Natural tone preamp

To go along with a 6 string build I am [slowly] working on, I have also been thinking about a new on-board preamp design. Unlike conventional preamps, this one focuses on preserving the natural tone of the pickups while using multiple pickup wiring combinations to provide tonal variety rather than tone controls. It is entirely digitally controlled, and thus programmable, but uses an entirely analog signal path.

The idea came from the problems I have with most existing basses, both active and passive:

• Limited to 2 or 3 pickups
  
• Too many controls and switches with too few good combinations
• Inconsistent pot tapers
  
• Level variations across different pickup combinations
• Tone variation and high frequency loss due to coil loading, especially for increasing numbers of pickups
• Poor S/N ratio due to low signal level and unbalanced cabling
• Noisy electronics
  
• Scratchy pots
• Complex wiring
• Dead batteries

The natural tone preamp attempts to solve those problems by implementing these features:

• Support for up to 4 pickups with up to 6 coils total
  
• Minimalist controls with a single master volume and single rotary preset selector
  
• Any combination of split/series/parallel humbucker coils with separate phase and pickup levels per pickup for every preset
  
• Programmable pot taper
  
• Level normalization across all presets so there is no change in volume when switching presets
  
• Dedicated buffer amp for each pickup for true no-load sound for all combinations
  
• Ultra low noise op amps
  
• Line level output for high S/N ratio
• Balanced output option for reduced EMI
  
• 100% digital potentiometer signal path for "scratch-less" operation
• Clean control wiring with 2 - single gang potentiometers and 4 total connections required
  
• Phantom power option for battery free operation

Surface mount is unavoidable with this amount of silicon but the relatively low component count and nothing with a pitch less than the standard SOIC at 1.27 mm and 1208 size packages should make it manageable to solder. Unfortunately this also means I need to finish a PCB design and get it fab'd before I can see if it even works. As such this is still in pre-prototype phase but the design I have includes:

• 1 - ATmega328P microcontroller (BYOB for this part, ex: Arduino Pro Mini)
  
• 1 - AD5206 hex digital potentiometer
• 3 - MAX395 8-channel SPST analog switches
• 3 - OPA1692 dual op-amps
• 2 - single gang linear potentiometers (almost any common value will work)
  
• A handful of passive components

The digital controls will use an off-the-shelf Arduino, which allows nearly complete separation of the analog and digital sections and also prevents having to solder even more tiny components. The remaining analog and power components will be on another board and the two boards linked with about half a dozen wires.

I have not completely decided on the pickup switching options. One option not possible in the current implementation is reverse phase for series connected humbucker coils. The entire pickup can be reversed, and the two coils can be phase reversed when wired in parallel, just not when in series. Anyone particularly like out of phase series humbucker coils?

Having never designed a preamp before it is safe to say I am in way over my head. However, I learned to use KiCad to do this project so at least I have a schematic to share:

Click to zoom. Even then it is an eye chart!

I am interested in feedback on the design and component selection and also if a project this crazy has appeal to anyone besides me. If there is enough interest I wouldn't mind publishing an open source design so anyone can build their own. Don't hold your breath though, as having this finished is still a long way out.

 

Thanks Charlie! Yes, a good portion of the right side of the schematic (digipots, balanced out, phantom power) is stuff I had already been working on, and still plan to use on another bass, as a "bolt on" to your open source preamp.

The left side began as a thought experiment on what it would look like to replace 10 physical switches (3 phase, 3 series/parallel and 4 on/off switches) on a 4 pickup bass with a single "preset" control containing all the good combinations with none of the bad.

Glad to hear your success with the OPA1692. It is low noise and low power compared to many really low noise amps, but at a typical 650 uA per amp it is is still on the ragged edge of my phantom power budget when the microcontroller isn't sleeping. If the digital and analog parts really don't get along I might have to program the microcontroller to power up the op amps after it finishes booting right before starting to sleep to attempt to limit max. current.

Although I am planning phantom power for myself there is nothing saying, with a few component changes, it couldn't run off a 9 V. Steady state draw should be around 4 mA, and it will run down to 5 V or less, so an alkaline battery should last about 150 hours.

 

Haha, yeah, I might have jumped into the deep end a bit early. Here is a scenario instead:

The tone from a bass is largely about the type of pickups and their arrangement. Tweaking the tone knobs on a preamp won't make a P bass sound the same as a J, or a MM etc. So, for a while now, I have had an idea to build a bass with huge tonal range by using more pickups and combinations. Spread them out a bit so you have a couple where they go on a J, one around where a MM sits, maybe another near the P position.

After the pickups are in place there are other ways to get an even broader range of tones. Things like series/parallel switches on a couple of single coils or a humbucker, or reverse phase for a pickup or two. Coil taps. You get all sorts of new, and occasionally awesome, tones.

That can be done already, and some have tried, but as pickups and switches are added problems begin.

The more volume knobs, the more highs are lost from the load put on the pickups by the pots.

The more switches, the more possible combinations but also the more complication. Wiring becomes a rats nest and you need lots of space for all those controls. Even after the cavity cover is back on to hide the mess, remembering where the best sounds are is nearly impossible -- at least for someone with as bad a memory as me.

Was that cool fat sound sound with the switches on off on on, with the volumes set at 5, 7 and 10? Or was that on off on off? Get them wrong and it sounds muddy, dead, etc. Trying to switch while playing live is tricky because you have to reset so many things and then try to get the levels right so you aren't too quiet or too loud when you start playing again.

With this preamp you wire up the controls with just 4 wires to 2 pots. Before closing up the cavity, hook up a USB cable to your laptop, plug it in, crank up the volume and start playing.

