Intro: Zeta Reticuli: MIDI Controlled 10-band EQ and Multiple External Effect Interface
Approximately half my lifetime (20 years) ago I used to have literally drillions & drillions of ideas that I still feel aren’t a complete waste of time. One of these brilliant ideas was to be able to control a guitar wah pedal from my computer. At the time I had scant musical equipment to match my scant muscial abilities. I had scavenged what I could afford in my extremely limited means from pawnshops, flea markets, thrift stores, & the Penny Power. Among this discarded mess were an Atari 1040st computer which has built in MIDI, a Korg Poly800MkI synth, & a Boss DR550 drum machine. I had been successfully using the Atari to sequence the synth & drum machine, but wanted to also control effects for my guitar at the same time and in the same way. But this was the early 1990′s. I didn’t have the skill nor the funds to do anything of the sort.
For those interested here are some YouTube links to a blurry picture of my son playing with the Korg over very randomly spliced together and poorly preserved material salvaged from a cassette made at the time: SLSC-B01 & SLSC-B02
I recently realized that some of these old ideas, due to a better financial situation and the current space age state of technology, are now easily within reach. The first of these ancient plans to be made finally real is a variation on the original MIDI wah pedal idea. Instead of using an analog sweeping wah I went with a 10 band equalizer. This can serve as a choppy wah effect, but I preferred the 10 band EQ as the foundation for the project because I felt that it would have a wider range of applications and, as will be explained later, the board can also simultaneously control an external real wah as well as a whole bunch of other effects through an expansion board capability.
The first 17 second demo recording I made of the Zeta Reticuli on SoundCloud: “This is me doing nothing but playing a single chord over & over to a drum machine. My fret hand does not move at all. I used my sweet-ass lucite guitar through a Proco Rat distortion pedal plugged into my MIDI EQ which is plugged into my micro-Marshall. This is mic’d into a mixer with flat EQ & recorded directly by my desktop computer. No other effects or trickery. All the EQing was programmed into a sequencer on a different computer 5 minutes before I recorded this.”
Step 1: Overview & Operation
I believe that a user will need to be somewhat familiar with using MIDI to get any real benefit of this device. It could be used as a very basic tone control, but those ends would likely not justify these means. The next simplest case I can envision is using a small MIDI keyboard controller to control the Reticuli in real time. My personal plan, for which I made it, is to create complicated tonal patterns and subliminal currents using a software sequencer where I can control precise values for each individual controller at precise times. This scenario has a lot of prerequisites in time and equipment. Although, truth be told, I’m still using mostly cheap items gathered over the years from flea markets, thrift stores, and pawn shops; and I much suck at music. The real measure of this project is that it cost under $100 and has exponentially paid that back in education and entertainment value. I spend that much just taking my family to a bad movie.
Another way to use the Zeta Reticuli is with a purpose-built or virtual controller. Hopefully I have done this already and included video(s) below to demonstrate and clarify the different modes and MIDI cc usages.
So anyway: The first selection that needs to be made to use the Zeta Reticuli is mode which is explained in more depth below and is chosen by issuing a MIDI patch change to patch 0 or 1. The firmware currently gives 2 options for mode, plus 10 user programs that can be either mode.
- patch 0: DIRECT: Each frequency band is controlled directly. This can be either by using MIDI continuous controllers 22 through 31 or by the velocity value in Note On messages for Middle C (Note #60) through A (Note #69)
- patch 1: PSUEDO-PARAMETRIC: An imitation of parametric behavior which uses the same Controller and Note On velocity options as above, but processes them differently:
- cc22: center band: The center frequency of the peak/notch.
- cc23: center level: The level of the center frequency. If it is a higher value than off-center level there will be a band-pass effect, if lower there will be a band-block or “notch” effect.
- cc24: off-center level: The level of the furthest off-center frequencies outside of the width parameter.
- cc25: width: The width of the parametric curve effect, or number of bands above & below the center frequency.
- patch 10-19: USER PATCH 0-9: Storing user patches is discussed below. Once a patch is in memory it can be recalled by sending the Zeta Reticuli a MIDI patch change message for patches 10 through 19 (user patch # plus 10). Only the mode and associated EQ parameters are stored. None of the parameters for controllers associated with an expansion board are stored.
