Here are the build notes for my MIDI Matrix Patch Bay PCBs. This post is in two parts (below) covering the two boards required. A further part will consider the mechanical build issues.
Recall that there is a mistake in the first version of the OUTPUT boards. Although in the design I've presented the fixed version, my physical boards are all "V1", so I'll talk around options for what to do about that as I go. I've not had any V2 boards made.
- Part 1: Design.
- Part 2: PCB assembly.
- Part 3: Mechanical design and assembly.
Note: If mounting holes are to be drilled, then it is probably easier to add them prior to soldering components.
Warning! I strongly recommend using old or second hand equipment for your experiments. I am not responsible for any damage to expensive instruments!
OUTPUT: Bill of Materials
- MIDI Matrix Patch Bay OUTPUT PCB - recall this description is for the V1 boards. I haven't; had any V2 boards made.
- 2x 220Ω resistors
- 2x 4-way rotary switches, PCB mounted (shown below).
- 2x 3.5mm, stereo TRS sockets, PCB mounted (see photos for footprint)
- Optional: 2x 180, 5-pin DIN sockets, PCB mounted (see photos for footprint)
- Optional: pin headers.
These are the 4-pole rotary switches I've used. They were described as "4 band switch" online and my specific purchase had the description "Band Switch Rotary Switch Gear Change Switch 2 Pole 4 Position RS101 YT".
The pins are at the standard 2.54mm (0.1") pitch and the two rows are 5mm apart. The common pin for each row is at the opposite ends, as indicated by the "square" pin in the footprint above (labelled 1 and 6) and the yellow circle in the photo.
OUTPUT: Errata and Patching
As described in the design notes, pins 4 and 5 of the MIDI ports are wired the wrong way round. This is really annoying and something I really should have spotted, but rather than waste the boards with a re-spin, here are some options.
- Accept that DIN MIDI won't work and I have TRS "Type B" MIDI sockets on the OUTPUT stages.
- Do a simple patch that just swaps the 5V/TX to the connectors. This can be done with one track cut, re-routing the resistor and a single additional wire, per MIDI port.
- Do a more complete patch that also allows the switched connections from the TRS socket to switch the DIN sockets in and out. This requires four track cuts and four additional wires, per MIDI port.
Both patch options are shown below with track cuts indicted in orange and new wires (and the re-routed resistor) in yellow and blue. The simpler patch (2) on the left; the more complex patch (3) on the right.
OUTPUT: Build Steps
If any patching is to be done on the PCB I recommend doing the track cutting first (see above). I've also drilled out some mounting holes (see next post) which again is probably worth doing prior to the soldering!
Then it can take a typical "low to high" soldering approach. This is the suggested order of assembly:
- The resistors - unless they are being used for patching, in which case these will probably have to be on the underside of the board and should be added at the end.
- TRS sockets.
- Rotary switches.
- Pin headers (if used).
- DIN sockets.
- Patch wires (if patching) and resistors (if being used for patching).
Here are some build photos of my original (non-patched) build.
And here are some photos of a fully patched board.
OUTPUT: Testing
I recommend performing the general tests described here: PCBs.
A multimeter or continuity tester can be used to check the switches are working correctly, connecting pin 4 of the DIN socket and the ring (R) of the TRS socket to each of the four inputs in turn.
Note that input 1 is actually "fully clockwise" for the rotary switch and input 4 is "full anti-clockwise". I'd have probably labelled these up the other way round on the header connectors if I'd realised this was the case.
INPUT: Bill of Materials
- MIDI Matrix Patch Bay INPUT PCB.
- 2x H11L1 optoisolators.
- 4x 74HC14 (HC or HCT variants).
- 2x 1N914 or 1N4148 signal diodes.
- 22x 220Ω resistors.
- 2x 1K resistors.
- 6x 100nF ceramic capacitors.
- 2x 3.5mm, stereo TRS sockets, PCB mounted (see photos for footprint)
- Optional: 2x 180, 5-pin DIN sockets, PCB mounted (see photos for footprint)
- Optional: pin headers.
- Optional 2x 6-pin DIP sockets.
- Optional 4x 14-pin DIP sockets.
For the power circuit:
- 1x 7805 5V regulator (TO-220 package).
- 1x 100nF ceramic capacitor.
- 1x 10uF electrolytic.
- 1x 100uF electrolytic.
- Barrel jack socket (see photos for footprint).
- Optional: pin headers.
INPUT: Build Steps
The first thing to decide is if the additional power PCB is required or not. If not used, then a stable, regulated 5V supply is required from somewhere. If two input stages are used, then a single power supply PCB can supply both.
If not required a sharp knife or hacksaw can be used to separate the two parts as shown below.
Also, if additional mounting holes are to be made (see next post) then that is worth doing prior to soldering too.
Then taking a typical "low to high" soldering approach, this is the suggested order of assembly:
- Resistors and diodes.
- DIP sockets (if used - recommended).
- TRS sockets.
- 7805 (if used).
- Ceramic capacitors.
- Header pins (if used).
- Barrel jack socket.
- Electrolytic capacitors.
- DIN sockets.
Here are some build photos. Note that I didn't think about soldering the 7805 until after the electrolytics in the photos, but I recommend soldering that before them, as described above.
INPUT: Testing
I recommend performing the general tests described here: PCBs.
If using the power circuit, I strongly recommend testing its output before installing the optoisolators and 74HC14s. If not using a switch for power then a jumper needs soldering/adding across the "PWRSW" header pins. Recall that the power circuit requires 7-12V.
Then a single output stage could be used to test each of the 10 THRU ports by either using jumper wires (if pin headers have been used) or the "fishing wire trick" to temporarily connect it to each THRU port in turn.
Recall that the TRS and DIN sockets cannot be used at the same time.
PCB Errata
There are the following issues with these PCBs:
- The TRS sockets were positioned to be panel mounted, but are not panel mount variants.
- Pins 4 and 5 for the DIN sockets on the OUTPUT are backwards (this was fixed in the V2 design).
- Due to the above, the TRS sockets are wired as MIDI "Type B" sockets (again, fixed in the V2 design).
- The labelling for the four inputs on the OUTPUT board is backwards to what might be expected from the rotary switch position (fixed in the V2 design).
- The rotary switches are set too far forward really to allow the TRS sockets and switches to be mounted neatly on a panel.
Enhancements:
- Switched TRS sockets could have been used to disable the DIN sockets when a 3.5mm plug is inserted (included in the V2 design for the OUTPUT board, not for the INPUT board yet).
- Mounting holes could be added, including to support stacking of the OUTPUT PCBs (added in the V2 design for the OUTPUT board, not for the INPUT board).
- I'd also add a 5V/GND connection on the opposite side of the INPUT boards to make linking them up to the OUTPUT board power/ground easier.
As there are a number of mistakes, I've not included these boards in my GitHub repository. But I'm very happy to share the "design" if this sounds like something you're interested in and you'd be happy to fiddle around with as described above!.
Closing Thoughts
Whilst I have a fixed design for the OUTPUT boards, I'm loath to order some new ones. With a bit of relatively simple patching I can get them working for what I need, so I think I'll probably just do that and put up with it.
As I say, if you are interested in doing something with this design too, feel free to ping me an email - diyelectromusic at gmail - and we can chat about it.
The final part in this series will look at the mechanical assembly.
These boards have been manufactured using the Seeed Fusion PCB service, which I am happy to continue to recommend. They have been supported with discount vouchers that I've been sent by Seeed for my previous projects.
Kevin
No comments:
Post a Comment