Kat 🇨🇦
@northernpaws.bsky.social
She/They | 🏳️⚧️ Software Engineer | Building various audio and music devices
Website: https://northernpaws.io
Patreon: https://www.patreon.com/c/Northernpaws
GitHub: https://github.com/northernpaws
come see me bap at projects with varying success
Website: https://northernpaws.io
Patreon: https://www.patreon.com/c/Northernpaws
GitHub: https://github.com/northernpaws
come see me bap at projects with varying success
Prototyping a front display and speaker mount - didn’t leave myself a whole lot of room for solid connection points to the board, so having to use some creative geometry and a layered sandwich design to make a case that fits around all the components
November 7, 2025 at 4:14 PM
Prototyping a front display and speaker mount - didn’t leave myself a whole lot of room for solid connection points to the board, so having to use some creative geometry and a layered sandwich design to make a case that fits around all the components
Finally have some LCD drawing working! The Rust embedded-graphics crate made it surprisingly painless to start drawing graphics primitives, only needing to implement one trait for my display framebuffer.
November 7, 2025 at 2:41 PM
Finally have some LCD drawing working! The Rust embedded-graphics crate made it surprisingly painless to start drawing graphics primitives, only needing to implement one trait for my display framebuffer.
Progress! After some confusing memory address alignment problems caused by DMA, I’ve got the LCD running at around 120fps. Some slight tearing occasionally, but I think that can be mitigated by syncing the start of the DMA transfers based on the current scan line.
November 5, 2025 at 1:40 PM
Progress! After some confusing memory address alignment problems caused by DMA, I’ve got the LCD running at around 120fps. Some slight tearing occasionally, but I think that can be mitigated by syncing the start of the DMA transfers based on the current scan line.
The current scope for the project is minimum 8 tracks, but ideally 16 mixed audio and step sequencer tracks per pattern.
A project will consist of a max of 128 patterns, with only a couple loaded into RAM at one time (currently playing + next queued).
A project will consist of a max of 128 patterns, with only a couple loaded into RAM at one time (currently playing + next queued).
Because of the complex scope I have for the audio processing side, offloading as much processing as I can peripherals like the FMC (Flexible Memory Controller) and DMA the start is a priority to keep as many cycles as possible free for audio track and instrument processing.
November 3, 2025 at 4:53 AM
The current scope for the project is minimum 8 tracks, but ideally 16 mixed audio and step sequencer tracks per pattern.
A project will consist of a max of 128 patterns, with only a couple loaded into RAM at one time (currently playing + next queued).
A project will consist of a max of 128 patterns, with only a couple loaded into RAM at one time (currently playing + next queued).
After a couple days of work, I’ve to write a custom FMC-based display driver - using an SRAM configuration to talk to the display’s 8080 parallel interface.
Now I can start setting it up to use DMA to write framebuffers to the display RAM without using CPU cycles :3
Now I can start setting it up to use DMA to write framebuffers to the display RAM without using CPU cycles :3
November 3, 2025 at 4:44 AM
After a couple days of work, I’ve to write a custom FMC-based display driver - using an SRAM configuration to talk to the display’s 8080 parallel interface.
Now I can start setting it up to use DMA to write framebuffers to the display RAM without using CPU cycles :3
Now I can start setting it up to use DMA to write framebuffers to the display RAM without using CPU cycles :3
Photo of the bodge job I also had to do on the LCD connector pins to fix the mode select pins I accidentally inverted. I recently bought a small grinding pen which did an amazing job taking off a small amount of solder mask and exposing the via to patch the wire to.
November 2, 2025 at 6:52 PM
Photo of the bodge job I also had to do on the LCD connector pins to fix the mode select pins I accidentally inverted. I recently bought a small grinding pen which did an amazing job taking off a small amount of solder mask and exposing the via to patch the wire to.
Finally got the LCD working consistently! Turned out that there was a significant amount of flux in the FPC port I didn’t initially see that was interfering. Now to get it driven by the FMC’s parallel interface instead of bit-banging, and running at a decent clock speed.
November 2, 2025 at 6:44 PM
Finally got the LCD working consistently! Turned out that there was a significant amount of flux in the FPC port I didn’t initially see that was interfering. Now to get it driven by the FMC’s parallel interface instead of bit-banging, and running at a decent clock speed.
Dealing with some odd LCD behaviour... after fixing a couple swapped pins, still couldn't get an image to appear on screen but could see it was receiving commands (sleep off, tearing output enable, etc). Cleaned the port and FPC with some IPA and suddenly I consistently got a white image. 1/2
November 1, 2025 at 1:41 AM
Dealing with some odd LCD behaviour... after fixing a couple swapped pins, still couldn't get an image to appear on screen but could see it was receiving commands (sleep off, tearing output enable, etc). Cleaned the port and FPC with some IPA and suddenly I consistently got a white image. 1/2
Added some boot status LEDs using the keypad matrix - I have it programmed currently so it waits for a successful SD card initialisation. But the few times I’ve forgotten to put the card in it was really confusing why it looked like the board didn’t work. Now it’ll show a red LED for the SD card!