- cc6: gain level: Controls input level to preamp. See Input and Preamp below
- cc7: volume level: Controls output level. See Output below
Storing User Patches:
As with everything else in the Zeta Reticuli there is more than one way to do this to accommodate different equipment and usage possibilities.
The first way of storing a user program to memory is by using MIDI continuous controllers 70 through 79, sending a value of greater than 64. The controller to use for a specific user memory location is just 70 plus the number of the memory slot. Slot 0 is MIDI cc 70, slot 1 is 71, and so on. When the Zeta Reticuli receives one of these controller messages it stores all of the EQ parameters to the appropriate EEPROM addresses. As noted above, only the mode and associated EQ parameters are stored. None of the parameters for controllers associated with an expansion board are stored.
The second way to store user patch data is through MIDI SysEx messages. This may be more complicated, but more convenient – at least in my case. A 7 byte SysEx message that will cause the Zeta Reticuli to store the current EQ settings to a specified user patch is as in the image titled "SysEx Store User Message"
- The Zeta Reticuli’s SysEx implementation allows it to also get patch data both to and from external software such as a MIDI librarian. This will only work over the USB port since there is no standard MIDI out port on the device. The format to retrieve a user patch over USB is exactly the same as above with the exception of byte 4 Command. This byte will need to be 0×11 (dec 17) in this instance, to signify a user patch request command. As soon as the Zeta Reticuli receives a properly formatted SysEx request it will respond over USB with an 18 byte reply as in the image titled "SysEx User Patch Data Message"
Controlling Additional Effects:
No point having idle pins, so I’m putting them to use extending the control capabilities through an 8 pin header. The header pipes the 5 volts, the 2 signals necessary for SPI (SCLK & MOSI), and enough chip select lines to control 4 more 5206 chips installed on expansion boards. The expansion board could be an individual stand-alone effect, which is the way I’m planning to implement an analog wah. It could also be just some pots accessed through a header to control an existing external effect that has been modded to be controlled by the Zeta Reticuli, which I’ve breadboarded using a cheap chorus pedal with great success.
Other SPI-controlled chips could be used in an expansion board, such as digital outputs to control relays for an MIDI controlled automatic signal router (Another plan in the works!), but the firmware would likely need changed to handle this. Another option that I’m considering is an SPI multiplexer to drastically increase the number of available CS lines. At this point the number of MIDI continuous controllers may then be the bottleneck and mapping multiple MIDI channels may be required. I’ve mentally bookmarked these and other options but to be real I haven’t even finished the updated version of this controller yet, let alone sketched out the first expansion board.
Step 2: The First Version Hardware
- Power: 5VDC is brought in from a wall wart through a barrel connector.
- CPU: The Teensy. I love these lil guys! I started using them both because of cost and because their on-board USB (NOT FTDI) allows using various USB profiles natively, thus not having to do any fancy driver/emulator footwork. This IS a plug’n'play Arduino MIDI device.
- User interface: The Teensy sends display data to an HD44780-compatible LCD in 4-bit mode and takes input from two tactile push buttons. The push buttons choose which MIDI channel the Zeta Reticuli responds to MIDI input from. The LCD screen displays the current channel, level of each of the 10 EQ channels, and the current operating mode.
In this version of the circuit I have all 16 pins for the LCD pushing out on one header and a second 4 pin header taking 5v +/- out to the push buttons and returning the two inputs for the Teensy. After ordering the PCBs I quickly kicked myself realizing that I didn’t need that many lines as some are redundant and some are unused. This has been changed in the next version.
- MIDI data: Input from (most of my) MIDI gear comes through the standard 5 pin DIN connector. I used a PCB-mount on the first version of this board, but will almost definitely use a panel-mount connected to a 90 degree header on the next version to free up board space as well as make enclosure layout more flexible. This data is isolated from the Teensy by an H11 optocoupler which I also robbed from the old MIDI interface.
