QuadSID, a week of fail
Recovering from my christmas holidays, I decided to continue this project early this week by doing quick mockups of the switch boards for the display section (LED/button matrix, SID selection, menu navigation). After all, this switch board section was the second most critical block after the display element. The mockups would help me to fine tune the final board placement and I could also use them as drilling templates for the front panel at a later stage. So once again, pieces of empty circuit boards to the rescue! Choosing a 1cm (test) spacing for the LED grid, I drilled some holes to the boards to simulate the buttons and LEDs.
And what a great idea this was! In no time it proved itself a good one as it became instantly apparent that I had mounted the display to the wrong place! There was just no way to fit the entire button/LED grid below the display. Yay \o/ .. Somewhat sucks that I didn’t think of making these mockups at a earlier stage, but anyway good that I did it at this point (before drilling the front plate or other less reversible tasks) . Anyway, after relocating the display mounts (other DOUT-module needed to be moved as well) and cutting the top half of the enclosure to accommodate the changes, I continued shuffling the buttons around in order to end up with the final panel layout. And guess what? It turned out that the switches bought for the project (panel mount type like this one here) are mechanically too big to fit the keyboard frame, even despite the relocated display element left a lot more room to manoeuvre. Hence, a new switch model had to be selected as there was no more room to move the display element. Back to square one it is then, double yay for me. \o/
The only suitable switch I had available was the rather bog-standard Tyco Electronics FSMJ-series tactile microswitch. This type has a lot shorter button travel and shaft, so the whole keyboard frame needed to be moved closer to the surface of the upper half by trimming the mounting towers on the upper half of the enclosure as short as possible (encoder nuts and some cabling still needs to fit in-between). Even with the frame moved closer to the surface, I’m not yet 100% sure that the switches will be usable/tactile enough given their short travel and shaft. I’ll probably need to drill the switch holes on the front panel slightly concave with a flush mount drill bit. Because of the switch change I also had to redesign the LED matrix and make completely new mockup boards.. Sigh. Anyway, I finally completed the new switch boards today for the display section (LEDs still waiting to be installed), so here’s a few pictures to summarize all this rambling:
This board placement approach allows me to skip doing any mechanical drawings on how the the front panel and the keyboard frame should line up, nor do I have to precisely measure the individual board locations either. When assembling, the boards can be simply lined up to the holes on the front panel (the spacing will be the same thanks to my mockups) by moving the boards before the glue used to fasten them to the keyboard frame dries up. Once again, CAD, now wtf is that ;) .. The last image is a side-view of the switch shafts; like said it still remains a mystery whether they’ll be usable enough. Initial test do indicate that the slightly concave area on the panel surface should be enough though.
So, what a week of fail with the project. On the positive side, the front plate will look more clean as the switches will be also hidden under the front plate. Most importantly, the switches still remain “garden variety type” despite the change, so the synth will hopefully remain serviceable to the distant future.