October 30, 2025 at 8:42 AM
Added some boot status LEDs using the keypad matrix - I have it programmed currently so it waits for a successful SD card initialisation. But the few times I’ve forgotten to put the card in it was really confusing why it looked like the board didn’t work. Now it’ll show a red LED for the SD card!
Got all the lights onto the board last night! Realized I inverted the pin 1 marker on my footprint accidentally so the LEDs are all rotated 180 degrees from the marker. It was rather confusing soldering 1 LED on, and then finding a short on the board 😅
October 30, 2025 at 7:44 AM
Got all the lights onto the board last night! Realized I inverted the pin 1 marker on my footprint accidentally so the LEDs are all rotated 180 degrees from the marker. It was rather confusing soldering 1 LED on, and then finding a short on the board 😅
Had a very confusing 2-3 hours where the TCA8418 I’m using as a button matrix controller appeared to be filling its FIFO with nonsense events.. turns out it was an issue with memory corruption on my I2C input and output buffers causing bit 0 to always be a 1 🫠
October 29, 2025 at 10:21 AM
Had a very confusing 2-3 hours where the TCA8418 I’m using as a button matrix controller appeared to be filling its FIFO with nonsense events.. turns out it was an issue with memory corruption on my I2C input and output buffers causing bit 0 to always be a 1 🫠
Well.. I didn't expected implementing USB MIDI to be as unremarkable as it was.. only took a couple of hours to get it appearing as a MIDI device in Ableton and decoding messages 😄
October 27, 2025 at 5:49 AM
Well.. I didn't expected implementing USB MIDI to be as unremarkable as it was.. only took a couple of hours to get it appearing as a MIDI device in Ableton and decoding messages 😄
Finally have the first row prototype board for my pocket music computer in hand! Bring-up is going smoothly so far, SD card, keypad LEDs, power manager, and basic USB comms running in the first 24h 😄
Just waiting impatiently for the RF modules to arrive so I can get the last components on!
Just waiting impatiently for the RF modules to arrive so I can get the last components on!
October 26, 2025 at 12:08 PM
Finally have the first row prototype board for my pocket music computer in hand! Bring-up is going smoothly so far, SD card, keypad LEDs, power manager, and basic USB comms running in the first 24h 😄
Just waiting impatiently for the RF modules to arrive so I can get the last components on!
Just waiting impatiently for the RF modules to arrive so I can get the last components on!
The last year of my efforts, re-learning circuit design for the first time in almost 10 years, learning new (to me) protocols and interfaces, and tons and tons of research has culminated in this board designed to serve as a base for a portable, pocket-sized music sequencer!
October 9, 2025 at 2:30 AM
The last year of my efforts, re-learning circuit design for the first time in almost 10 years, learning new (to me) protocols and interfaces, and tons and tons of research has culminated in this board designed to serve as a base for a portable, pocket-sized music sequencer!
Reposted by Kat 🇨🇦
If it is not clear, if you think resizable logos/graphics in KiCad PCBs would be a useful feature consider upvoting the linked issue! Thanks~~~ ;) #kicad #pcb #electronics #opensource
September 28, 2025 at 11:26 PM
If it is not clear, if you think resizable logos/graphics in KiCad PCBs would be a useful feature consider upvoting the linked issue! Thanks~~~ ;) #kicad #pcb #electronics #opensource
Got a huge amount done on this project the last couple of weeks - finalized the peripherals, the schematic is 90% finished and most of the board is routed! Aiming to have a prototype ordered by the end of the month.
A few things changed for this project after planning this week - I've taken it from just an FM synth to a full mini-groove box with sampling capabilities, which is much more in line with my long-term project planning from the last several years!
Starting to poke at a new electronics project that's more in-line with my current abilities - pocket-sized FM synthesizer with an integrated 16-step sequencer, SD card, USB-C charging (and possible future USB audio interface), on-board speakers, line-out, headphone amp, and possibly Bluetooth audio.
September 8, 2025 at 12:16 PM
Got a huge amount done on this project the last couple of weeks - finalized the peripherals, the schematic is 90% finished and most of the board is routed! Aiming to have a prototype ordered by the end of the month.
Reposted by Kat 🇨🇦
Rust is such a cute language
September 6, 2025 at 11:49 PM
Rust is such a cute language
Working on a small prototype using a lattice of copper traces and some resistive film as an aftertouch/velocity sensor for mechanical key switches. In theory, by using the two halves of the lattice as a variable resistor in a voltage divider, you can read pressure on the pressed switch with an ADC.
August 16, 2025 at 3:28 PM
Working on a small prototype using a lattice of copper traces and some resistive film as an aftertouch/velocity sensor for mechanical key switches. In theory, by using the two halves of the lattice as a variable resistor in a voltage divider, you can read pressure on the pressed switch with an ADC.
