Laser Cutter Woes…
About 4 years ago THS raised funds to buy a cheap Chinese K40 laser cutter. These laser cutters have become very popular within the hacker/maker spaces because of their cost and hack-ability. For most of 2017 the laser cutter was inoperable. We had decided the software that came with the laser cutter had to go, and replace it with something that was open source and easier to use. We had finally achieved that goal when 3 weeks ago disaster struck. One member and a non member tried using it to cut something, and it wasn’t working. Another tried the next night with the same issue. I personally was frustrated as I had spent a lot of time working with 2 other members to get it working, and to understand the new software. I was worried that the new controller board we installed had failed. Going back to the old controller board was not an option.
Early the following Saturday one of the members who found problems with the laser cutter and I showed up at the space to debug the problem. We turned on the laser cutter loaded LaserWeb put in a piece of scrap MDF and ran the compiled g-code. The motors moved, and the power supply relay clicked. All a sign that the controller board was most likely not the problem. That left one of two things. Either the high voltage output on the power supply or the laser tube. Either one would cost us at least ¥10,000 (USD$100) to replace. Pulling the laser cutter away from the wall and removing the exhaust fan we opened up the tube housing. The question for me was how do I test a 10KV power supply without killing myself. With the laser cutter on low power I decided the first step was to just watch and see if anything could be visually seen. I honestly didn’t expect anything because lasers are invisible to the naked eye. With the laser on fairly low power I clicked the test fire button. To my surprise the laser tube glowed purple. For me this was great news. It told me the power supply was outputting voltage. Was it enough to create the laser for cutting? I didn’t know, but it was something. I then took a piece of scrap mdf and output it at the end of the laser tube and clicked the test fire again. No marks on the wood, no burn, no smoke. I had hoped that maybe the mirrors were just real dirty but having the MDF right in front of the output proved that the laser was not exiting the tube. Not wanting to accidentally injure myself I turned off the laser cutter and began to visually inspect the laser tube. I would soon discover that the output end of the tube had broken off. I hadn’t noticed it before because it is held in place by a piece of silicone tubing used for passing the cooling water from one aspect of the tube to another.
You see a CO2 based laser tube (the kind found in the k40) uses distilled water for cooling. The water is pumped from a holding tank into the output end of the tube first. The output end of the tube is actually a piece of glass that is glued with silicone based glue to the output lens on the end. The water runs through this section of glass and keeps that mirror cool. There is then a small piece of silicone tube that passes the water to the main tube. The water then travels down the length of the laser tube and then goes through another piece of silicon tubing to another piece of glass tubing that is silicone glued to the high voltage end.
What we suspect happened with our laser tube was from improper use. Our water reservoir is located below our laser cutter. If we turn it on water will flow, but when we turn it off gravity feeds water back into the reservoir. When the water is turned on air bubbles get stuck in the tube and unless you flush the air bubbles correctly proper cooling doesn’t happen. In our case the air bubbles got stuck at the input stage, minimizing cooling of the output lens and eventually it heated up the glue and the cooling ring disconnected. This prevented any cooling of the output lens and lead to it’s inability to pass the laser out of the tube.
What the problem was, was solved. The laser tube was broken. This meant we needed to buy a replacement. The member helping me debug the issue searched for tubes online in Japan, but could not find any. That meant shopping Aliexpress. I spent the next night looking at various vendor offerings. Prices ranged from $65 to $200, but most the ones with free shipping averaged around $110. I wanted to manage the shipping so I chose one that didn’t offer free shipping, and in the end paid $100. $65 for the tube and $38 for shipping. As many who order from aliexpress and know often when you order something with free shipping you are at the mercy of the Chinese postal system. Your package is given minimal priority and could take up to 30 days to arrive. I wanted it here in no more than 2 weeks. I selected FedEx as the shipping agent, and sure enough it took a little under 2 weeks to arrive.
One of the features I chose when I ordered the tube was that it already had wires wrapped around the high and low voltage ends. After watching a few youtube videos I decided it was a less fussy way to go, making the install so much easier.
I arrived at the hackerspace this afternoon. It was our monthly open house. There were two other members there and 4 visitors. Two of the visitors both Americans were on their way out, so the rest of us all went downstairs to watch/help with the installation. First we pulled the laser cutter away from the wall and removed the exhaust fan housing from the back and put it aside. I cut the wires close to the laser so we didn’t lose much length, and then unscrewed the mounting clamps that keep the laser tube in place. The tube still had water in it so I carefully disconnected the tube and brought the tube somewhere I could empty the water.
Cutting the high voltage wire was a bit of a pain. You see laser cutters use extremely high voltages. The wires they use have traditional insulation on the outside, but a hard plastic insulation on the inside. I didn’t know to expect this, so I ended up trying to use traditional wire strippers and ended up breaking off about 2cm of the wire accidentally. It didn’t help that the shop has poor lighting. Eventually the other members cleared off some space near the garage door and we moved the laser cutter over so I could see what I was doing. Finally got the insulation off both wires. Did something similar to a western union short tie splice, and then soldered the splice for good measure. I didn’t have any hard plastic to put over the splice, so to try and be safe about it we put 2 layers of heat shrink wrap on it, and then used a layer of electrical tape. I figure no one has any reason to be in the tube chamber so it should be safe. I made sure the new insulation went 2cm from the splice in either direction just to make sure. Doing the same for the other end completed the electrical aspect of installing the new tube.
Next was time to get the water tubes back on. This was a bit of a pain. I struggled to get the small tube on the end of the glass nub. Another member suggested using margarine. I laughed. That seemed like a bad idea introducing food particulate to the cooling system. Apparently it’s common in Japan to do that with this kind of tubing. Instead I sent him off on his motorcycle with a sample to find something bigger. He came back with some pink hose used for gasoline lines. It was perfect. Took us a few minutes to fit the water tube on the laser tube water inlet/outlet, but soon we were pushing water through the tube.
When I turned on the water pump it pushed water into the tube. However air bubbles were not escaping. I theorized that the tube needed to be rotated a little so the bubbles could escape. My theory was correct a quick twist clockwise about 10 degrees and the water was flowing. However we still had bubbles in the tube. These ones were back down where the output lens was. Lifting the laser cutter about 20 degrees on one side the bubbles came out and all that was left in the tube was water. Great just what we wanted.
This is where I learned something important that if we had known previously then this probably wouldn’t have happened. Turning off the pump gravity did it’s thing and pull 98% of the water out of the tube. Turning on the pump created the same situation with air at the output end of the tube. This has probably been going on all along with no one noticing. This was probably why our original tube failed after so little use.
“Ah gravity, thou art a heartless bitch” — Sheldon Cooper TBBT
Rebalancing the tube so that it was all water, I turned to testing the laser. I put tape over the first bounce mirror to make sure it was still aligned. Test firing proved that the laser was indeed working and that it was about as close to center as I was going to get it. Ramping up the power we put a piece of MDF in the cutting chamber and test fired again. Bingo we had burn.
Another member pulled out his laptop and fired up laser web, and within a few minutes we had cut out the logo for a local festival we participate in every year. The install was a success. We had never done this before. All kinds of things could have gone wrong, but we took our time, we researched the process and in the end were rewarded.
While the laser cutter is now working there is still some work to do with solving the air bubbles. For now we’ll just have to remember to rebalance the tube every time we use it, but long term solution is to bring the cooling reservoir to a location that is higher than the laser tube inlet so that gravity works for us instead of against us.