MFOS Envelope Generator


After a lot of test failures after another, I decided to go to Grant’s to see if he could identify what was going wrong with my prototypes, whether it was using the wrong values of parts, the wrong configuration and pathways of components, or something else. We began working on the circuit, and after about an hour I had completed most of the build, but I saw that one of the pins I had to use was already occupied, and after searching, I found out that I had been reading the pinouts wrong; I had been reading the numbers from top-to-bottom-to-top-to-bottom, like a book, and not reading the numbers ascending in a U shape around the ICs. I probably didn’t have this problem in the beginning of the quarter because I had very few tests I was able to complete before I started working on the book work.

After the build, the output was very low, so we prodded along the signal path, and found that one diode was installed backward. After that, we had a working attack-sustain-decay generator. We tested all of its functions on his Serge system, including envelope repeat mode, and both “gated” and “triggered” modes. One thing we noticed was that the output for the self-triggering mode was not as high as the ‘gated’ and ‘triggered’ modes.

One interesting part was using the breadboard to ‘flip switches’ and trigger the envelope instead of connecting hardware switches.

I then took it home and plugged it into the Oscilloscope and the modular and took a few pictures and videos showing the features:

CAM00663 CAM00662



Bench Progress and Tour

My current setup is more than adequate, I have found. This quarter, once the materials arrived, I found that for the tasks I have set out for myself, as well as occasional repairs and other electronics work not connected to the ILC, I did not need anything else, and more and more I have used every single tool on this bench.



I’ll be going over every piece on this bench in detail, what I have used each tool for, what problems I have had that these tools eliminate, and how I organize the space.


The first, and arguably most important tool on the bench, is the soldering iron. I opted for a mid- to low-grade Weller, and it was a major improvement over the soldering iron I had. I immediately found that after I bought a good iron that my soldering skill improved immediately, showing that the tools really do make a difference. I ended up buying a special tip for it as well, as the one that came with the iron was too wide. The temperature control dial on the front of the base I usually keep between 4 or 5 unless I am desoldering.

The solder I chose was the thinnest available. The spool on top of the larger spool is slightly thinner (0.022″-Dia), and I’ve had it longer, so I have been using that. The larger spool is a half pound of solder and is slightly thicker (0.025″-Dia). Both have a roughly equivalent percentage of tin, lead, and rosin, so between the two, I shouldn’t have to use much else for a while.

The little blue plastic thing is 5ft of braided, rosined desolder wick. I really don’t like using this very much, I find it sloppy, and the rosin leaves huge marks on the PCBs. I, hopefully, will be able to replace a lot of the duties of this stuff with the desoldering pump I bought last month.

A friend gave me their whiteboard, so I have been taking notes about needed supplies on it, as well as experimenting with circuit diagrams on it.


The PCB holder I’ve bought has become an indispensable tool for populating PCBs and soldering parts to them. I don’t think I would have a need to upgrade it, unless I had a larger PCB than this could accommodate. Behind it is a jar full of things, Sharpies, fine point pens, and zip ties. Also in the jar is the desoldering pump, which I have yet to use.

The plastic box to the right contains some vintage Russian NPN transistors for experimentation.

To the right of that box are the two bench power supplies that I have bought, one I have received in the mail today. One problem I’ve been having with these are that I don’t really have banana jacks or cables to reliably connect the power supply to the breadboard. I will have to purchase leads, or connect them to the board with single-core wire. Another thing is that these supplies aren’t digitally controlled, so to reliably get similar output from both I will have to use multimeters to test the output and adjust the two supplies.

The jar on top of the supplies contains alligator cables and breadboard jumper wires. The alligator cables allow me to connect larger components that cannot be affixed to the breadboard, and the jumper wires allow a quick connection between point on the breadboard. I had originally opted to get a different jumper wire setup, there are wires that are flush with the breadboard instead of looping around, but I felt that it would be harder to keep those wires in good shape than these. I find that it’s a double-edged sword, because it is far harder to reliably prototype circuits without a clear visual idea of what’s going on. Instead I have a lot of wires that end up nesting around ICs.


