search

Lefuneste83 / KLOR

KLOR is 36-42 keys column-staggered split keyboard. It supports a per key RGB matrix, encoders, OLED displays, haptic feedback, audio, a Pixart Paw3204 trackball and four different layouts, through brake off parts.
GNU General Public License v3.0
17 stars 2 forks source link

Rev 1.5 #1

Open pmd-coutinho opened 1 week ago

pmd-coutinho commented 1 week ago

hey hey @Lefuneste83 . first of all thank you for this amazing fork. very nice QOL improvements particularly the LED changes xD. I just wanna to ask if you are planning on distributing the gerber for the 1.5 rev any time soon?

if not i will order rev 1.4 pcbs.

thanks in advance

Lefuneste83 commented 1 week ago

Thank you for your feedback. I haven't validated the 1.5 hardware integration yet. It is required as the linking cable is now transferring USB VCC (5v) instead of post regulator VCC (3.3v in most cases) from master MCU so I need to clamp the voltage down to 3.3v in order to protect the RX/TX GPIO rails in case of cable disconnection while the keyboard is powered. I will be using a dual Schottky diode to clamp the voltage on these rails but obviously this must be validated in real life. For the moment I recommend sticking to 1.4 which is good enough for most users. The main drawback of these pre 1.5 versions is the overload on the master MCU voltage regulator when using a 3.3V rated MCU. And also the less than optimal voltage to drive the LEDs strings. Also keep in mind that the 1.5 version will have an integrated haptic buzzer and driver along with several SMD passives which require advanced SMD soldering skills and equipment.

joeblackwaslike commented 3 days ago

@Lefuneste83 reading the changelog the improvements included in 1.5 seem worth waiting for, especially since my previous 5 attempts to build the original 1.3 all failed for different reasons, sometimes the mcu would overheat instantly, sometimes the oled would fail to work, sometimes a short would develop among the mcu pins of both sides, sometimes they would be fine on their own but would stop working the moment they were connected, etc.

It took me 5 attempts to get both sides of the board working w/o overheating, encoders, oled, switches all worked but when I connected them together over trrs, it stopped working at all. I tried 3 different trrs cables and even a trs for good measure. I enabled debug in qmk and connected to the debug monitor over usb and the errors were all related to the split keyboard communication.

I realized later that I had made the mistake of using the mini-e led modules, although the original docs said not to use these, the illustrations showed the mini-e modules (with the 4 leads coming out) instead of those on the BOM list.. very confusing. I'm really happy to see that your refactor included these instead, as I have little interest in using a solder iron to hand solder smd leds one leg at a time.

I strongly prefer the workflow of ordering a smd stencil, using it to apply solder paste, placing the components with a tweezer and using a hotplate to melt the paste vs hand soldering countless smd components one leg at a time.

When do you expect to have 1.5 validated and a zip of 1.5 gerbers available? I don't have kicad experience, so I'm worried using the gerber files directly will miss something important. I think I will hold off ordering from jlc until 1.5 is verified.

Thanks again for helping revive the klor, we're lucky to have someone as skilled as you to update the original design!