The current firmware allows for simultaneous MIDI to be received over the USB port. Theoretically you could add a MIDI out port and send and receive over the USB and MIDI ports, using it as a rudimentary MIDI-to-USB adapter. I plan on exploring this option in future versions, though I fear that timing and lag will be a serious issue.
- Equalizer: The EQ section is based around (2) Rohm BA3812L chips. Honestly it’s just taken directly from the 10-band app note in the datasheet. The basic premise is that each frequency band is set by two capacitors. The first (A) sets the resonant (targeted) frequency and the second sets it’s bandwidth (Q). The datasheet has the formula for determining these values. The values and their resulting frequency bands that I used are listed below. Once a frequency band is isolated from the rest of the signal in this way a potentiometer is used to vary it’s volume before it is recombined with the rest of the signal and set out through the 1/4 inch jack output. The AD5206 chips (6 10k digital potentiometers) are used instead of standard pots so they can be controlled over SPI by the Teensy.
- Input and Preamp: Audio input enters through a 1/4 inch jack and is routed into one of the 10K digital pots which is controlled by MIDI continuous controller 6. I call this the “gain level” as this moderates the input through an LM386 circuit to boost the signal. Before deciding on using a preamp stage, while developing the circuit in a breadboard form it worked beautifully as long as I had a distortion pedal hooked before the Zeta Reticuli spaghetti. A clean signal just wasn’t up to driving it. This version of the PCB has the option to solder a cap across the opamp pins 1 and 8 to have a gain of 200. Without the cap the preamp has a gain of 20. The additional gain creates a lot of distortion, so in reality I am instead using a toggle to switch the chaos in and out. In the next version I plan to use a 3-position switch for preamp bypass/low/high levels. Even better would be a relay controlled by the Teensy so the options could be selected over MIDI.
- Output: Also routed into one of the 10K digital pots which is controlled by MIDI continuous controller 7 and referred to as “volume level”. I’m unsure if my separate use of controllers 6 and 7 could be problematic in some situations since, if I’m understanding their intended use correctly, are meant to be able to be combined together allowing a single 14 bit volume level.
Frequencies and Capacitor values:
While researching my project I also found a similar, albeit manually adjusted, kit that has some great info in its instructions PDF including many pre-calculated capacitor values.
The values I used in this version:
FREQ, CAP A, CAP B
33Hz, 3.3µF, 0.082µF
56Hz, 2.2µF, 0.047µF
100Hz, 1µF, 0.022µF
250Hz, 0.47µF, 10000pF
500Hz, 0.22µF, 5600pF
1kHz, 0.1µF, 2700pF
2.2kHz, 0.068µF, 1000pF
4.1kHz, 0.027µF, 680pF
8.2kHz, 0.015µF, 330pF
16kHz, 6800pF, 180pF
The schematic pictured here is the first version. The only correction or change from those listed in the PCB section is the pinout of the power connector.
EQ chip,eBay,Rohm BA3812L
The first draft of this is also the first PCB I’ve ever had professionally manufactured. This has been a bucketlist-level dream that has intimidated me for years and finally going all in on it, despite my mistakes, has been very eye opening.
At the recommendation of a friend I used ITead Studio to make them and I am turbo-happy with the job they did as well as the cost. It took about a month to get the boards back. The vast majority of that time was the shipping. Even before they arrived I had already made a number of changes to the design. Once they were in hand I found that I had also made several rookie mistakes as well:
- The Teensy is positioned behind the MIDI connector in a way that blocks any normal micro-USB connector. I was able to get a reliable connection by bending one of my cables’ connector. A 90-degree connector is definitely called for in this layout.
- I couldn’t find a pre-made library for the AD5206 digital potentiometer chips, so I made one. I had changed the grid size in Eagle and had forgotten which means that the width between rows of pins is way wrong. I was able to bend the pins and make it fit though, so no harm there.
- The MIDI DIN socket that I snatched off an old Mac MIDI interface has slightly different spacing on the 2 front mounting tabs than the Eagle library component I used. I neutered the one I used so it would fit.
- The worst and most embarrassing mistake is that I got the power connector wrong. So so soooo wrong. This means that instead of the through-hole connector that I wanted I’m making do by using a panel-mount connector that doesn’t care how bad I messed up in the past as long as I hook it up correctly in the present.