Next up is a breadboard (and a TL072 Op Amp). This breadboard is huge, but I’m quickly realizing how small it actually is, especially with the jumper wires that I bought. I find myself trying to stretch the circuits out over the whole board instead of conserving space so that I can reliably see the pathways the jumper wires are making. Still, as I get better at prototyping circuits I’ll most likely get used to using less and less space, and this board will be all that I need.


Here are the two batteries I have been using to power the prototypes. Two batteries are necessary to get +9V, GND, and -9V.

The box to the right of the batteries is something I made, it’s a passive mult that connects 3 1/8″ jacks and one 1/4″ jack to probe connectors that go to the oscilloscope. This way I can hear and see signals coming out of the prototypes at once. The 1/4″ jack is so I can analyze signals coming from my synthesizers, and any effects pedals I have. For the probe connectors, I have mounted a screw for the alligator clip, and a loop of stripped wire for the hook attachment. Thus far, I have used the box to analyze every signal, I haven’t yet had a reason to detach the probe from the box.


This is the parts drawer I have labeled and organized. Each kind of part is arranged vertically in ascending value, starting with passive components and ending with transistors. There are a few other, smaller components that don’t really need labels, as the parts will probably be switched out at some point. I would like more space, as I’ve organized each drawer as shown below:


All of the static sensitive components are in labeled static resistant poly zip bags (pink), or in their original packaging (as can be seen in the IC drawer). The resistors are in normal plastic bags, but I ran out of them before I could bag the capacitors, so they’re in the pink bags as well. Because of the bags and organization I have employed, I’m finding that I really would like more drawers (or bigger drawers). The organization system has deteriorated somewhat the more I use the parts, but I am really happy I decided to do it this way, as parts are easy to find and easy to return back to order.


I’m extremely happy with this oscilloscope, it was the cheapest digital scope I could find, and still it is possibly the most useful tool on the bench. I can immediately see any DC bias, any noise, count the frequency of any waveform, and analyze two different signals on the same screen. Another useful feature is that at any given time I can capture anything that is on the screen and record any data that accompanies the capture onto a USB storage device. There are so many functions I haven’t used yet, I will try and work through all of them before the quarter is over.

Below the Oscilloscope is some ribbon cable and a few perf boards I bought in anticipation of needing them for next quarter’s work.


A labeler. Pretty essential stuff for a workbench.


Two digital multimeters is a good idea, but I initially wanted to buy just one. I had ordered the smaller one from Amazon last quarter, but it was coming from China and it took a long time to get to the states, so I bought the larger one at ERI. I end up using the smaller one out of convenience, and because it has a better display (it’s backlit), but the second one will come in handy when I need to keep an eye on a steady voltage while testing parts of the circuit. Both of the multimeters have transistor checking and diode checking, but both lack frequency counting (which is something the oscilloscope has).


A messy box of tools. This has the Dremel and all of the dremel attachments, screwdrivers, pliers, wire, snips, cutters, strippers, tweezers, an x-acto knife, and, for some reason, a flashlight.


Here’s a box of larger parts that wouldn’t fit into the mini drawers. Top left to bottom right:

  • Rack screws
  • Potentiometers
  • 1/8″ panel mount jacks
  • Davies 1900H clone knobs with set screw
  • Davies 1510 clone knobs with set screw
  • Heat shrink in a few different sizes (spent too much for these)
  • ON/ON SPDT switches
  • some 9V batteries
  • Some AAs
  • IC sockets
  • Header pins and sockets, standard, battery connectors and an LCD screen taken out of a toy
  • Pushbutton switches
  • 5-pin DIN jacks
  • Serial cable connector

All-in-all, I’m very satisfied with the tools I have, I’ve used every single one during this and last quarter, and I’m finding that I require less and less, and the needs become more and more specific. For example, I’m no longer thinking that I need something to hold another thing, or something to snip this or that, it’s more that I’m missing a specific part of a specific value.

Space is another issue, I’d like to progress vertically, find a way to store my reference books near the workbench, get a better or bigger toolbox, and more tiny drawers. I found a desk at Habitat for Humanity, but I didn’t have the cash for it. It was perfect, had three levels and the desk space was roughly equivalent of this one.

