HP WMR – Alternate Headstrap Mod

The stock halo style headstrap for the HP Windows Mixed Reality headset is fairly comfortable, but it positions the lenses quite far from your face, and tends to droop off your face when looking downwards. The lens position tends to shift a lot too.

I decided to remove the stock halo, and attach a more conventional fabric headstrap. The stock halo can be removed with 4 screws inside the faceplate, located under a sticker. The headstrap was salvaged from a cheap phone VR headset from Walmart, which can often be found for only $1, and is actually quite nice. The brackets are 3D printed and glued to the headset, and 30mm M3 standoffs and screws are used to attach the headstrap to the brackets. Overall, the appearance is quite nice, and looks almost stock, with the exception of some glue leakage.

After a play session, there are definitely a few compromises I noticed. The straps take a bit of adjusting to get the right tension without being too tight. I also could not as easily wear my glasses, though I personally don’t need my glasses in VR. Once I had everything adjusted correctly, the benefits definitely outweighed the downsides. The perceived field of view is much larger, as my eyes are a bit closer to the lenses. Moving my head feels more direct, due to the headset being more secure to my face with less wobble, and the lenses don’t shift away from my eyes when looking down anymore. Light leakage is now almost zero.

Parts used (Amazon Sponsored Links):

Standoffs: https://amzn.to/37GtMRJ
M3 Screws: https://amzn.to/2N4PhUx
3D Printed Brackets: WMR Bracket STL

Pilot Trak 3.0 – Final Design Update

Well, it all fits together! Just add receiver and a battery, and it could be up and running, though I still want to try making my ESP32 wifi fpv control unit work. Chassis / Cab / Rails print in 3 pieces and snap together, wheels screw on.

Scratch that whole old V3 design. This does still use SG90 servos, and snaps together as intended, but resorts to using M4 screws as axles, as the plastic axles did not work out so great. I also flipped it backwards and extended the servo cover into a cab. Instead of going full utilitarian with this, I figured making it a scale model would be more desirable for those who don't want or need FPV, or just those who want something that looks kinda cute. It's currently designed to house an ESP32-Cam in the cab, otherwise I'll have to make an adapter to sit a camera on the seat or above the cab.

This design uses two SG90 continuous rotation servos, as they're very cheap and do not require a separate esc purchase. I'll have to make an adapter to fit an FPV camera in the cab or on the roof. Maybe poke a lens through the back window! It also houses two 5mm LED headlights. The most basic build could be as simple as a 1s Lipo, camera, and both servos plugged into a receiver.

For those who don't have FPV equipment or a transmitter, I will also produce a version which uses an ESP32 to power a wifi-controlled interface.

PilotTrak V3

So I think one of the biggest challenges in getting into a Tiny Trak is cost and complexity. I previously used N20 gearmotors, but the cost of them plus an ESC is a little overwhelming. Not to mention, having to buy hardware / screw kit to assemble. I'm working on a TinyTrak design which will be 100% 3D printable including TPU tracks, without use of screws or tools for assembly. All axles clip into the frame, and all frame parts are designed to snap together.

This design uses two SG90 continuous rotation servos, as they're very cheap and do not require a separate esc purchase. The front accomidates a typical micro FPV camera lens, along with four 5mm LEDs. The most basic build could be as simple as a 1s Lipo, camera, and both servos plugged into a receiver.

For those who don't have FPV equipment or a transmitter, I have also been planning on using an ESP32 as a solution to the electronics. This is a cheap microcontroller with built-in wifi and is often bundled with a camera for less than $10. Ideally, the ESP32 would run a webserver to provide controls and video feed to any computer or phone with a web browser. This also opens up the possibility of using the device for remote monitoring & control over the internet.

PGR – 3D Printed Revolver

Still in development, a few light primer strikes, but works mostly okay! Have since lengthened the firing pin to account for variations between cylinders. This isn't designed to be practical or useful (it's neither), more of an engineering challenge to myself. 3D Printed designs are a bit of a challenge for something like a gun, because plastic really doesn't have very good tensile or shear strength, so you have to overbuild components, and carefully design parts that support compressive forces. This generally leads to pretty bulky designs, but with a bit of metal in the right places, a compact design is possible, albeit with less features than a standard revolver. Ammunition is currently only 22 shorts, but a larger frame and a metal insert for the breech face could change that. 22 Shorts are my favorite test ammo because when a part fails, it's usually less catastrophic than larger, more powerful ammo.

