly...

I am complicated person...

Cuki, You're a perfectionist and that's Good!

The Dazzo is an easy installation, and if you really want to follow Teddy's instructions about moving it around while amped up before the Loctite 6-Minute Epoxy Gel sets up, you likely get the results you're hoping to hear. Me, I just aligned the Dazzos the same as Teddy did in his Taylor GS Mini installation video and got good results. Some day, I'll post direct-to-computer Flat EQ, No Effects sound files that of course will sound like a pickup system (like every pickup system does), and that is partly why I haven't done so yet along with my left hand still healing from carpel-tunnel surgery. I do, however, really appreciate your time and effort in trying to build a better mousetrap!

[Cuki79]

Originally Posted by SpruceTop

I do, however, really appreciate your time and effort in trying to build a better mousetrap!

Well I am pretty sure my mousetrap will not be as good as commercial stuff, be it will be mine

I spent a lots of time making pickup v5 a perfect match for the Eastman and found out I had almost to restart from scratch for the D-18.

Making one good pickup is possible (but still hard), making a pickup that sounds transparent in most guitars is really really hard.

Doug Young once told me that there is no universal "best" pickup: there are good match.

These days, I think the guy who installs the pickup is as important as the pickup itself... I've heard wonderfull Anthem installed pickup (In Gibson and Martin guitars installed by the same luthier)... and some factory installed awfull sounding ones...

[Cuki79]

I made another pickup (similar but made sure to put superglue all over the surface of the discs)

I did a 2nd full install (forgot to sand the bridge top... ended with -6dB) but less resonance. I recorded every possible step (high E and low E strings with each sensor set, clamped, clamped overnight, un clamped... Honestly it is impossible to tell if it is good or not.

ended up again with too much treble.

So I started wondering about preamp and tried a Belcat IRIS-2 preamp with condenser mic onboard. It was noisy but the tone was much more balanced. I measured the IRIS-2 preamp and it turned out it was VOICED.

So I decided to make my own preamp:
* It's a charge amp so I don't have to care about impedance issues. The gain is set to +16dB.
* The EQ is a Big Muff tonestack mod. I have one pot for scooping around 1kHz and another to set the Bass/treble Balance.
You can see it here, and play with the 2nd pot. Mine is modified so I can scoop the mid to taste.
https://www.guitarscience.net/tsc/bigmuff.htm
* I added a passive High-pass filter and tuned to keep just enough "woof". (Maybe too much)

I made a demo. Of course the SM57 is way more natural althought bright. However I am pretty sure that pickup will sound great through a PA speaker. (did not try yet).


It's setup for fingerstyle... Turned out to be too nasal for pick. I imght need to scoop even more...

[SpruceTop]

Cuki, your system doesn't sound all that bad at all. I'll bet if you keep at it, you'll develop something that may surpass what's currently offered by many makers!

[Cuki79]

The scoop pot was at 0.. No mid scooping therefore it is nasal.

Pot to max (mids flat)


Pot to min (scooped mids)


Truth is probably somewhere in between

Just to add the technical aspects:
1) First stage is a charge amplifier based on TL072 (J-fet opamp). Pickup v8 has 7nF capacitance so I used a 1nF capacitor to obtain +16dB gain. It is super simple design and very efficient (no Impedance to worry about)
2) To power the circuit I used the KLON centaur internal power supply schematic. It uses a MAX1044 chip to produce 16V -8V and 4.5V so you get about +/- 12V supply for your TL072. On the internet, they say it matters for the slew rate. Basically, it means that if the opamp runs on 9V only, the opamp becomes slow if the votage is high. It means fast & strong transient are lost (think hard pick struck). I tried both with and without and it made a difference. It kind of removes the "clic" tone when plucking and adds "air".
3) Between the first and second stage there is a decoupling capacitor. I adjusted the value of it to tame the cut everything below 59 Hz. Again I tried different values until I found a sweet spot. Note that I could put a pot there to adjust to taste.
4) The second stage is a Big Muff tone stack. The big muff has a scooped 1KHz response, with one pot to boost the lows (and reduce the treble) or inversely reduce the lows (and boost the treble). Exactly what I needed for my pickup (found out about it by investigating the Belcat IRIS-2 preamp).
I just added a mod I found on the internet to be able to adjust the mid scoop with a second pot.
5)The last stage is a buffer to help the preamp drive any load. (No need for Hi-Z behind the preamp)

I think a BOSS enclosure pedal could be a good candidate for this circuit. It would make changing the battery a bliss and I could add a pot to tune the HPF cut off and one for volume. I'd have the 4 pots every BOSS pedal have.

To be continued... in another thread (I think the maximum postnumber is 100).

[SpruceTop]

I have no problem with sticking my neck out and looking foolish so here goes! I think current acoustic-guitar amplification technology is limited by the materials being used for sensing guitar vibrations. We have piezo crystals, gas-based films, and other-based films. Maybe, some new vibration-to-electric-signal material will come along or a new method or configuration of using the current materials needs to be considered and developed for reproducing a more natural amplified acoustic tone if that is a goal. For a pie-in-the-sky example, how about someone developing a spray-on, vibration-sensing/electrically-conductive film for the underside of a guitar's top? It could be used on a greater area of the top and using electronic control (a preamp) to limit feedback issues and to tweak the tone through EQ, perhaps produce a more realistic amplified tone of that acoustic guitar?

[Cuki79]

Originally Posted by SpruceTop

For a pie-in-the-sky example, how about someone developing a spray-on, vibration-sensing/electrically-conductive film for the underside of a guitar's top?

Good idea, you would have to spray the electrodes too.

I think in this case printing would be easier. Like the Martin printed top (think 00L Earth Guitar).

One could think of printing with electrodes to print and pole a thin piezoelectric "films". Then print the electrodes with a pattern to make a thousand sensors. Then the electronic would set weights on the the signals coming from the different sensors.

However I still think it would not solve the problem of the strings. IR can do this job. IR does not "sens" the strings direct vibrations.

For the only real problem is the comb filtering problem of magnetic pickups. If you solve that (you need to break a symmetry somewhere), you can just put a mag pickup on a guitar (like Taylor used to do with ES1) and then make it sound the way you want ith digital signal processing...