Sound generators to PCBs


Next phase in the analog drums project would be to design printed circuit boards for the sound generators. This will again need some planning.

I’m using the free version of Eagle as my PCB CAD, and it has some limitations. The biggest of those are: only two layers and board maximum size 8*10 cm. Two layers is almost never a problem in my projects. Some people consider the board size as a big limitation, but it is not always that straightforward. If you find yourself in a need of bigger board, ask yourself: did my hardware modularization go as it should have? This limitation forces you to think your hardware as reasonable sized modules, and it is never a bad thing. Of course, there are situations when the PCB must be just certain size, but most of the time, this limitation is not a big deal.

Considering that the generators, mixers and others consist of 19 op amps, 13 transistors and something like 300 passive components, they surely won’t fit on an 8*10 cm board. So modularization is needed.

I normally use SMD components for passives. ICs and transistors are SMD or through-hole, according to what I happen to have or get. I have stock of TL072’s in DIP-8, so they will be used. BC5xx-transistors are in TO92.

So, after some thinking with those preconditions, I came up with these principles:

  • There will be two or three boards containing the 10 sound generators
  • There will be one board containing the common parts: main mixer, accent mixer, accenter circuit and noise generator.
  • In order to simplify the wiring, each sound generator board will have an accent submixer for mixing that board’s generators together.
  • In order to minimize disturbances, there will be a local supply voltage regulator on each board.

And after some more thinking and test layouts, I came up with this 4 board setup:

  • DR-55 soundgenerator board:
    • Boss bass drum
    • Boss snare
    • Boss hi-hat
    • Accent submixer
    • 12 V supply regulator
  • KR-55 + ASR soundgenerator board:
    • Korg bass drum
    • Korg snare
    • Andromeda Space Rockers hi-hat
    • Accent submixer
    • 12 V supply regulator
  • Percussion soundgenerator board:
    • Korg rimshot
    • Korg claves
    • ASR hi tom
    • ASR lo tom
    • Accent submixer
    • 12 V supply regulator
  • Main mixer board:
    • Main mixer
    • Accent mixer
    • Accenter circuit
    • Noise generator
    • 12 V supply regulator

Four boards with this amount of components means that the boards will not be too tightly populated. It will be possible to design them single-sided. I could get them manufactured in a cheap Chinese PCB factory, but I think this time I will use traditional home etching. There are no small traces, all ICs are DIP-packages, and I’m trying to standardize my passive SMDs to 0805 size. So, nothing too fine for home etching. Boards may also need some extra rework after building, you never know about analog electronics. Soldering extra capacitors, rerouting wires etc. might be needed. All this will be easier without solder mask and other pro things that a PCB factory will put there.

This “factory or home etching” decision must be done before designing the layout because many things must be done differently in the layout, depending on how the board will be manufactured.

Sitting several hours drawing schematics, placing components and routing traces with Eagle:


 This resulted in a bunch of ready designed boards.

After some more hours of exposing with UV, etching and drilling the boards were ready to be populated with components and soldered.

An finally, here are two of the resulting boards, DR-55 and percussion boards. The same boards from component side are in the beginning of this post. The empty IC sockets will be populated with TL072’s.

 

The boards are meant to be stacked on each other using spacers, like this. This "generator stack" is quite handy and is very space-effective inside the enclosure. Even if my Eagle had not the size limitation, I probably would have used this structure anyway. Again, an example that the seemingly serious limitation is in practice not that serious.


In the Github repository, there is also a picture showing principles of the connections between the boards. This is basically the same picture as the high level diagram in the previous post, only with more details. 

Each sound generator board has only one accent output because all board's generator outputs are mixed together already on the board. So the accent mixer on the main mixer board has only 3 inputs. This reduces wiring between the boards.

The connection diagram also shows how the dual-sound channels (channels with two selectable sounds) are connected. There are selection switches both on the trigger side and the output side. While this may look like a bit overkill, this has to be done like that. If there were selection switches only on the output side, both sounds would be triggered always, and the sound of the non-selected generator would be audible through the accent mixer. On the other hand, if there were selection switches only on the trigger side, things on the output side would get more complex. We would have 10 generator outputs, the dual-channel level control pot would have to be tandem and main mixer would have 10 inputs. While doable, unnecessary complexity anyway.

The trigger pulse polarities are one thing needing some consideration. Boss, Korg and Eric Archer have had problems in deciding, should the trigger polarity be active high or active low. Boss and Archer use active-high trigger and Korg uses active-low. This means that on some of the dual-sound channels (bass drum and snare) the sequencer must produce both versions of the trigger pulse. Not a big deal, but must be taken into account while designing the sequencer.

 

Now that the sound generators are finally ready, it is time to start designing the sequencer part of the machine. This will happen in the next post.


Comments

Post a Comment

Popular posts from this blog

Secrets of the YM2154

Image
  After a break, I returned to the PSR-70 reverse engineering project: t he YM2154, a.k.a. RYP4 is still an unknown territory. The time has come to reveal its secrets. R YP4 is used a bit more widely than OPQ. Besides PSR-60/70 it is used at least in PSR-80 and RX- 11/15/21 drum machines. This time I did remember to check for service manuals, and found one for RX-11: https://elektrotanya.com/yamaha_rx11_sm.pdf/download.html . This contains a full schematic and other useful information.   T racing register writes Because I had good results with OPQ by tracing what the PSR-70 firmware is writing to the chip registers, this would be a good starting point also for RYP4. The chip has 128 registers which reside in Z80 I/O address space at addresses 80H...FFH. Looking at the disassembly listing reveals quickly that there is no single function which would take care of all the register writes, like OPQ had. There are many IN and OUT instructions accessing the RYP4 register
Read more

Reversing the PSR-70 hardware

Image
  First step in the actual reverse engineering is to start studying the PSR-70 hardware. This is quite service-friendly (meaning hacker-friendly) device: all circuit boards are single-sided and copper side silkscreen has IC numbers and jumper wire positions marked. There are also many useful looking signal names marked, like “D7”, “A12”, “IOWR” etc. It would be possible to draw the whole schematic from one high-quality photo of the circuit board. My intention is not to draw the whole schematic, but at least the address maps should be figured out. It will make the interpretation of disassembly much easier. All address decoding is done with 7400-series discrete logic, no PAL-circuits or other nuisances used, this is very good.   Block diagram Most of the electronics is on the large main board . The other cards are: Power + audio board contains the audio amplifiers, some kind of analog chorus-effect based on a BBD-circuit, power amplifiers feeding the speakers, and voltage
Read more

Digging into YM3806

Image
In the previous post I got the PSR-70 original firmware modified so that it dumps to serial port all register writes done to OPQ chip (YM3806). Now I’m getting the data I wanted and plenty of it! A simple keyboard key-down or key-up produce only under 20 register writes but changing sound from the front panel results in about 300 writes. Why 300, there are only 256 registers? Yes, but there are many registers which are written several times. T he raw dumps from the modified firmware are quite hard to read, they just list the register address and written value in hex in the order they happened: 1F=20 5E=1F 56=1F 4E=20 46=20 5E=81 56=81 4E=81 46=81 9E=1F … To make some sense out of this I wrote a small python script to analyze the dump. It lists all registers in numerical order and all writes done to that register. In the script output the first column is OPQ register number, second column is the number of writes to that register and then a list of the written values.
Read more