Non Plastic Parts used: Drill bit (Firing Pin) Steel tubing (Cylinder Liners) Springs (Striker & Cylinder Indexing) 5/16" Ball Bearing (Cylinder Indexing) M3 Screws (Front Sight, Trigger Assembly, Rear Sights) M5 x 65mm Screw (Cylinder Center Pin)


Autonomous Tiny Trak, a quick and dirty project adding simple autonomous function to one of my PilotTrak platforms. As always, no microcontrollers here, just a simple digital circuit. Sorry to disappoint, there's absolutely no tracking or mapping going on here, he's honestly got the intelligence of a rather bright potato. Easy to make though, all off-the-shelf circuits.

Pathfinder Mini – Logic Gate Robot, no CPU!

Based off my larger walker, the Pathfinder Mini is a cost-reduced version which requires far fewer parts, much less plastic, and is a bit more simple to build. I will be releasing the plans for it as soon as I can, and considering the reduced part count, I may be able to sell complete kits instead of my usual plans or plastic-only kits.

If it were economically feasible, I would also like to make a batch of circuit boards to include with the kit. As of right now, it's too complex to build in point-to-point wiring, but a PCB with proper traces would be nice. I'd have to ditch the LEDs as a cost saving measure though as well. Or, you know, I could just use a microcontroller like a sane person.

This circuit was really just a fun project to prove I could build this idea from my head into an actual working robot. It does work quite well, and schematics and an explanation of the design is listed below.

Clock Gen

Astable 555 timer which generates the clock signal for SR04 Sonar Sensor and 4017 Counter.

Sonar Sensor

HC-SR04 Sonar Sensor. A clock pulse triggers a sonar ping. Pulse width on echo pin is proportional to echo time.

Missing Pulse Detector

The echo pin from sonar seamlessly refreshes the monostable 555 timer, keeping the output high. A resistor-capacitor circuit snubs out short pulses from close objects, causing the output of the 555 to go low. This causes a reset for the “Counter” and “First Loop” circuits.

First Loop

RS flip-flop “register” using a 555. Upon reset, the output is set high. The last output from the counter will set the output low


“Program Counter” using a 4017. Each clock pulse pulls one of 10 pins high, in sequence. A reset brings the output back to 0.

Diode “ROM”

Diode logic is used to hard-wire program instructions to each step of the counter. These instructions consist of “Left, Right, Stop, Up, and Load MEM”.

Light Sensor

Functions outside the program loop. Injects instructions for “Stop” and “Up” when light is above a threshold. Not strictly necessary, but adds a bit of interactivity.

Last Direction

RS Flip-flop “Register”. Retains the last direction executed from ROM after a reset.

MEM Compare

Combinational logic to determine the results of the MEM instruction. If the “First Loop” register is true, the “Last Direction” register is interpreted as either a “Left” or “Right” instruction. This provides obstacle avoidance when a reset is triggered by the sonar sensor, and is ignored in subsequent loops. Also only allows STOP instruction on first loop.

Motor Driver

A L293D motor driver circuit. I recommend capacitors on the motor outputs.

Servo Driver

A 555 timer in astable mode. The output is pulse width modulated via a voltage on pin 5, resulting in basic servo control for head tilt.




Ideas for a small BB shooter, possibly for a TinyTrak.
Integrated hopper, auto-feeding, uses single N20 GearMotor. Requires 8x3x4mm ball bearings and M3 screws.

Hopefully someone else can improve on this?


AutoBB Download

PG22 Short Update

Well, I built myself a second personal PG22 Short. Doing this allowed me to follow my own instructions, and find any remaining flaws in design or instructions. This did result in a minor revision of the location of the set-screw for the guide-rod, along with more explicit notifications on cleaning and sanding parts.

I’m happy to report that my second build not only fired properly on the first shot, but ejected and continued to function throughout 20 consecutive shots, with no visible wear or damage to any parts.

Because this gun is now a spare for me, I decided to have a bit of fun when printing it, and went for a more wild color scheme. I hope others find this as amusing as I do 🙂