MFOS Multi-Mode Filter (Dead)

Having received all of my parts, I’ve attempted to prototype something a little bit more complex, with a total of 19 parts. One of the big problems I have had is that the parts I ordered haven’t been of the best values. I seem to be missing some very common part values, like 1uF capacitors and 20k resistors, even though these are very common, and are used in many of the circuit diagrams I have access to.

So, with that out of the way, and knowing that I can easily put resistors and capacitors in series or parallel to get other values, I started to work on this filter.

It turns out that most of the important capacitor values in this circuit (the drain capacitor for the biased output of the LM13700, the capacitor at the input stage of the filter) had to be made using other capacitors, as I was not shipped 560pF capacitors, nor 1uF capacitors. I was also not shipped 20k or 200k resistors.

This made prototyping the circuit unnecessarily complicated and large. It also made diagnosing errors a more complicated task.

I started from the pins of the LM13700 and worked my way out from there, from pin 1 to pin 16. Just as an idea, there are 4 pairs of resistors meant to make up for the lack of 20k resistors, and 3 pairs of capacitors making up for the lack of 560pF and 1uF capacitors.

Once the circuit was built and I went over the diagram one more time, I powered the circuit.

The oscilloscope showed a DC bias, which was not expected, as I have set up the batteries to allow +9V, GND, -9V. It would also show some kind of oscillation, even though no input was connected.

The LM began to smell, and I touched it and it was hot. So I fried the OP-Amp somehow, and it is impossible for me to tell where I went wrong because, as you can see, there are no voltage reference points on the diagram so I can measure with the multimeter, and the breadboard looks like a nest of wires.

I would like to revisit this circuit once my second power supply comes in the mail, but until then, I have to disassemble this and throw away the LM.

MFOS Square Wave Oscillator

After the 555 Timer Oscillator experiment, I decided to do some real audio oscillators. This time I was able to breadboard from a circuit diagram and not a video or tutorial. The design I chose was a simple square wave oscillator from the Make: Analog Synthesizers book by Ray Wilson.


This Oscillator requires a bipolar supply with a ground, and, unfortunately, the power supply I decided to buy did not have a ground, it instead has two leads, one for positive voltage, one for negative voltage. I decided to use two batteries connected in such a way that the 0V reference point was between the two batteries (you can see this from one of the images below). It also requires use of an op amp to generate the signal, and I decided to use the TL072.

To get the circuit correct, I decided to look up the datasheet on the TL072, and used the pinouts found there to model the diagram on the breadboard.

My one worry while wiring up the breadboard was that a 9V bipolar supply would not be enough to power this circuit (this was before I was able to calculate the load and power loss of a circuit), but it was a complete success.

As you can see, the oscillations were clean and musical (other than being relatively stuck at high frequencies due to the audio taper potentiometer I had), and the square wave kept its duty cycle all the way up into the high frequencies.

I was able to get the circuit to oscillate between  40Hz and 2.1kHz, which was surprising to me, given that the last oscillator I prototyped had a relatively low frequency range (probably due to my modifications and nothing else).

More than ever I am becoming aware that I need a bipolar power supply and proper potentiometers or trimpots.






555 Timer Oscillator Experiment




After the long wait for components to come in the mail, I decided to take a stab at making a 555 Timer Oscillator, one of the easier projects to do with very few components. I ended up modifying the circuit that was contained within the following tutorial:

During the video, you will hear him talking about the discharging capacitor used. He said that he picked an electrolytic capacitor that was 22µF so that it would blink the LED at a rate he could see.

Now, I knew this would be a square oscillator because without waveshaping, the 555 timer outputs digital pulses. Since I had to adjust the frequency of the oscillator in order for it to be audible, I decided to use a capacitor that discharges at a lower voltage. I also modified the design by adding in a potentiometer by replacing the second resistor with a 100k potentiometer, which adjusted the rate that the capacitor discharged, thereby adjusting the frequency of the oscillator.


I had trouble finding the right place to insert the potentiometer, but after I tried “replacing” one of the resistors with a variable resistor, I noticed changes. Apparently, after the voltage already went through two resistors, there wasn’t enough of a voltage to change with the potentiometer.

