stuartl/atinysynth
1
0
Fork 0
mirror of https://github.com/sjlongland/atinysynth.git synced 2025-09-13 18:13:16 +10:00
Code Issues Releases Wiki Activity
60 Commits 1 Branch 0 Tags 235 KiB
98df61229d
Commit Graph

60 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Stuart Longland
98df61229d
attiny861 main: Use switch instead of if. 
Years ago, and on MSP430 not AVR, I found that gcc produced tighter code
using `if` instead of `switch`.  Thus, I found myself always using `if`
and never `switch` if I needed branching, since smaller code is faster
code, and we needed all the speed we could get.

Turns out, AVR does not have this problem, and we get better performance
using `switch`.
 2017-07-15 15:59:49 +10:00
Stuart Longland
b6e678f045
attiny861 poly_cfg: Drop sample rate to 8kHz 
With the extra processing needed to get bleed-free LED control, we can't
quite make 16kHz sample rates.
 2017-07-15 15:55:06 +10:00
Stuart Longland
0ef982d463
attiny861 main: Fix adjacent light channel bleed. 
Move to a 3-state machine for PWMing lights so that we don't get bleed
from one channel into the next due to propagation delay in the 74HC574.
 2017-07-15 15:41:20 +10:00
Stuart Longland
7bb914755c
attiny861 main: Hard-code buttons to channels. 
Rather than hunting for a free channel, just hard-code the button to the
channel.  This gets polyphonics sort-of working: both notes are
out-of-tune which suggests the CPU isn't keeping up.
 2017-07-15 15:22:21 +10:00
Stuart Longland
7039dd1307
adsr: Add state helper functions 
Add helper functions that report back whether the ADSR generator is in a
given state.  Namely, if it is "waiting" for a trigger, or it is idle.
 2017-07-15 15:19:54 +10:00
Stuart Longland
aeefd70fe3
attiny861 port: Fix annoying pop on each note. 
It seems when the NJM2113 is first powered up, it emits a pop noise.
Not much I can do about that.  The thought was I should power down the
amplifier when it is not in use to save power.

The outcome is on each note, it would power the amp up, and it would
pop.

Solution: create a 1msec timer for general timing events, and use that
to count down one minute before powering off the amplifier.  The toy
will pop on first power on, then if left alone for a minute, will pop
again on the first note, but will not pop so long as there's only short
gaps between notes.
 2017-06-11 16:09:42 +10:00
Stuart Longland
1cc66ace92
README.md: Add some more detailed notes. 
This documents the theory of operation and how the ports work.
 2017-06-02 20:36:07 +10:00
Stuart Longland
f982e0c31f
attiny861: Bump amplitude up. 
Since the number of notes possible with this device is small… in fact,
it seems to only want to do monophonics, not sure why, we can bump the
amplitude up a bit more without risk of saturation.
 2017-05-28 11:08:18 +10:00
Stuart Longland
cf587e3fd8
attiny861 port: Clear button hit/release events at power-on. 
This prevents all the notes being played all at once on power-on.
 2017-05-28 11:01:32 +10:00
Stuart Longland
e75d75d20b
attiny861: Enable pull resistors when reading. 
Set PORTA to all ones to enable the pull resistors on all pins.
 2017-05-28 10:43:27 +10:00
Stuart Longland
762b07177b
adsr: Fix "infinite" delay/sustain. 
Seems `~value` doesn't quite have the intended effect.  The intent was
that, being the 1s compliment of `value`, the expression would return
zero if `value` was all ones, and non-zero.

This doesn't quite work the way I'd like it to.  A comparison should be
the same number of CPU cycles, and is more readable, so we'll do that.
 2017-05-28 10:33:52 +10:00
Stuart Longland
0bed82802b
attiny861 port: Bump sample rate 
Seems we can handle 16kHz sampling… but not 32kHz.
 2017-05-28 10:25:21 +10:00
Stuart Longland
75fe949f27
attiny861 port: Semi-working version. 
This implements detection of the pressed buttons, with software
debouncing, and plays a note indefinitely until the button is released.
The corresponding LED's amplitude is modulated by the ADSR envelope.
 2017-05-28 10:19:05 +10:00
Stuart Longland
d9bfcb7991
adsr: Support "infinite" delay and sustain. 
This is useful if you don't know how long a given note will be held for…
you can just trigger the note upon pressing the button and have it
progress to the "sustain" state, at which point it'll hold the note.