Now, instead of physical knobs and switches, you set switches and knobs in software. Change a pickup config, say switching a humbucker from series to parallel coils, and see how it sounds. Tweak a volume, tap a coil, reverse phase on a coil. Keep noodling until you find something you like and then save it.

Repeat until you run out of cool tones. Hopefully there are a handful of good ones. Make sure the levels are consistent across all of them. Save it all, unplug the USB cable, stick the cavity cover back on and enjoy! Your favorite presets are now just a tweak of the preset knob away.

At least that's the idea!

 

Well... I just placed an order for 22 pots, 40 switches, 20 op amps, plus a bunch of other components. Once it arrives I get to have fun hand wiring a bunch of stuff to build a prototype. Assuming it works, creating a PCB will be the next daunting task: the current layout has 221 pads!

 

UPDATE: I received the parts order, and got a bit of time for building a proof of concept for the preamp on the weekend. So far I have hooked up the Arduino, 1 pot chip, 1 analog switch chip and one dual op amp. If I can get the Arduino to control the pots and switches then it's just a matter of configuration and adding parts to make the rest work.

The first hurdle was soldering my first SOIC packages and, good news, no problems -- with just a cheap pencil tip iron. I even managed a SOT23 regulator.

Here is my first attempt with the op amp. For size reference the vertical rows of 4 pins on the left and right are DIP8 spacing (0.1"):


and pots:


and the VREG:


Not the prettiest looking, but every pin was connected with no bridges.

After I got the hardware wired up, I needed some software. After a few evenings it isn't finished yet, but it has a CLI to configure it, reads pots to adjust volume and presets, and has a custom pot taper function. Probably more than I needed given the early stage of this.

Next step is testing that the software can talk on the SPI interface and can write to the pots and switches. Hopefully I can get that done in the next few days. If that works then I will wire up a two coil prototype that can switch between these 7 configurations:

• No coils connected (pickup off)
• North coil only
• South coil only
• South coil only with inverted phase
• Series
• Parallel
• Parallel with South coil inverted phase

...and check how it sounds hooked up to an actual pickup.

 

Thanks! Unfortunately my vision definitely exceeds my ability

 

Quick update: I put together a few basic parts but cannot control the pots over the SPI interface. I am pretty sure the pot chip is fried, and I think I know why. When I designed this I planned on phantom power, but I did not build in a mechanism to power the digital pots and switches when only the microcontroller was powered. As such, when I programmed the microcontroller it then applied digital signals to the unpowered chips, which is generally a no-no. I will make a design change to connect the power circuits so that when the microcontroller is externally powered all the ICs are powered as well, and try this again.

 

Yes, the idea of the XLR is to go straight into a low-Z input -- mixer, DAC, amp, etc. My original idea (and the reason I called it "natural tone") was to have no "EQ" at all. Everything coming out would be as accurate a representation of the true sound of the pickups as possible. However, as I went through the design I decided I could add a tone control (LPF) to allow limited local tone control equivalent to a typical passive bass.

That's not to say it wouldn't be possible to add EQ as it would just take a handful of extra components. But, as you point out, there are about a million options out there that already do that, and I generally prefer to do tone and effects off-bass.

That Vaudoo stuff is interesting, and I already need a USB jack somewhere to do programming anyway. One issue is that hooking up USB is a bit of a pain, and the connectors tend to be very fragile. I would probably be more inclined to use a phantom powered charger instead. But the main issue is that I just generally don't like batteries. Even rechargable ones typically only last a few years before they need to be replaced. So I started thinking about alternatives and it should be possible to use a super capacitor instead to provide power for several hours when going into something unpowered (unbalanced input or balanced input without phantom power). I will look into that a bit more for the next version.

 

Good to hear from you Charlie!

Thanks for the tip, I hadn't seen the 2145 yet. So a quick look at the datasheet shows a bit lower quiescent current, which is nice, a bit higher voltage noise, but much lower current noise which is probably dominant on a high-Z input amplifier. Looking forward to hearing about your experience in case you find anything else notable about it.

 

I continue to make design improvements, having now changed to using a high efficiency switching regulator to maximize the tiny amount of available power when running on phantom power.

Having made that change I also decided to add a battery and have the preamp automatically choose the right source based on available options, with battery being the last resort. That way if phantom isn't available everything still works.

Disliking dead batteries, I also decided to add a smart sleep mode into the controller. If you leave the bass plugged in for an extended time, something that would kill the battery on a typical active bass, the controller will detect that it is running on battery and, after the volume has been left turned off for a few minutes, will shut down the LEDs and op amps and put itself into a low power mode that will allow the battery to last more than a year. Turning up the volume up past 30% will wake everything back up with the settings back exactly where they were.

Yesterday I also did some experiments with addressable LEDs to use as high-tech fretboard marker alternatives to luminlay. I wanted to see how much light I could get out of the leftover energy after the microcontroller and analog circuitry had taken their share. The LEDs themselves are trivial to setup since I already have the microcontroller available to drive them. To my surprise, the amount of light is substantial even at low power levels:


That picture shows 12 addressable RGB LEDs (enough for the typical markers on a 24 fret board) running on just 1.5 mA each for a total of 93 mW. For perspective, a 9 V battery could run those for about 30 hours.

They aren't going to blind anyone but they should be clearly visible in normal lighting, and quite bright in the dark. Plus, because I can control the brightness and color of each LED individually, they can be used as visual indicators. For example they could indicate which preset is the preamp is set to when changing settings, or provide a battery level indicator when first turned on.

I also made some progress on the PCB. Having given up on hand-routing I tried freeroute with enough success that I should be able to get a working 2-layer board in about a 1"x3" footprint.

 

Yes, I could put all sorts of gadgets onboard, but fundamentally I want to use all the technology to keep this easy to use, so I am resisting urges to build more complexity in.