Reposted by Kat 🇨🇦
"the AI knows it did something wrong"
Nah man, the AI knows nothing. It has no understanding. No critical thinking skills. It's just giving you a reasonably likely response to the garbage you yell at it. It's a Ouija board with terminal access.
Nah man, the AI knows nothing. It has no understanding. No critical thinking skills. It's just giving you a reasonably likely response to the garbage you yell at it. It's a Ouija board with terminal access.
July 21, 2025 at 1:38 AM
"the AI knows it did something wrong"
Nah man, the AI knows nothing. It has no understanding. No critical thinking skills. It's just giving you a reasonably likely response to the garbage you yell at it. It's a Ouija board with terminal access.
Nah man, the AI knows nothing. It has no understanding. No critical thinking skills. It's just giving you a reasonably likely response to the garbage you yell at it. It's a Ouija board with terminal access.
Spent the last 2 days trying to get the Barebox bootloader running under PTXDist for x86. Every time I tried to image the bootloader would freeze with a blank screen. Still not entirely sure what fixed it, but overwriting PTXDist’s barebox config with the config from Barbeox’s source fixed it!
July 19, 2025 at 10:41 PM
Spent the last 2 days trying to get the Barebox bootloader running under PTXDist for x86. Every time I tried to image the bootloader would freeze with a blank screen. Still not entirely sure what fixed it, but overwriting PTXDist’s barebox config with the config from Barbeox’s source fixed it!
A hour with the scope later and I’ve finally got this ST7789V2 adapter reverse engineered! One step closer to getting my retrofit PCB for this laser tag system done…
July 17, 2025 at 12:06 AM
A hour with the scope later and I’ve finally got this ST7789V2 adapter reverse engineered! One step closer to getting my retrofit PCB for this laser tag system done…
I've been picking up embedded Rust via Embassy for some work projects, which has expanded into writing a crate for talking to Digi XBee 3 devices.. So this week I've been learning how proc macros work so that I don't need to hand-write parameter encoding for all 178 commands 😄
![use xbee3_rs_macros::command;
#[command]
pub struct ACKFailures(pub u16);](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:og5c7wdglbue7abgtzjuvr4x/bafkreidgbhh55pqt7r62t3f4lvehj5wsz2uwl7tqbfihbvwsh7kpnv55mm@jpeg)
![use xbee3_rs_macros::command;
pub struct ACKFailures(pub u16);
use crate::at::commands::Command;
#[automatically_derived]
impl crate::at::commands::Command for ACKFailures {
fn identifier(&self) -> crate::at::commands::Identifier {
crate::at::commands::Identifier::ACKFailures
}
}
#[automatically_derived]
impl From<ACKFailures> for crate::at::Command<2usize> {
fn from(cmd: ACKFailures) -> crate::at::Command<2usize> {
crate::at::Command {
identifier: cmd.identifier(),
payload: Some(cmd.0.to_be_bytes()),
carriage_returns: cmd.carriage_returns(),
}
}
}](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:og5c7wdglbue7abgtzjuvr4x/bafkreibzahfiw5oqfs63u2mhywee5m4ck4zvbzjxl3v2efyqremsrgkijq@jpeg)
July 15, 2025 at 10:19 PM
I've been picking up embedded Rust via Embassy for some work projects, which has expanded into writing a crate for talking to Digi XBee 3 devices.. So this week I've been learning how proc macros work so that I don't need to hand-write parameter encoding for all 178 commands 😄
Hand-soldered one of my smallest connectors yet for this BuyDisplay LCD dev board! Connector was slightly crooked and had a couple bridges, so I pulled out an old scientific microscope we had around and managed to fix the bridges. Really reinforced that I need to buy a decent binocular microscope!
May 22, 2025 at 12:57 AM
Hand-soldered one of my smallest connectors yet for this BuyDisplay LCD dev board! Connector was slightly crooked and had a couple bridges, so I pulled out an old scientific microscope we had around and managed to fix the bridges. Really reinforced that I need to buy a decent binocular microscope!
Reposted by Kat 🇨🇦
Rohde & Schwarz AMIQ Modulation Generator - Teardown and Analog Deep Dive
tomverbeure.github.io/2025/04/26/R...
tomverbeure.github.io/2025/04/26/R...
April 26, 2025 at 9:38 PM
Rohde & Schwarz AMIQ Modulation Generator - Teardown and Analog Deep Dive
tomverbeure.github.io/2025/04/26/R...
tomverbeure.github.io/2025/04/26/R...
Got my test PCBs for my capacitive sensing rotary encoders!
Only made one small mistake and selected a very small header footprint instead of standard breadboard size… gonna have to find some small wires to get them connected 😄
Only made one small mistake and selected a very small header footprint instead of standard breadboard size… gonna have to find some small wires to get them connected 😄
April 25, 2025 at 12:20 AM
Got my test PCBs for my capacitive sensing rotary encoders!
Only made one small mistake and selected a very small header footprint instead of standard breadboard size… gonna have to find some small wires to get them connected 😄
Only made one small mistake and selected a very small header footprint instead of standard breadboard size… gonna have to find some small wires to get them connected 😄