This page is here to catalogue my build of the Music From Outer Space Synth DIY Experimenter, click any pictures to enlarge them.
I officially started soldering this project up on the 18th November 2013, many weeks of planning had already taken place at this point however!! I previously attempted Ray Wilson’s WSG project, this was years ago and my soldering was far from perfect – the filter never worked and it eventually died of unknown causes, but mainly linked to my lack of patience and amateur soldering!! This never put me off Ray’s work however and now, with a good few years of experience I decided to dive into something complicated.
The SDIY EB is described as a modular sound effects synthesizer – this is because it’s oscillators are quite simple and don’t tune or track in a musically accurate way, but other than that the project builds a very flexible standalone synth! Altogether the synth has:
- 2 voltage controlled oscillators
- 2 voltage controlled filters
- 2 AR envelope generators
- 2 voltage controlled amplifiers
- 1 noise generator
- 1 cv mixer
- 1 audio mixer
- 1 attenuator
- + an optional sample and hold module (who wouldn’t?)
You can think of these as individual modules and they can all work independently of each other, and built separately following Ray’s schematics if desired. The kit came in 3 separate PCBs, the main board, the power supply and the micro sample and hold
The Case Designs
I have asked a friend for help here, I always struggle with enclosures, and don’t have the best setup to get into building them at home, so I drew up some ideas and passed them on to Matt Harris. I wanted a wedge shape for the box to make performance more comfortable, and to use wood wherever possible including the front panel. Matt has a workshop based in a yard that works with reclaimed wood furniture – so there is a chance we will use materials from here.
Edit: I measured up the main panel mounting components and have found that the front panel and the back utility panel could only have a maximum thickness of around 4mm – this has made me feel a little concerned about using wood for either panel as I plan to take this instrument around a lot, and I opted for banana jacks for the patch points which tend to need a bit more welly when plugging and unplugging than other patchers. I am basically concerned that a lot of hard work might be ruined if the wood splits. I am know thinking about using something a little more standard like aluminium sheeting or perhaps aluminium dibond.
Here are the cabinet designs I passed on to Matt:
The back of the case will also hold a panel will have the main output, a ground link for patching to other devices and a power switch. The front of the case will have a carry handle.
– – –
I started with the power supply PCB so I would be able to test the main circuit when needed – I have run into a snag as I bought the wrong capacitors, the voltage rating required for capacitors on the board is 35v, mine are below that so I have had to reorder. In the mean time I added the diodes, resistors, tantalum capacitors and regulators, in that order:
– – –
As I’m waiting for the capacitors for the power supply PCB, I decided to start on the main board. Today I completed the two VCO’s, it took an hour and a bit. I soldered the resistor rows, then added the capacitors, the transistors and finally the IC sockets for the TL074 and LM3900:
– – –
I started the dual LFO section today, working my way down the main PCB. I used the same approach adding the resistors first, followed by the ceramic capacitors, the transistors, the electrolytic capacitors and finally the IC socket for the TL072:
– – –
I soldered up the dual VCA section today. Spent some time reading over the documentation first as I wasn’t sure how the amplitude modulation inputs worked, being slightly confused by markings on Ray’s front panel design, but I understand now – one (or more) of the inputs link directly to the IC but there is also a dedicated input that goes through a depth control pot that sweeps from the VCA being 100% on to being 100% modulated. I followed the same technique as above, first soldering the resistors, then the ceramic capacitors, the bi-polar electrolytic capacitors and finally the IC sockets:
– – –
I made the decision that a wood front panel would be too risky because of the stress of constant plugging and unplugging. So I ordered some 3mm aluminium sheeting cut to size, 3mm is almost certainly a little excessive but it will be solid. I need to apply the panel design and drill the aluminium myself, I am going to try a technique using lazertran paper to get the panel design on to the aluminium, there is a good tutorial here: http://clacktronics.co.uk/diy/lazertran. Anticipating this, I have been updating the panel design, it now features drill locations as opposed to my artistic renditions of knobs and buttons, this is by no means final but I am getting closer to finishing it:
I have run into another component related snag – I don’t have the correct component package for the trim pots on each filter and the white noise section, so I have had to order those. I am also running out of solder and I’m getting a little low on cash to throw at the project for the time being. I did however complete the dual AR generator section today and I will probably get round to doing the two mixer sections tomorrow:
– – –
I have officially run out of solder and money at this point so I have to wait a couple of weeks before I can crack on with electronics. Last night I did some more work on the front panel and I’m almost happy with it now save for a bit of space that needs filling in the bottom left, a logo perhaps? Here is how it is looking: (it will be printed in one colour so any white will actually show the aluminium – unless I decide to paint the aluminium first), not sure whether to go for the inverted look or not as the components on top will be colourful
Capacitors arrived, fairly fast as well! Shipped from Hong Kong. To celebrate I used the last of my dwindling solder reserves to finish of the power supply:
– – –
My aluminium and special Lazertran paper arrived so with help from Matt Harris and his printer, I attempted a print of the front panel today. This was my first ever attempt working with the lazertran method, it was nice to see the result which looks very professional, however I wasn’t happy enough with the result. I believe my mistake was not being careful enough removing water and air from below the adhesive before putting the panel in the oven, as a result the panel is covered in tiny bubbles and cracks. I imagine these will chip away very easily and take the ink up. I am going to sand it down and try again. Here are some pics from the work:
The image is mirrored as the ink side is applied to the surface and the adhesive is clear.