As you can see, the resulting wave looks and sounds “dirty”, meaning that as the oscillator gets higher in frequency the resulting wave gets distorted from a square shape to something else. The duty cycle of the higher pitched oscillations is more like 70/30 instead of the 50/50 of a traditional square wave. It is also pretty telling that the crest of the wave looks to be rising from what would be ‘100%’. This is probably due to the capacitor used and it’s tolerances. The duty cycle change is most likely due to the 555 Timer not having any biasing on the input pin.

Another weird thing is that the potentiometer I used was an “A” potentiometer, which has a logarithmic response curve and not a linear one, hence the twisting of the knob going from a gradual increase to a relatively quick increase to maximum frequency output.

The leads from the circuit to the oscilloscope went from the drain pin of the capacitor and the negative supply of the battery. This caused a DC signal. If a bipolar supply were used, this would have been an AC signal, but at the time I didn’t really know how to do that with batteries.

CAM00497 CAM00498CAM00499 CAM00500

Week 5 Distortion/Overdrive Pedal Build

Not having many materials, by Week 5 I decided to purchase something to construct.

After doing a bit of research on which pedal to buy, in part guided by the folks at’s “TB-303 Pedal Distortion Test“, I decided that I liked the resulting sound of the Fulltone OCD the best (near the bottom of the page). It seemed like a good-sounding circuit that didn’t rely on germanium parts to add to the tone, it does it quite well with ICs and modern transistors. It features an LP/HP switch, which stands for “High Peak/Low Peak”, as well as a “Tone” knob, an input level knob, and a distortion/drive knob.

Having bought the kit and chassis from Mammoth Electronics, I was able to not fuss around with locating parts and having them all come in separately, I could just open the box up and start soldering, however…

The kit, the website I ordered from, nor any of the materials I was sent in the box came with a proper Bill of Materials. Sure, everything was there, but no actual link between the parts and the numbers on the board (i.e. what value is C5?). After that I realized I had ordered a kit from Mammoth, but the design was somebody else’s, who preferred that everyone building his pedals had to sign in to a forum and hunt around for a proper bill of materials.

Why? Not sure. Either way, many things COULD have been avoided had Mammoth had permission to distribute their kits with a BOM. I emailed Mammoth to find out exactly why they didn’t indicate that it wasn’t going to ship with a BOM, and they basically said that the owner of the designs at preferred to not have anyone distribute ‘his’ circuit designs. I think, more than protecting his property, he wants an ego trip. Either way…

The BOM I managed to find was, as I would later learn, for a different revision of the pedal than the one Mammoth had sold me. So, I ended up “not having” a capacitor I needed, while having a capacitor I “didn’t need”. Fortunately, so I thought, I had a capacitor lying around that was of the value the BOM asked for, so I soldered that in and switched it on.

The LED lit, but no signal was passing through the pedal when the SPDT switch was “ON”. So, I went through a very laborious and painstaking process of desoldering a small capacitor I had inserted from a fully populated board and trying out the one I was supplied. Seriously, the board has so much flux from desoldering braid on it right now that it looks like I baked snot on the board. Because of this I have purchased a desolder pump and some 99% Isopropyl alcohol for cleaning boards.

I think this goes to show that no matter what preparations you might do, how much research you’ve done in advance, there is always something that’s out of your hands you won’t think will end up fucked up, but will be fucked up somewhere down the line. Had I had a proper Bill of Materials of the version I was constructing (and there is no indication anywhere of the version I had bought, nor any links to any other than the most current version on the forums), I would not have had to spend an hour desoldering one component from the board. I am baffled that Mammoth, nor have received hundreds of emails a year about how they distribute their information. It’s such a daft process it doesn’t even logically work out.

Regardless, I finished the pedal. Since I wanted to use 9v batteries instead of plugging the pedal into an outlet, I didn’t attach the barrel adaptor for the power supply. What I found was that, even with the pedal off, the battery drains anyway… Who thinks that’s a good idea? Seriously. The battery is draining because the circuit (I guess) needs to be active even when off to pass the signal through it (as it doesn’t do this when the battery is removed)… Probably one of the worst design choices I’ve ever heard of. Why even include a battery clip at all if you have to expose the board and open the entire chassis everytime you stop using the pedal? I can’t believe it! I have an Ibanez digital delay pedal that does the same thing… Just pass the signal through a pole of the SPDT switch that’s connected when the pedal is turned “off”. It’s not tough, or complicated, and I would do the modifications, but all of the SPDT terminals are already occupied, so that would mean adding another switch, or buying an SPDT switch with more terminals.

