Finally I decided to make a recording...
This is the sound with the new transistor in the input stage (2N5087). I had to amplify the effect part by 3dB to make it sound a little more even with the clean signal. I'm playing a fretless yamaha bass with EMG-DC pickups through an Ampeg Rocket Bass amp (I love this amp). The signal is going into my computer via the line out, so there's no EQ. I recorded with Audacity. Here's what I'm doing:
0:00 - 0:30 Clean
0:30 - 0:54 Effect with no resonance
0:54 - 1:05 Some resonance
1:05 - 1:23 More resonance
1:23 - 1:53 Max resonance
1:55 - end Distortion
For the distortion part, I keep on changing tone, depth, rate, and resonance. I know I start with no res, then I increase the tone, then I adjust the settings until I get something I like. I had to change the playing style because slapping doesn't come out so nice with distortion.
Here are some pictures of an unpopulated board. The Fuzz Factory part is blacked out with sharpie. My next step is to populate the board with different JFETs and see what that does.
Tuesday, July 24, 2012
My main thing is guitar pedals. This is one of the most ambitious ones I have done and I still have a lot of work to do on it. It's a clone of the AP-7 Jet Phaser from Roland. It's from the 70s and if you can find one on eBay it would cost you around $800. This pedal is really old school and sounds awesome!
I'm close to getting that sound, but there are a few problems I have to figure out first. The preamp stage which you can see is off the pcb is part of the original schematic, and it wasn't on the main pcb either. I added that to beef up the input since only the distorted setting sounded loud enough. Adding the input stage didn't make much difference. I think the problem is because the schematic asks for a high gain RF PNP transistor, and I used a 2N3906. Yeah... It didn't help the volume, but the tone got a little worse.
I also tried modifying the gain of the opamp in the mix stage, but that made the distorted setting too loud, and the phase effect couldn't be heard in the normal setting. I tried changing the resistor values in the output, and nothing. I have a feeling that if I use a different transistor in the input gain stage it would solve my problems. I bought 30 new transistors (2N5087) I'll replace for the 3906 soon.
For the JFETs, I used 2N5457s and they work well and sound fine (very clean without the 3906 input stage). The problem I had was that the circuit needs 8 matched JFETs and I only bought 30. I chose the ones that matched best, but the average VGS was around -0.4V. I think the circuit was designed for -4.0V VGS so the biasing was off. I just modified that without a problem by replacing one of the fixed resistors tied to the bias trim pot.
The distortion, on the other hand, sounds SO good. It has a really vintage sound and I love it. I think I'm gonna take it out and make a distortion pedal out of it. Maybe.
Anyways, it phases, sounds good, but not good enough for me. I designed the circuit board and got the PCBs made through DorkbotPDX (OSHpark), so I have three boards and I'm going to populate the next board soon. I bought 100 new JFETS (BF245) with around -4V VGS. I hope they make it better.
I also put a ZVEX fuzz factory distortion on the circuit board because it is so small and I always wanted to adapt that pedal for bass. I haven't tested the fuzz factory since I have no Ge transistors. I'm not an intense audiophile, but I'm pretty sure this pedal needs germanium to operate properly. If anybody knows how to get it to work with another transistor, please tell me.
Anyways, I put the fuzz factory on because the circuit is super simple and could easily fit onto my board. It would make my pedal look pretty funny with 10 knobs! I was thinking I can use that circuit for the gain stage for V2? Here are some pictures. I'll put up a video or sound file when I get it working the way I like.
This is a really simple project I did at work in my spare time (after hours). I was working in a lab and there were all sorts of components and machining tools at my disposal, so I decided to put together a circuit I had in mind for a while. I've been using the triangle wave generator as an oscillator with great success in the past, so I decided to put four of them together, each being controlled by a separate LDR (Light Dependent Resistor) for the frequency. This makes it light controlled. They all have the same base frequency (slightly offset due to component variations) but they are so sensitive that if you hold your hand over it just the right way, you can get a chord or some weird screaming sounds.
This is my MIDI sequencer I built using an arduino microcontroller. I put it in a cigar box which I bought for $1 from a tobacco shop in Harvard Square. It generates a random sequence of notes (tone row) without repeating any pitches and spreads them over three octaves. This is the fundamental rule of 12 tone music, the idea being that there is no tonal center.
The algorithm takes the chromatic scale (numbers 0-11) and randomly selects one note to put into a new array, shifts it by 0, 1 or 2 octaves and deletes it from the original array. This repeats, each time selecting a different note and randomly selecting the octave. This algorithm is a function which is called whenever the current sequence is finished if the tone row button is held down. The sequence is sent out to MIDI, and I've been using it with a Midisport 2x2 connected to Reason.
The controls give you the ability to change the duration of the notes (they're all the same length), velocity of the notes, reverse the sequence, modulate the whole sequence up or down by a semitone, produce random note durations (arrhythmic) and generate a new random sequence. The ability to modulate by a semitone means that in that transition there will be a repeated note, but heck it sounds cool so I just left it.
A fun thing to do is to crank the duration way down and modulate the pitch using the pitch knob. Also holding down the "new tone row" button while you do this makes it sound pretty crazy. It's also fun to flip the arrhythmic switch and let the Arduino make it's own music. Sometimes I like to let computers do the composing just to see how soulful a dead thing can be.
For version two, I'd like to get rid of the repeated note when reversing the sequence and also allow the ability to invert the sequence. I'm not really sure how to do the inversion if the tone row is spread across three octaves though. Maybe a switch to randomly select octaves while the sequence is running would be cool too.