You can then call `adsr_continue` to "release" that note when you detect
the button release and have it decay.
 2017-05-27 15:19:40 +10:00
Stuart Longland
f2fac87ba3
main: timer: Invert button states 
There's a pull-up with the button asserting a logic low, so invert the
inputs.
 2017-05-27 15:01:41 +10:00
Stuart Longland
f8346cdea9
Add ATTiny861 port. 
The ATTiny85, as good as it is, has too few pins to really do a lot.
You can buy I²C GPIO expanders, but many cost more than programming an
ATTiny24A (which has more brains).

The nearest equivalent is the ATTiny861, it is basically the same core
as the ATTiny85, features the same PLL for high-speed PWM, but comes in
a 20-pin package.  They cost ~AU$3 in individual quantities.  (An
ATTiny85 costs AU$1.74, an ATTiny24A costs AU$1.57; thus you save about
20c and a lot of interfacing effort in going to the '861.)

Thus it can interface to 8 individual switches with ease.

The circuit here uses a 74374 D-latch to drive up to 8 LEDs with PWM and
two 4066s to isolate the 8 inputs.  Idea being, when the GPIO_EN signal
is high, the 4066s are turned on and port A sees the 8 GPIO lines on the
other side of the 4066s.

4066s were used because I have 4066s up the wazoo… bought a box of
random ICs off eBay many years ago and it came with 5 (!) tubes of
MM74HC4066s with 25 ICs each (amongst other parts, some hard-to-find).
The bonus being these can be ADC inputs too if desired, allowing sensing
of piezo sensors.

When we want to switch which LEDs are turned on, we bring GPIO_EN low,
switch port A's pins to outputs, assert the desired LEDs, then bring
GPIO_EN high again.  The 74374 latches those pins, and we are free to
put port A back to being inputs.

This happens with each sample from the synthesizer, alternating between
input and output.  Thus the effective rate seen on the LEDs and inputs
is half the sample rate.

A spare GPIO is available for turning on and off an amplifier (I use the
NJR NJM2113D) to save power.

The 74374's nOE pin is connected to the PWM output for the lights, thus
using pull resistors, one is able to use the one PWM channel for all 8
lights.  The lights are turned on in round-robin fashion, so effective
duty cycle is ⅛ and the refresh rate is ~500Hz.
 2017-05-20 20:48:19 +10:00
Stuart Longland
6ddaad18f2
attiny85: Add ADC-based keyboard support. 
This is a work-in-progress, but we're now able to play some notes by
pressing keys.
 2017-05-01 17:38:17 +10:00
Stuart Longland
3ff3e349f5
attiny85 Makefile: Add main.o and adckbd.o to object list. 2017-05-01 17:37:48 +10:00
Stuart Longland
8333f32d26
pc Makefile: Add main.o to port object list for PC 2017-05-01 17:37:22 +10:00
Stuart Longland
01a1bc8b70
Makefile: Make object list configurable. 
Allow ports to insert their own object files to the list.
 2017-05-01 17:36:56 +10:00
Stuart Longland
aa3771dc78
adckbd: Add first stab at ADC-based keyboard. 
This works by relying on a resistor ladder that is part-shorted-out by a
button to Vcc.  The ADC detects the button pressed from the voltage seen
on the ADC.

There's some work to be done on this yet.
 2017-05-01 17:35:19 +10:00
Stuart Longland
44e2d937d6
gensine: Add in sine wave generator. 
I left this out because I thought the idea of modulus and division on a
MCU that lacks a hardware multiplier would be too much for it… and
indeed it was too much with my other project.

Thinking about it this afternoon, I had an idea.  If I have 2^N samples,
then the modulus can be optimised to:

```
	mod = sample & ((2**n)-1)
```

and the segment can be figured out by:

```
	segment = sample >> n
```

The segment is two bits.  A function that returns the scaled sine value
for a given scaled angle can be given as:

```
	int8_t fp_sine(uint8_t angle) {
		uint8_t segment = (angle >> POLY_SINE_SZ_BITS) & 3;
		uint8_t offset = angle & ((1 << POLY_SINE_SZ_BITS)-1);

		switch (segment) {
			case 0:
				return _poly_sine[offset];
			case 1:
				return _poly_sine[
					POLY_SINE_SZ - offset];
			case 2:
				return -_poly_sine[offset];
			case 3:
				return -_poly_sine[
					POLY_SINE_SZ - offset];
		}
	}
```

