Snapmaker 2.0 Laser module Part 1: The autopsy.
Described failure: The laser output can no longer burn materials.
Let’s dig into it.
Opening the side panel reveals the laser housing. Looks intimidating, but it’s mostly just a giant aluminum heat sink. The actual laser diode is quite small. It’s likely the part that failed. Just a guess. We’ll get into this later. Can’t go any further this way.
Taking the panel off the other side reveals a board. Let’s get it out and see what it does. At the bottom you can see the camera ribbon cable.
Removal of the board shows 3 cables? One 4 pin, and two 2 pin JST connectors… What? That’s unusual. JST connectors aren’t really suitable for camera data purposes, not to mention carry the the ability to supply power to the laser diode, fan output, along with led light and camera controls? This going through 8 pins through 3 connectors simply isn’t enough for all that.
Curious…
Looking through the microscope shows what looks like an ARM Cortex-M3 style microcontroller. This is what handles can + and can - communications from the control module.
Well well well, what do we have here? An XL3005. Here’s how the laser’s power output is controlled. This is a PWM (pulse width modulation) power control. A simplified way to explain how it works is let’s say we have 36v coming in (100% power) and I turn the power on and off quickly an equal amount of time then pass this through a capacitor to smooth it out. This averages out to be about 50% of its initial power or 18v. There’s a bit more to it, but this is the gist of it.
A little homework reveals this chip has a 220 KHz fixed frequency step down PWM buck converter LED constant current driver. These are capable of driving a 5A load with rather high efficiency. They have an internal regulator and include internal frequency compensation and a fixed- frequency oscillator. This chip alone controls the duty ratio linearly from 0 to 100%. An over current protection function is built inside. Any upgrade to the laser will require some changes. Fun fact, these chips cost a few cents if ordered in mass. I found them for about $4 with shipping. Kind of hope this is the problem, nothing wrong with an easy fix. This explains one 2 pin JST connectors of the 3 JST connectors.
Removing the bracket that held the first board reveals a second board and answers some questions.
This explains how they went about controlling the camera and led light. The laser module communicates to the module via Bluetooth! An inventive, albeit unusual way of dealing with the communication bottleneck of sending all that data back through the can lines. There was no way the cortex chip with an operating frequency of 108mhz could handle that camera efflciently… So it likely just gets power from the 4 pin cable and simply talks directly to the controller. The other 2 pin JST connector goes to the fan.
Needed to unscrew the tip from the laser to get enough clearance to remove it.
Here you can see the camera and led board. No need to mess with it for now.
After removing the fan you can see the business end of the unit. And that’s still not it. Under that glue is the diode itself. Like I said before, it’s small. I will be powering it up to see if the failure point is the diode or see if the supply voltage is large enough to cover the voltage drop.
Stay tuned for part 2: Component testing.