So instead my solution was to simply add a switch that disconnects the circuit from the battery…

Anyway, once I had finished the circuit and everything was plugged in, I put a square wave from the Rubicon oscillator into the distortion circuit and looked at the output through the oscilloscope.

The LP/HP switch basically changed the gain of the circuit, there was a noticeable difference in the tone as well. The lower the signal was (in Low Peak mode), the distortion was less clipped and more rounded, the square wave had a tiny peak at the rising edge and a rounded fall. In High Peak mode, the distortion was louder and more pronounced, and the waveform was slightly mangled. The ‘resonant’ peak of the square wave was higher, and while the fall was rounded, there was a slight dip in the waveform.

That slight dip on the falling edge of the waveform was adjustable by the “Tone” knob, it moved the dip’s position from about 25% of the cycle to about 50% and then back again.

Overall, all of the tests turned out well, the circuit sounds nice, and I might try swapping out a different op-amp to see what that changes. I tested it with a square wave oscillator, tested it with one of my synths at line level, and took it to a friends’ house and ran it through their TB-303.

I suppose that I’ve learned that while kits in general promise a lot less fuss, it actually invites quite a lot more fuss, because less of the project is in your hands. So there’s really no way to win, other than having a huge parts collection and having a lot of time to design circuits. That or just having a lot of time to wait for parts to come in the mail that you’ve specifically ordered.

Passive components Capacitors Active components Aptly titled pedal

Panels and Nothing

The panels and brackets came in a few days ago, so I put the bad stereo jacks I had and put the pots in to make myself feel better… Ha.

Even if I were to try and source parts to get this finished, I’m finding even more things that are wrong.

For one, the Rim Shot panel came with 4 holes for potentiometers and only 1 came with the Rim Shot. Why? Well, apparently the other three pots are for a tuning mod that I didn’t know about because there was absolutely no information about it whatsoever in the documentation. That would be all well and good, I’d just have one knob and three holes, except the mounting bracket that’s keeping the PCB from flopping all around behind the panel is bolted onto the panel using the potentiometer. So without at least two, the bracket swings back and forth. I’d do the mod, but these look like a specific kind of pot with a specific kind of shaft, and there isn’t much wiggle room for fudging a few of the parameters.

Another thing, the mounting bracket and the spacers that came with the panel for the Snare don’t allow any room for the noise board behind the panel. Meaning, even if I did have headers, I would still have to buy shorter standoffs to install it into a rack, or the noise board sticks out so far that it would short against any adjacent module.

I love the 909, I love soldering, I applaud for what they’ve done for the Eurorack community, I think Hex is a genius of sorts… But obviously not the practical kind. These modules have been hell (nearly) from start to non-finish. I couldn’t order the PCBs because the website wasn’t clear when things were out of stock, the BOM and the Mouser cart he put together were both missing integral parts that are mentioned nowhere in the documentation, the panels come from some guy who does a really good job but doesn’t talk to you at all after you order and doesn’t notify when he ships things…

The PCB and the designs are absolutely perfect. I love the attention to detail, the matte soldermask, the incredibly attractive arrangement of the parts, the wide solder pads… For a half a day I was in electronics heaven… for two months I was a nervous wreck.

I think this goes to show that the more minds involved, the more hands that have to come together for one particular task, the greater there is that something won’t get done by a deadline. If Hex offered complete kits from his website that someone looked over before shipping them out, I’d have two wonderfully noisy 909 drum voices. But… things don’t work that way, and I thought that 10 weeks was ample time to acquire anything I’d need to finish, I just didn’t anticipating needing more than I needed (?). This makes me think about work situations, about being asked to get a product ready before a deadline… How much work is actually involve in double- triple- and quadruple-checking absolutely every minor and seemingly insignificant detail before pressing onward. It requires a lot of skill, and probably requires more than one brain to do it properly.

Well, I hope I don’t lose credit, but if I do, I’m not sure who to blame…