How do I get into synth DIY?

This question pops up regularly. So why not write up a blogpost about the topic and save myself the effort of replying the same thing every time. Heck, maybe drive some extra traffic to the site and propel this blog to new unknown heights paving the path to world domination. Maybe I should start plastering adds all over the place and cut up this post into 10 smaller ones on different pages… and…

Anyhow. How to start with this whole DIY synth thing. It’s seems such an insurmountable amount of stuff to learn. And yes, it is. It’s a lot of stuff to absorb. But it has never been easier. Thanks to the wonders of the internet, most information needed is only a Google search away, for some more esoteric questions you might need to scroll to page 2 of the search results though.

Now, first things first. We’re all unique snowflakes, so everyone’s progress, starting point, ambitions and all that crap will be different. There’s not a single path to enlightenment and what might come natural to me might be hard to grasp for someone else or vice versa. That said, If you’re one of those people that just have to look at a piece of electronics for it to spontaneously combust, you might want to reconsider diving into this without very good home insurance.

The most important thing is this: Don’t try to learn everything at once. There you have it. Understand this and most importantly, accept it. Take it with small steps and eventually you’ll get there. You won’t be designing your own synth from scratch just yet, but it is an achievable goal in the long run.

This blog post isn’t intended to learn you everything you need to know, it’s rather a framework or maybe even, dare I say, a guideline to starting with synth DIY. It’s a non-complete list of things you will need to pick up on the way, in a more or less logical order.

1. Soldering

The most important thing to remember about soldering is that if it’s smells like chicken, you’re doing it wrong. There are a ton of videos on youtube on how to solder and it’s advisable to watch a few until you know that melting metal with a very hot stick can be dangerous, and are fairly sure you grasp the basic safety precautions involved, such as being able to identify the hot side of said stick.

The only thing left to do after you imprinted those things in your mind, is practice. It’s something you just need to do to learn. A friend of mine just gave me 20 jack plugs and a stretch of cables to solder them on. Good cables always come in handy anyway and by the 20th cable you kinda get the hang of things. Most of these cables didn’t last long, but that’s beside the point.

Alternatively, search for some easy kits, like the VCF-1. You want kits where there aren’t too many components cramped in a small space giving you some room to work. Preferably not with expensive components either. Avoid anything SMD, you don’t want to try and solder those tiny bits by hand as an introduction to DIY. You might do things you will regret later out of sheer frustration.

2. Identifying components

Get a multimeter, life is too short to deal with those pesky color schemes on resistors. Also get a magnifying glass, apart from making you feel a bit like Sherlock Holmes it also makes those markings a lot easier to read. I really think some manufacturers make it a challenge to see how faint and tiny they can label their IC’s and still get away with it.

You know how I said it has never been a better time then this to learn this shit. Here’s where that comes into play. Whatever doubt you have about something, the answer is a Google search away. So, right now, you might want to google for ‘guide to identifying electronic components‘ and come up with plenty of relevant information on the basics. If you need to know something about a specific component, just type in whatever markings you see and the datasheet about said component will surely pop up. Google is just awesome that way (or Bing if you happen to be so inclined, not judging here). If you’re not sure about something, always check and never assume anything. I’ve seen identical components from different manufacturers with a different pin-out for example.

Why do I put this much emphasis on this. Because polarity matters people. Some components need to be placed in a certain way. Electrolytic capacitors, leds, diodes, transistors, IC’s. If you don’t, things can go wrong. So when opening up your first kit, check what each component is and whether or not polarity is an issue. If it is, figure out how to place them. Often that’s indicated on the PCB, but sometimes the designer will assume this basic knowledge so you better have that handy.

3. What first kit?

So where’s a good place to start? That’s open for debate, but what I do know is where NOT to start. Your first kit should NOT be a power supply! I know it’s tempting, they look easy to build (and they are), and it’s the first thing you’ll need. But I can’t stress this enough. Don’t go play with the big 220V/110V boys when you’re not up to it.

