While I was working on the trépied wires Neve woke up and kept sticking her cute wet nose perilously close to the jaws of my pliers. At one point she upended the tray and I had to go searching for wires on the floor. The video I showed was of Neve being rewarded for not being a complete nuisance.
Minutes after I recorded that she went rolling on the floor pouncing on her squeaky toy and barking. My wife took her outside for a little exercise and I constructed the Y cable in the relative peace and quiet that followed.
Today was a relatively quiet day. My caring duties were light.
The rings arrived, and they look very good. The inner edge of the self-adhesive paper ring fits underneath the metal lip of the jack barrel, and the outer edge overhangs by about 1mm. I coloured one of the rings using a washable marker and left the other untouched. This is a very low tech, cheap and adaptable way to colour code banana jacks.
There are 2000 paper rings in all, working out to a fraction of a penny per ring. I’m old enough to remember when rings like these were gummed and you had to lick them before applying them.
In the Potting electronics thread you can follow my progress creating mignons à vitrine, but that is mostly waiting and occasionally checking the results of the curing. I think I’m nearing the point where I can reliably encapsulate whole batches of small electronic circuits in epoxy resin, so now my attention turns to soldering together such circuits.
As discussed here in the Potting electronics thread, I mistakenly ordered some surface mount transistors a week or two ago, and as their really small size matches my needs well I think I’ll try incorporating them into a suitable circuit.
The initial mignon I have in mind is one from Bela Salt. I’m posting here a blowup of the trigger input of that schematic under the GNU Public Licence version 2. Copyright belongs to Rebel Technology dated 2018 and a full copy can be downloaded and used under the same licence from github at:
The trigger output is also shown but I intend to work on the input first, then move to the output afterwards in a separate mignon. My main concern will be to fit the circuit into the space, which is about 20ml. Maybe I should also consider surface mount resistors. A ceramic 100nF capacitor will also be placed across the +3.3V and 0V power rails within the mignon.
I’ll test it in chapeau form, which is a small box with a mini breadboard on top. This will give me a chance to measure the quiescent current and heat dissipation, which are important considerations for this construction technique.
Here’s a picture of my chapeau. It’s a small box (boîte) with a separate lid. A small breadboard on the lid makes it easy to configure and test.
So here I am this afternoon with some spare time. I’m ready to try soldering the circuit together, but first I need to learn how to identify the legs of the transistor.
The BC846 datasheet tells me that the base and the emitter are on the same side of the transistor, while the collector is alone on the far side. But it’s difficult to work out from that which is the base and which is the emitter. Supposedly the base is on the left hand side, but which way am I supposed to be looking at the transistor?
I could guess and say that as it’s a surface mount component the view is from above the transistor which is standing on its pads (as in the first photograph below) but I think it’s worth asking around. The second photograph shows the transistor lying on its back.
That was my instinct. Meanwhile this is the first result of my soldering adventures. The collector pad is soldered to a 10K resistor, labelled R906 in the Salt schematic I posted above.
Off to a serious start, I see! Glad to hear you got the free time this afternoon. Your suggestion that you might try to use the surface mount resistors too kind of excited my imagination though. Using all surface mount components dead-bug style would probably be it’s own art form!
The attraction is of course the space savings. If you have only 20ml to pack your circuitry in, the smaller the parts you use the more wonderfulness you can pack in. As an example, the original Salt provides 4 CV out using a single TL074, 16 resistors and four 2.7nF capacitors. I might just have a chance of fitting half of that circuit into the 20ml vitrine process I’m currently developing, but only with surface mount components. More likely I’ll be forced to use mignons à boîte with through-hole components until I perfect a process to produce vitrines larger than 20ml.
I had forgotten the output lead, and also needed to prove to myself that this wasn’t a fluke. Here’s another 846 with a 10K resistor and an output lead. I soldered the output lead first, then wrestled with the transistor.
I need to work on my technique. Also a proper thermostatically controlled iron and a finer tip might help. I’m currently using an Antex C15, basically the same model of soldering iron I used as a boy.
