A few months ago, I started to modify a record player. Like all good short projects, the scope grew until I was making a new player from scratch. All parts are 3d printed, and electronics can be soldered at home. (more info and plans)
I’d wanted a record player with an exposed platter and drive belt. I had an old piece of tulip wood from a jewelry box that had fallen into disrepair, which seemed like a good starting point.
I started with some rough sketches and used my low-end $30 Jensen record player as a known-working template. From there, I redesigned each part, fitting together into the new design.
From the original player, the only remaining parts are the cartridge, motor, and drive belt- all easily sourced online.
What does this mean? We can make more record players!
Why build a record player?
For a long time, I had no interest in vinyl records, and wrote it off as silly retro fad.
My wife bought me a tube amp for my birthday and a record player to go with it, and I started to get the appeal. I as much as I don’t buy the sound quality argument, I see the romantic appeal of sound waves etched into a disk, and the joy of watching it spin.
The scale and simplicity of a record player made it an ideal, satisfying project. No need to bring out the microscope to solder stuff. Still, it’s unforgiving, and it’s easy to make the music sound terrible.
Design and Printing
I went for a mid century modern look with clean angles. All of the parts are exposed- built around a single piece of wood, there’s nowhere to hide anything. Parts were designed in Autodesk Fusion 360, and printed on an Ultimaker 2.
The platter rotates around a skateboard bearing- super cheap.
This was the most involved physical thing I’ve designed and built. There are about 15 printed parts, most needing to snugly fit together or rotate smoothly. Fusion 360 made most of this a breeze, there’s still a steep learning curve.
The Ultimaker 2 was a beast and printed everything without a hassle. There’s a little bit of a learning curve, though. Parts would fit together too snugly, and I’d need to build in some tolerance. Tiny circular parts are also slightly out of true.
Looking forward, I should have spent more time planning where to mount things. The motor is a mess, and placing the circuit board was an afterthought.
I’d initially planned to just use the original circuitry from the Jensen player, but it looked pretty ugly.
The electronics feature motor control and a preamplifier. I built in support for to stop the motor when the record is finished, but didn’t get around to building the mechanical parts for it.
Amplifiers are not my thing, but I think it was a decent attempt, and I think it sounds a bit better than the original. I used a LM386 op amp, and put in a basic bass-boost. There’s probably a lot of room for improvement. (Have any experience here? Let me know!).
I don’t do a lot of electrical design, but the tools always seem really dated and frustrating. I’m prone to errors when manually designing part packages, and long turn around time mean slow iteration. I wonder if there’s an opportunity here for sharing part package repos (i.e. NPM, Rubygems, RPM, etc.), or tools to quickly digitize mechanical designs.
After the final built, the music sounded terrible. Just this awful, wavering quality, especially on piano music. I spend a solid month obsessing about platter wobble. I designed new stabilizers. Bought new drive belts. Fretted about circles that were out of true by 50 microns. You know- crazy person stuff.
This is the point where the project stopped being fun and started being work, and exactly what I promised myself I wouldn’t do. I quit and didn’t touch the player for a week.
When I came back, I realized the player arm wasn’t steady. The original prototype had a small piece of sticky tack on it stopped the player arm from wobbling. I popped the sticky tack back on, and it sounded great again.
Watch out for target blindness- take a break and look at the problem again with a more open mind.
I’m starting on a new player as a gift, and there are few things I want to try:
- Magnetic stabilization: Using a ring of magnets beneath the platter to prevent wobbling.
- Optical Timing: It’s 2017- why do we need a tune knob? Monitor the rotation and adjust the motor speed in real time.
Interested in a kit or want to buy one? Let me know!