Those voltages can be lethal or cause any number of other issues, you know, like fires and stuff. Even just getting a little nudge from 220V isn’t a very fun experience. But regardless, it’s also the foundation on which everything else is build, so you want to be sure it’s done properly and it reliably delivers clean power at the specified voltage to whatever you’re building. So you won’t need to worry about this when troubleshooting.

Eurorack is a great place to start with DIY because, due to it’s nature, the whole circuit of a synth is broken down in its various modules. So as you advance in your understanding of things you can tackle more and more complex parts, yet even the simplest things you build will remain useful in the future. We all need a simple mixer in our rack or a buffered mult. So whatever you’ve build the first day is still useful for the years to come. You’re also not limited to your own DIY efforts to populate your rack, you can add some modules from a nearly endless list of manufacturers.

What I’d look for in a first kit is this:

  • Low part count
  • An easy circuit to understand
  • No complex calibration process. You want an oscilloscope first.
  • Avoid things with thermal compensation circuits and paired transistors such as oscillators and certain types of filters.
  • Don’t go for modules generating CV controllers unless you have an oscilloscope to visualise the output.
  • Make it something useful in your particular setup

Now, you build up complexity as you get yourself a few more kits and soon enough even a more complex kit like the metal-o-tron II will feel a lot like building a lego set. So, time for some blatant self-promotion, some easy kits to start with:

4. Troubleshooting

After building your first kit, flying high with pride over your creation, not unlike that Icarus dude, the crushing weight of reality you will sober you straight up. It probably won’t work.

Maybe you’ll even release a bit of the magic smoke, start a small electrical fire or blow a fuse. Or it just stays dead as a doornail (who came up with that) . You might be the exception, ‘the chosen one’, who makes electronics work the first time around, but statistically speaking you will have fucked up in one way or another.

So troubleshooting circuits is a skill you’ll need to develop alongside your building skills. It’s mainly based around logical thinking so that might be great or awful news for you depending. What’s most important though is that you’re get used to reading schematics which basically starts with memorising a bit of those weird symbols and figuring out how these symbols relate to the actual components on the board. hint: It’s pretty straightforward.

Now, you will see that some of these weird symbols come into similar configurations over and over again. Look up the following stuff: Opamp configurations, filters, mixers to get you started. And somehow, over some time you will start to understand what all these configurations actually do and how they work. This process involves a lot of poking around with a multimeter, scratching your head, some mild cursing and trying to understand specific parts of a circuit. Google is your main tool here. Whatever you are trying to understand, there will be at least a dozen or so people explaining that very thing in a wide variety of ways, so there is bound to be some understanding in there if you’re dedicated enough.

The main take-away is this though – cheesy quote alert: Don’t look for the solution, try to understand the problem. If you do understand the circuit design and intentions, the solution will present itself. And now I’m going to look where I can get a motivational poster with this slogan printed, maybe one with a bonsai tree or something.

After a while, troubleshooting finally start to become easier. It will remain something of a black art though, so keep your local demons happy and well fed.

5. Ordering components

That’s a whole can-o-worms on it’s own. It’s not a bad idea to buy a few complete kits until you at least have some basic grasp on how components look. Then you can start browsing through the immense catalogue of Mouser, Farnell, TME, … or whichever component provider you prefer and delivers to your neck of the woods. Here you’ll pay the price of learning quite literally. You will order the wrong components, it’s like a certitude. You’ll feel bad about it and you’ll learn to check things like physical dimensions as well since, apparently, they do matter. Mouser has, as of writing, no less than 660.377 resistors in their catalogue and a filter system of dozens of parameters. So picking the right one can be a little bit of a hassle.

But ordering your own components pays off in the long run. It’s cheaper than kits if you only need to buy the PCB & front panel. You don’t have to pay for someone to count out and put components in plastic baggies per kit. You also can build a bit of a stock of certain components you use often. Prices drop significally if you order larger quantities, so there’s that.