Step 3: The First Version Firmware
The firmware is very much a work in progress. If you are using a Teensy for the first time be sure to first follow the Getting Started guide from PJRC.com and that all of the appropriate software is installed. When using the Arduino IDE make sure that in the Tools menu that Board is set to “Teensy 2.0″ and USB Type is set to “MIDI”.
Step 4: The First Enclosure
I struggled with the design of the first enclosure for this. I was putting so much weight on the the appearance and durability (and still am, if I’m being honest) that I was almost frozen. While I am somewhat happy with the way this particular enclosure has turned out, I really don’t feel like it is it’s final incarnation. Aside from the base design aesthetic, I’m not super happy with the alignment of everything and there are some obvious places where I filed unevenly. I’ll likely make another of this exact enclosure just to prove to myself I can get it right and then start making others of different designs until I hit on the design that completely satisfies me.
The original idea was to capture some of what I feel looking at both vintage home stereo equipment from the 60′s/70′s and vintage synth equipment. We have some SAE home audio gear from the 1970′s in our home that I really appreciate the look of. They’re rack mountable, but if you don’t want to mount them in 19-inch racks they have an additional enclosure with wooden sides that the equipment will mount in just like an ornamental single-unit 19-inch rack. After several frustrations first cutting then bending the metal I decided to go a different route until I can revisit that design with more focus on building up the needed tools beforehand. This particular enclosure is made from the coated aluminum case of an old temperature controller from my work.
I still wanted the wooden sides. I toyed with ideas for more rectangular, and even triangular, ends. I ended up starting out with very simple plain ends so that I would get a bit more insight into what I was dealing with. When I remake the case I will most likely try making something with built-in cable routing.
One thing that was always certain and unwavering was that I would be ammonia staining the wooden parts. If you have never used this method I would highly recommend trying it out. I’ve refinished many things using this technique and developed my own twist on it. In my opinion there is nothing that matches the depth and intricacy of it. It’s very simple, works on about any hardwood, and is a relatively slow process which can take up to 48 hours to complete a dark piece. This means you can stop it at exactly the level of stain you want. Personally, I especially like it dark on oak and walnut.
My process is very simple. Once the piece is ready to be stained I rub it with 2 or 3 coats of linseed oil, letting them dry out in between. This first oiling step really adds depth and seems to accelerate the process as well. You will need to have an airtight place to let the fumigatory staining process happen. This can be a trash bag framed out with scrap wood or, as in this case, a Tupperware-type container. Just make sure it’s airtight so the ammonia doesn’t evaporate away and leave the area smelling like a meth lab. For large pieces it may be beneficial to rig up a small case fan or some other way to get the air to circulate inside the ‘tent’ so that the bottom doesn’t stain faster than the top. After the last coat of linseed oil, dry or not, I put a random, unmeasured amount of ammonia in the ‘fume tent’ and space the article up out of the ammonia. In the Tupperware I had the case ends sitting on wingnuts. You should try to keep the contact area between the piece and it’s spacers as small as possible. The points of contact will receive less ammonia and may show lighter. For this reason it’s also not a bad idea to reposition the piece on the spacers every few hours.
Step 5: Examples
Zeta Reticuli drone demo 01
A quick demo I made of the Zeta Reticuli MIDI EQ running a short, looped pattern. Again, I'm strumming one chord. The guitar is plugged directly into the Reticuli with "chaos gain" switched on. The Reticuli is plugged into a small Marshall amp with it's gain turned all the way up as well. Sound was recorded by the camera, so is not the best. Not that it needs to be.
2012-12-13: I added an oscillator for the pseudo-parametric mode in the firmware. I'll update the Instructable soon, but in the meantime here is video of the Zeta Reticuli manipulating raw guitar feedback with it's new oscillator settings. It's in it's temporary "fail case" and I'm using a separate interface I made with a second Teensy to control the Zeta.
Testing the new oscillator firmware with guitar feedback. I made a separate MIDI controller for the Zeta so I would have portable hands-on control of each setting.
Here's a clip of a better recording of a short keyboard loop being manipulated by the Zeta, as well.