Where to go after Nauvis
I've just bought space age after a couple of playthrough and I'm currently researching the spaceport. What should be the modus operandi expanding into space?
I've just bought space age after a couple of playthrough and I'm currently researching the spaceport. What should be the modus operandi expanding into space?
This is incredibly frustrating and expensive for me. I have my camera mounted in the same spot with manual focus when I'm working. I turn it off constantly and pull the USB disk during longer sessions where I don't expect to record anything, because otherwise it for some reason drains the battery if I simply turn it off.
Every time I do this I need to refocus. Sometimes I forget and the footage is ruined with no way to reshoot. This is becoming the bane of my existence.
Is there some way to fix this?
Hey!
I'm doing a bit more advanced stuff now on my 3-axis mill, but I've been having problem turning the piece over and milling both sides. So I am looking for some PROFESSIONAL advice, to aid my self learning process.
The material is a 200x60x20 AW6061, the vise is a small 50mmx25mm, and I'm using parallels that go up from 14mm in 4mm increments. The raw stock cannot be milled on its side as the vise builds too much height for it to be practical.
- First setup I carve out a reference edges to get it at a predictable position in the vise.
- Second setup I turn the whole thing around and mill the backside, leaving another handle and two small dimples to use as reference for setup 3.
- Setup 3 I turn again and mill away the rest of the material, leaving 0.5mm for the two pieces to stay attached to the "handle".
So what mistakes or extra work am I doing here?
I have been designing a trackball in my head for a while. I ordered the billiard ball and the bearings, and made this CAD design as soon as they arrived. I took apart the optical mouse I planned to use for the project and discovered:
- it doesn't track on the shiny surface of the billiard ball
- the PCB of a mouse is too large to fit in the bottom part of the design.
- the sensor has a LOT of pins, so desoldering and moving it to a breakout board seems like a big project
I have a few options I've been thinking about
- buying a used Logitech mouse with the "darkfield" technology, which supposedly is good at tracking shiny surfaces (unsure if it will work). For example Logitech MX Master 2S
- using old-school mouse rollers (complex mechanical design)
- buying a commercial trackball to reuse the sensor (unsure if it will work, and expensive)
I'm sure there are a ton of solutions I haven't thought of. So to the clever people of Reddit, please help!
Hey!
I've been trying to upgrade from Sonoma for a while now. No luck, I'm getting "failed to personalise" errors. I've downloaded the installer and ran it manually, no luck.
I've booted into rescue mode and tried reinstalling Sonoma and I'm getting the exact same problem there.
The logs are extremely spammy, but complain about some missing /var/mobile directory (from the top of my head).
Is there some repair procedure? Some way to get more information of what is wrong?
God I miss Linux.
This is a Teleguide, or rather it was a Teleguide; a complete flop of a project from the Swedish state telecom operator, released in 1991. They pushed around 10,000 units and a lot of them still sit in storage in unopened boxes around the country, so I got this one for myself, new in unbroken packaging with styrofoam and spiders and everything.
Inside it sits a couple of Intel P8031AH microcontrollers, one for the card reader and one for the main program which allowed you to bet on horses, order pizza, keep a phone book, and call your enemies. Those are now depowered and replaced with a Raspberry Pi 5. The godawful keyboard was replaced with a really nice [EV63 Hall Effect keyboard](https://iqunix.com/products/iqunix-ev63-hall-effect-keyboard), sponsored by IQUNIX after seeing my [previous article](https://www.reddit.com/r/retrocomputing/comments/1rrpyiw/how\_do\_i\_turn\_this\_thing\_brown/).
The goal of this project was to turn this machine from an interior decorative detail into a gaming machine, and to do that I needed to somehow get the monochrome 10" CRT working with the Raspberry Pi. The monitor does not follow any standard composite or VGA mode; instead it runs at roughly 18.75 kHz horizontal and ~57 Hz vertical sync, which is just far enough from standard timings to be annoying.
After a lot (and I truly mean a lot) of trial and error I managed to get a prototype working with a Raspberry Pi 3. That prototype used the Broadcom VC4 GPU’s DPI (parallel RGB) output. To reduce the number of GPIO pins I implemented a software LUT in the driver that mixed RGB into grayscale, and then output that grayscale value on all three channels, and then only muxing a single channel to GPIO pins using a device tree overlay; so 8 bits in total.
That approach worked, but it relied on behavior in the VC4 pipeline that doesn’t carry over to the Raspberry Pi 5.
The next step of the project was to move to a Raspberry Pi 5, mainly for performance. The problem is that the VC6 pipeline doesn’t expose the same LUT mechanism in a usable way, so the grayscale conversion had to move out of the driver and into hardware.
The new PCB uses a 16 bit dac; 5 bits red, 6 bits green, and 5 bits blue. This mode leaves the PCM pins available for use with an audio DAC, and a few pins for an input or output device, such as a joystick.
You can find the [Kicad schematics here](https://drive.google.com/file/d/1b4ZBXvEhbKEOGUg36lYMHon5TDy6n9KJ/view?usp=sharing).
The solution I ended up with was to treat the Pi as a proper RGB source and build an external analog front-end:
Each color channel is converted to an analog voltage using a resistor DAC
The three channels are resistively summed into a luma (Y) signal
The result is AC-coupled and buffered using an NPN emitter follower before being injected into the CRT circuitry
This turned out to be much more stable than the original prototype. The key issue is that the resistor DAC has a relatively high output impedance, and the CRT input is not an ideal high-impedance load. Without buffering, the voltage level shifts depending on what the CRT circuitry is doing. The emitter follower isolates the DAC from that load and stabilizes the signal.
The injection points are:
Luma (intensity) is injected at the wiper of the brightness control potentiometer
Sync is injected directly into the sync input node after disconnecting the original video source
For sync I used a simple diode combiner:
HSYNC and VSYNC from the Pi are OR’d together using two diodes
The resulting composite sync is fed into the CRT sync input
This works because the monitor expects a negative-going composite sync signal rather than full composite video.
At this point I ran into a new problem. I designed the board around RGB565 (5/6/5 bits) to reduce GPIO usage and leave pins free for things like audio and input devices. However, there was no straightforward way (that I could find) to get the Raspberry Pi 5 to actually output RGB565 over DPI.
The workaround was to patch the kernel driver to add support for this mode, and then provide a matching device tree overlay to configure the GPIO interface. Slightly painful, but it works.
You can access the [kernel driver patch here](https://drive.google.com/file/d/1HCAgP8rILkHzMR\_37rmhemRFdOMo2rYg/view?usp=sharing).
And now I can finally play games on this thing.