6. So what tools do I need?

It seems a good idea to add a small list of things you should get when starting out. There is a bit of an investment to be made at first, but it are all tools you’ll need sooner or later and you’ll use them for years. Many of which you might already have as they’re useful to have around the house anyways.

  • Soldering station: You can start out with a cheap ass one, but it’s a good idea to at least get something from a respectable brand if you plan on using it regularly. Expect to pay anywhere between 100€-200€ for a good model to start with, a cheap one can be got from 30€.
  • Solder: Lead based is easier to use but worse for the environment than non-lead based solder. I use non-lead, which I have to since I’m selling stuff, ROHS compliance and all.
  • A magnifying glass, because things can be tiny yo.
  • Some kind of cheap ass amp/speaker setup. Why do I say cheap ass, because you’ll be abusing the hell out of it. Pushing 24vpp square wave signals below 20Hz will be hard on any system – yeah, it happens, don’t ask – so you might as well save you the grief of blowing up an expensive system, once the initial testing phase is over it’s safe to bring it into the nice studio with the fancy speakers. Don’t use headphones if you can avoid it, unless permanent hearing damage is on your bucket-list.
  • Multimeter: Something you’ll need from the get-go. A small portable multimeter with basic resistance, voltage and capacitor measurement capabilities will set you back around 50€-100€, maybe even less. You don’t need extreme precision, so don’t go spending thousands on a finely calibrated Fluke meter. Or do.. what do I care, it’s your money, do as you like.
  • Various wire-cutters and screwdrivers and such. You’ll gather these as you go.
  • Oscilloscope: These things can be expensive, but audio usage isn’t very demanding, if you can view a waveform in the audio range and count frequencies, you already have 80 to 90% of your scope needs covered. A DSO138 can be had for like 30€ and will do just great. ‘real’ scopes will set you back €300 to infinity.
  • Bench power supply: Not a bad thing to have on this list because they have short-circuit protection. Saves a ton on replacing blown fuses on your PSU in the long run. You’ll need a double channel one for use as a symmetric power supply though.
  • Computer with Google, or Bing, or Duck Duck Go, or any other search engine. A tablet works as well, so you could go that route.

7. From schematic to PCB

When you go Googling you’ll notice there are a lot of resources out there. It’s easy to find schematics for most vintage synths and even for a lot of recent Eurorack modules. On this site, but also from other manufactures like Erica synths, Bastl, mutable instruments, … So there probably will be a time you think ‘Why can’t I make those PCB’s myself?’ And guess what, you actually can. It’s a whole new chapter of the DIY journey. How exciting.

First, you’ll need to get your hands on some software to do this ( let’s forgo the option of drawing them by hand, it’s no longer the middle ages ). The basic idea here is that you draw the schematic in said software, then assign footprints ( which is the place the components take up on the PCB ) then arrange them all on a PCB and export the files for fabrication. There are various platforms with all kinds of Pro and Cons. Personally I use KiCad since it’s free, open source and if it’s good enough for use at CERN, it’s certainly good enough for me. It has its quirks and weird interface thingies but it does the job. YMMV as they say. Eagle and Easy EDA seem both very popular amongst hobbyist as well.

Whatever software you chose to draw schematics, expect a moderately steep learning curve. UX design hasn’t seeped through yet to that part of the software world. You’ll need some experience in components and reading schematics to actually pull this off, hence it’s not one of the first things you should try. And it will require a bit of time and persistence to get the hang of things. There are tons of Youtube videos by dull engineers guiding you through the basics and advanced concepts of designing PCB’s.

8. Hello China?

Once you have found your way through the PCB design software, the time has come to get them manufactured. You could etch your own PCB’s which involves quite a bit of hazardous chemicals and even some voodoo, but I won’t go into that here as it makes little economic sense IMHO. It’s a time consuming process and getting everything needed is probably just as costly as getting them manufactured. But if you’re into hardcore DIY, who am I to tell you it’s a bad idea.