The panel after being buffed with meths
Applying the lazertran design
The result after a good few hours roasting
– – –
I have had another failed lazertran attempt since last posting but after some interweb conversations I have discovered that I was using the wrong type of lazertran!! The inkjet version isn’t suitable for aluminium baking so I am trying to get a company to send me some A3 lazer printed designs. In the mean time, I have been finishing the main PCB, today I completed the “kludge” signal/audio mixer section and the white noise section. I just have the filters left to solder up and then I have the micro sample and hold PCB to do, which is tiny. Here is how it is looking:
The mixer section in the prototype area
The white noise section below (the mixer uses two spare op-amps from the TL-074 chip that boosts the white noise signal) Notice the orientation of the transistor – I hope I have this right – I used the 2n5172 as an alternative to the 2n3904, which is supposed to be noisier. The three legs are arranged differently on the 2n5172 package, so it doesn’t align with the board markings, it is turned 45 degrees anticlockwise and the middle leg is bent forwards rather than back, you may not be able to make it out in the picture:
– – –
Today I finished the last of the PCB populating. I soldered up the filter section on the main PCB, then populated the Micro Sample & Hold. No more soldering to do now until the front panel is complete and the panel components are mounted, then I will start the epic wiring mission. Here are some pics of the completed boards:
The filter section:
The completed main PCB:
Main PCB with ICs mounted, mounting the ICs with more soldering to do is a little risky, but I decided that mounting them before wiring would probably be much easier than trying to do it through a jungle of wires
The Micro Sample & Hold PCB
– – –
A couple of days ago I received some new Lazertran paper prints from Bear Print. These were printed on the lazer printer version this time, (not ink jet like before) which is the type you are supposed to use for aluminium and oven heating as I now know. So last night I managed to get far better results transferring the design to the front panel, hooray!! Here is a pic, it needs a little touching up but it is very useable:
Next week I am going to drop into the workshop at the local uni and drill out the mounting holes, then I can finally start wiring!! Yes I know the panel has 2013 printed on it, we can keep that a secret right?
– – –
Today I took the panel to the local college and drilled out the mounting holes. I used a centre punch to help stabilise the drill bit and used a pillar drill to do the job. I wasn’t too happy with the results. Firstly the drill bit looked kinda wobbly when spinning in the machine, and a couple of the holes were a little off centre possibly as a result of this (I blame the tools). Secondly, I made a bit of a dumb mistake, I was drilling into a piece of scrap wood to prevent the aluminium from burrs as the drill punched through the other side. The first hole drilled perfectly, I then moved the panel for the second hole but left the scrap wood were it was, when I drilled into the panel, the drill went through the aluminium and then into the previous hole in the wood, making a nice burr in the panel. I then made another mistake of not fixing it there and then and just went on to drill the rest of the holes. The first burr pressed the panel up slightly creating a gap, this had a knock on effect and every subsequent hole created a burr. Fortunately aluminium is very soft and I discovered that I could clear this up very quickly at home using my craft drill with a sanding bit, disaster averted! Or at least many painful hours of filing. Here are the pics:
– – –
Today I mounted all the components to the front panel, looking good!
– – –
Spent a couple of hours soldering the panel today, I managed to finish all the point to point wiring for the components mounted directly to the panel. Next job is making the connections between the panel and the three circuit boards – not much left to do now!
– – –
So I haven’t updated this page for a while, the project got put on the back burner for a little while as I was busy with work commitments and other DIY projects but I have now nearly finished the synth – unfortunately I managed to fail to document Matt’s work on the case itself – I stained the case with a dark toner and cut the ugly hole out the back – no the best job but I still like the way it looks 🙂 This pic shows the PCBs mounted in the box for the first time:
And here it is with everything bar the Micro S&H and power supply wired to the panel (eek!!):
Upon testing the synth, a couple of things happened. First of all a capacitor exploded in the power supply – because I had somehow managed to solder it back to front!! How I managed this is beyond me. With that issue solved I have a number of problems to work through with the synth mostly revolving around very bad crosstalk between modules and the mixer section not working properly. It is almost certainly down to one or more simple short circuits that will be easily remedied, but I am yet to pluck up the courage to wade through that spaghetti!!
– – –
So… it’s now 2015… and I finally finished this synth!!
It turned out that I had made one biggie – I think in 2013 my brain told me that aluminium was not very conductive, so having the metal contacts of the banana connectors touching the front panel would be fine. Turns out I was totally wrong and as a result, every patch point was being shorted to every other patch point!!!
Dumb story I know! I actually figured this out whilst I was building a low pass filter pedal in an aluminium case, I included some banana points for external modulation and realised that there was a good reason some plastic sleeves were included – d’oh!
So I ordered a complete set of Emerson banana sockets from Thonk which have plastic shafts, and after an hour or two of soldering, surprise surprise it sprang to life!!
Once I got past this massive hurdle of fail, I found a couple of other problems. The mix controls were all wired backwards, the filters didn’t work at all and the sample and hold didn’t do a great job of the “hold” part…
Mix pots were an easy fix of course – the filters I battled with for a while, but it turned out they just needed calibrating and the trimmers actually had a pretty specific sweet spot. The sample and hold took a little longer to diagnose, but I was able to get to the bottom of the problem by sifting through the information on Ray Wilson’s site and probing the circuit. It turned out that the 2n5457 N-channel JFET was dead, and this was supposed to gate current allowing the weird polystyrene cap to charge/discharge – because it was broken it stayed in a locked “on” state, which explains why the “hold” didn’t really work as intended. I am guessing it broke when the power supply exploded… woops… Once I replaced that tranny, the sample and hold worked beautifully 😀
So here it is, all finished, all working… missing a few screws… (ahem) but finished!