In case anybody wonders why I would put myself to so much trouble, here’s a picture of the same soldering job alongside some through-hole equivalents of the same transistor.
I still use the Antex I got for my 10th birthday in my guitar kit along with the stand I built that weekend. If the soldering is getting fiddly can I suggest solder paste and a hot air reflow station.
I’m considering it. I’m definitely going to get some surface mount passive components and I have some ideas about how to solder surface mount in a point to point fashion. Using such tiny components would greatly expand the capacity of my tiny 20ml vitrines, subject to adequate heat dissipation.
Edit: I looked at the size of the transistors I’ve been soldering and matched them in size to the 1206 surface mount resistor package. I’ve ordered a whole kit of them for a few pounds. Time to start joining SMDs together in crazy tree-like arrangements, or making them look like balloon animals or skeletons or something.
Most resins can take the heat (just avoid heating gas bubbles which could crack your work - sorry, I swear if I mention bubbles once more…)
Hookup wire (the really thin glazed stuff on a pencil spool is also an option then you can use any material you like to support the circuit form.
The resin may take the heat, but thermal runaway may make the semiconducting elements misbehave when they warm up. I may need to incorporate heat sinks or convection vanes that extend outside the surface of the resin and keep the electronics cool enough. I’ll get to that when the time comes.
I’ve ordered some solder paste and I live in hope that the heat gun I already have can work well with it. If it doesn’t I may have to get a rework station. I’ll start developing the sapin de Noël circuit construction method once I have the right gear lined up.
The solder paste did arrive but I wasn’t kidding when I said my caring duties would take up a lot of time this month. So I’ve been thinking about power distribution because thinking is all I can do right now.
As I’ve discussed earlier in this thread, I don’t like the Eurorack connectors and ribbon cables. I want to use fairly hefty power cables, and for convenience I’m probably going to use the fairly standard low power three core braided cables used on the 240VAC circuit in the UK. My power supplies are likely to be bog standard Frequency Central convertors fed by 12VAC wall warts, exactly like the power systems in common use with Kosmo rigs. So the idea is that these three-core cables will carry -12V, 0V and +12V from the convertors to the modules.
There may also be a use for +5V and +3.3V distribution but that’s only pencilled in for now. I’ll assume at first that such requirements are best catered for by circuitry at the end of the +12V line.
Because my system is going to be a hybrid with a lot of software running on small computers like ESP32 and Raspberry Pi and Beaglebones and almost certainly some Arduinos and similar stuff, I’m probably going to have to take special steps to keep RF interference out of my audio path. That’s an entirely separate barrel of monkeys and I haven’t even started to study the needs there. Suffice to say, whatever I do by way of power distribution will have to be easily reconfigurable. I haven’t entirely ruled out isolating the computer power systems rigorously from the audio gear. Maybe I’ll pass my audio through an optical link, or use balanced cables in certain places. I’ll need to design my power distribution system so that it’s easy to debug.
I’m still looking around online to find a suitable distribution design. I think it should probably support triple bus bars running the length of my bookcase shelves, with clip-on or screw-fix connectors attachable at any point. How practical that would be I’ll have to find out. I want to use inexpensive off the peg hardware.
And here’s a brief glimpse of my current concept in development. My busbars are made up of female headers whose legs will be linked by soldering them to a length of uninsulated copper wire (not shown here.) They are to be spaced 5mm apart so that a 3-pin screw terminal can be plugged in at any point. I have a batch of 5cm lengths of single row female headers which I’ll be using to test this.
Another photo showing a second screw terminal and also a length of male header. The spacing is such that every second pin of the male header finds a socket. I’m now thinking of possible uses for the troughs in between. 5V and 3.3V perhaps? I may be using those voltage levels quite a lot.
In this photo I haven’t trimmed the male header to length as the intention is simply to demonstrate the general idea.
Edit: on second thoughts, I think a separate bus would be better for the digital supply of 5V and 3.3V.
My main concern here is now I have created a symmetrical connector, opening the door to reverse polarity catastrophe. Worse, if I used the same design for both synth voltages and digital voltages on separate buses, I would risk frying some quite expensive digital gear.