As with all things in electronics you’ll find the best deals in China, the manufacturing plant of the world. Ordering low-volume PCB (5 to 10 pieces) is something that used to be very expensive. Nowadays there are several Chinese manufacturers offering these kind of services for very little money. Again, do a google search, JLPCB, ALLPCB, … Expect to pay around 5€ – 10€ for 5 PCB’s, depending on size.

Off course, for these prices you can’t expect a very personal service. You send in the Gerber files and you’ve better thoroughly checked them, because that’s how they get made. Unless there is something wrong with the files you’ve send that would make production impossible, things just get build as specified.

Euhm Gerber, gerbil what? It’s the standard format for PCB production used by most PCB production facilities. You can generate them in your PCB design software of your choosing and they contain all data needed for production. This isn’t a how-to guide, so I won’t dive deep into that rabbit hole, but Google can help you there.

Just pointing out the obvious, you will make mistakes in your PCB’s. Depending on the complexity I typically go through several iterations before I have a finished product. Maybe that’s just me though, but I keep making mistakes by reversing potentiometer pins for example, so pots turn the wrong way. (I know, it’s not hard but my brain just refuses to process it correctly)

9. Designing from scratch

The holy grail of DIY-ing. Come up with an idea, design a circuit, build it and presto. Quick and easy.

If you’ve reached this far into the blog post, dedication isn’t something you’re lacking. That’s a good thing, you’ll need it. Get a breadboard, simulation software and start designing things. I would advise to start modestly though, and not directly try to design something rivalling the CS80 in terms of functionality. But given some time, there is little to stop you from doing so.

From this point on there is little advice I can offer. It comes down to studying existing schematics (luckily, easily available through, yet again Google) and figuring out how they work, then transferring that knowledge to whatever idea you’re trying to implement. Easy peasy.

I found circuit simulation software incredible helpful when designing new circuits. I use iCircuit a lot for trying out simple ideas and concepts. While not very accurate, the fact that it’s a realtime simulator with great usability makes more than up for it. It’s fast to use and that’s great for a quick idea check. If more accuracy is needed, LTspice is, horrible interface aside, the go to tool. I don’t use it often since the interface on OSX in mindbogglingly arcane.

While simulation can offer great insights into a circuits design, nothing beats actually building it on a breadboard so you can toy around with it in real life. Hook it up to actual CV/audio sources and see/hear it work, or not. Sometimes even a heartwarming fire can occur.

Whatever type of circuit you’re aiming for, you’ll always will find yourself up to the neck buried in information and discussions over all kinds of circuit details or approaches. It’s up to you to distil all that into a functional circuit. Analog circuitry found in synths has been roughly around for 60 years, so there is no shortage on implementations of various ideas and solutions to almost every problem you can run into. However, the perfect one hasn’t been found yet because ‘perfect’ is a very subjective concept. A filter might be perfect from an engineering standpoint, a musician might prefer a rougher approach that’s more ‘characterful and warm’ or whatever arty-farty term is used to describe it and that’s what makes it so much fun and why it makes sense to make yet another iteration of an active filter design.

A perfect design simply doesn’t exist. Circuit design is a continuous trade-off between the laws of physics, the laws of economics, the laws of well.. laws and a lot of other ‘law-type-things’ I can’t think of right now but sure as hell need to be taken into account. A lot of great classic synths own their unique sound to shortcuts engineers took to make production cheaper, easier or just meet the deadline.

10. Conclusion

It has never been this easy to start DIY-ing. The information available on this topic is, putting it mildly, extensive. Online shopping makes purchasing all needed goods and services more affordable as they have ever been. Technological advancements have made circuit design much easier and doable with a lot less math. Basically, shit got so easy even you can do it.

tl;dr: use Google