Recent searches
Search options
Found this beautiful unused keyboard in storage at work. Switches are all Cherry Brown and I really like the feel of them. Only gotcha: the interface may be an oddball.
Hey wait a gosh dang minute what is going on in here
The two holes are buttons and it looks as if two of the traces from the RJ-45 go into a transformer! There's also an unused header.
oh HEY. Turns out USB common mode chokes are a thing and that's probably what I'm seeing there!!
https://www.coilcraft.com/en-us/other/usb/
also I do wish Coilcraft's website favicon didn't look disturbingly like the Chick-Fil-A logo.
TRI-COLOR FOAMING WAX
if you can't tell that's being typed on the Ross keyboard and I love the feel of this thing. yes. After blowing through an embarassing number of RJ45 plugs (well only four) I got the pinout!
3. VBUS
4. GND
6. D+
7. D-
I guess if you use an STP plug you can ground the drain at both ends - I just left it open at the keyboard end and am allowing it to get grounded at the host end. Works fine.
I just unscientifically hit typingtest.com and did the 1 minute test on this board, scored 92 WPM 96% accuracy for an adjusted speed of 89 WPM.
....I think this key feel works for me
Interestingly, it comes up with a different model number.
[1637511.148933] usb 2-2: New USB device found, idVendor=173e, idProduct=0038, bcdDevice= 0.06
[1637511.148943] usb 2-2: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[1637511.148947] usb 2-2: Product: KMX-143-750-583
[1637511.148950] usb 2-2: Manufacturer: DEVLIN Electronics
[1637511.148953] usb 2-2: SerialNumber: KMX-143-750-583
Yeah, potato, po-tah-to
that model number brings this up, which mentions more of the OEM customizable features - also explains why I saw a bunch of unused pads on the switch pcb that looked like they were for some kind of LED drivers (maybe just shift registers?).
this mentions stock trading terminals as an application, which makes me wonder... if you do an image search for a Bloomberg terminal keyboard you'll find a few different versions.
Did Devlin also make at least one version of that?
There's one that does look kinda look the same keycaps but it's a Maxi Switch product.
https://www.linkedin.com/posts/adafruit_the-bloomberg-keyboard-classic-1996-https-activity-6901563827666661376-KTBG
What an odd keyboard! It must have some specialized purpose. Is it designed for working with some specialized device or operating system?
Christ on a cracker, look at that price tag! 
https://alliedbroadcastgroup.com/product/ross-video-xpression-custom-keyboard/
@DXMacGuffin welcome to the broadcast world! would you care for some questionable bitter coffee in the breakroom? lmao
How "questionable" is it? Like is it the cheap stuff that tastes like engine oil? Or is it the stuff that's like $400 a cup and is made from ground up cat shit?
@DXMacGuffin the cheapest of the cheap :D
Well you've got that going for you. They spend all the budget on super expensive keyboards, instead of coffee made of cat poop. The glass is half full.
@DXMacGuffin it's for a broadcast graphics/character generator kinda system. https://www.rossvideo.com/live-production/graphics/xpression/
The site for the current version of it shows a similar looking keyboard!
@DXMacGuffin it's the same layout and everything, same manufacturer, MAYBE even the same PCB. Now that the morning show is over I wanna try tracing this out. I'm pretty sure if it's just a USB cord terminated in an RJ45 I can ohm, sweet ohm, my way through the maze and gob a passthrough RJ45 connector onto the wires of a cast off USB cable.
yes I mean this https://www.youtube.com/watch?v=QLwEG3cdeRw
lol...godspeed
, powered by sarcasm @vxo that stm32 has USB support. Time to get qmk on it!
@sycophantic I was actually wondering if one of the buttons on the back was to hook the STM32 bootloader to flash it! I guess the next step if I can't find interface info would be to start tracing pinouts and figure out how to talk to the little guy
@sycophantic oh, I guess I could have just gone straight to their documentation!
https://documentation.rossvideo.com/files/Manuals/CG%20and%20Graphics/XPression/Hardware%20Specs/XPression%20Custom%20Keyboard%20Specs.pdf
it looks like at least now, they come with an attached USB cord - which probably lands right on that unpopulated header I found next to the rj45.
@vxo lol we could have but that's the fun
@sycophantic Well, I might still need to ohm out the +Vbus, D+, D-, and Ground since they're not marked out on J6 which I'm guessing would be for a JST header for the USB cable, hopefully that's all she needs
@sycophantic hey what if I'm just misreading the labels and TR1 is for TRansistor, not TRansformer, and that's actually just kinda a ferrite bead on the USB lines? hmmmm. I mean, the two wires that have gotta be data lines stand out on the board, they go through that and then through R9 and R14 and cross under the STM32.
@vxo looks like boot0 is tied to Vss. If I read the board and pinout right.
@sycophantic Thanks! I think the next step may be to try soldering a usb cable to J6 there - I suspect one of the electrolytic caps will be conveniently right across VBus and GND and will let me figure out the power right off the bat.
Is there any harm in accidentally crossing D- and D+ or will the device simply fail to wake up 'n' enumerate?
@vxo the transformer is used to electrically isolate the RJ45 signal from the cable. if only 2 traces are used then likely it's a telephone interface instead of an Ethernet interface
@vxo if it's for a broadcasting system like you say then the telephone interface makes sense - you can carry 4 telephone lines over a cat5 cable and use a switchbox to bus them together. likely that there's an audio signal going over that line in that case.
@vxo absolutely could be. could also potentially be some power over ethernet abomination
@vxo that also explains why the two signals pass through debug header j6, so that you can sniff the outgoing audio with an oscilloscope
@vxo this looks like an 80's toy musical instrument. 
@vxo common mode choke, you have usb
@vxo you’ve already found it, but another giveaway that it’s probably USB is that the traces are a constant width apart, splitting and merging an equal distance each time they hit a component. USB is a differential signal so it needs a calculated distance between the D+ and D- traces to meet the differential impedance, and also needs equal-length traces for the data lines.