… or something like that.  If `POLY_SINE_SZ_BITS=6`, then `angle=255`
represents 360°.
 2017-04-25 17:43:51 +10:00
Stuart Longland
2fe3ea766c
fifo: Drop unused library 2017-04-09 10:53:57 +10:00
Stuart Longland
17599177cb
ports/attiny85: Looping sound effects. 
It sounds like the crappiest doorbell you've ever heard, but we now get
a crude "ding dong" effect with this code, on a loop.
 2017-04-09 10:17:49 +10:00
Stuart Longland
7214ae7d78
synth: Make enable/mute bits volatile. 
Otherwise the optimiser doesn't bother to re-check them.
 2017-04-09 10:17:09 +10:00
Stuart Longland
fb59c52b71
ports/pc: Start playback when enable ≠ 0. 
After configuring the voices, we start playback until all channels stop.
 2017-04-09 10:13:29 +10:00
Stuart Longland
ba7d0598c9
synth: Turn off channels that are "done". 2017-04-09 10:07:27 +10:00
Stuart Longland
dd483630d7
voice: Add test for voice channel done. 2017-04-09 10:07:10 +10:00
Stuart Longland
57769d477a
adsr: Add debugging 2017-04-09 10:06:38 +10:00
Stuart Longland
5855eb3da0
adsr: Add test for "ADSR done" event. 2017-04-09 10:06:11 +10:00
Stuart Longland
093b6dd888
ports/attiny85: Set sample rate to 8kHz 2017-04-09 09:17:28 +10:00
Stuart Longland
3f7c6353c1
ports/attiny85: Remove debugging #if blocks 2017-04-09 09:17:04 +10:00
Stuart Longland
b1ddd0c099
ports/attiny85: Define program target 2017-04-09 09:16:50 +10:00
Stuart Longland
38bf5a4261
Makefile: Remove .dep output definitions. 
This confuses `make` and it winds up telling `gcc` to write the compiled
binary to the `.dep` output.
 2017-04-09 09:09:53 +10:00
Stuart Longland
9bba7c0c19
ports/attiny85: Work in progress ATTiny85 port 2017-04-09 08:24:14 +10:00
Stuart Longland
301f5d0b8c
fifo: Add FIFO buffer library 2017-04-09 08:23:52 +10:00
Stuart Longland
9f3cd37109
ports/pc: Set default target in Makefile 2017-04-09 08:23:33 +10:00
Stuart Longland
4bad2a4c91
Makefile: Re-work build system 
Running `make PORT=...` should be all that's required to build one of
the ports.
 2017-04-09 08:22:54 +10:00
Stuart Longland
67bf582ee4
adsr: Add functions for configuring the ADSR. 2017-04-09 08:17:03 +10:00
Stuart Longland
286caebd67
adsr: Consider sustain time in initialisation. 2017-04-09 07:45:32 +10:00
Stuart Longland
54f47795ac
waveform: Fix logic errors in triangle synthesis 2017-04-09 07:26:16 +10:00
Stuart Longland
aff9a6c00b
waveform: Flip phase of sawtooth wave. 2017-04-09 07:22:32 +10:00
Stuart Longland
6f7749c0fb
waveform: Increase step to 16-bits. 
Integer overflow was breaking sawtooth output.
 2017-04-09 07:19:04 +10:00
Stuart Longland
2a9c5ac4dd
waveform: Fix square wave generation. 2017-04-08 22:02:02 +10:00
Stuart Longland
06199b7da8
pc/main: Tweak output 2017-04-08 21:53:12 +10:00
Stuart Longland
2236e9eda8
synth: Show channel output in debug 2017-04-08 21:53:00 +10:00
Stuart Longland
d7ae5000ca
adsr: Tweak attack/release envelope computation. 2017-04-08 21:52:27 +10:00
Stuart Longland
2281f8b0e3
voice: Debugging 2017-04-08 21:52:10 +10:00
Stuart Longland
ba94adbfd1
adsr: Fix attack envelope. 2017-04-08 21:41:02 +10:00
Stuart Longland
b77e364f53
adsr: Define "done" state 2017-04-08 21:40:49 +10:00