The Improved Calibration Process, Some Beautiful Box Joints, and Complete Instructions: The April 26 Update

Hi everyone, happy Wednesday!

This week’s update talks about how we’re resolving the calibration issues we saw last week, the completed first draft of all the assembly steps, our progress with Autodesk Fusion 360, and some cool box joints made by the community. It’s a big one so let’s get into it!

Resolving Calibration Issues:

Right before last week’s update we were struggling with big issues with the calibration and settings process. Things were so dire that we even went so far as to lock the settings in last week’s release because changing them seemed to do more harm than good. I’m proud to say that within hours of the update the community had found the bug that was causing the settings issue. The community then went on to devise and test a calibration procedure that is making it easier to get the machine dialed in. We can and will continue to improve the calibration process, but the amount that was accomplished in the last week is astounding.

I would like to extend a huge thank you to everyone who has participated the 350+ post ongoing conversation into calibration. I say this every week, but I am continually blown away by the insight, quality of work, kindness, and willingness to get involved that our beta testers show every day. You guys make it so every morning when I wake up and check the forums I’m blown away by something you’ve created or thought of since I went to bed. When I go to bed at the end of the day I’m thinking about all the great ideas the the community proposed to make the machine better over the course of the day. Thank you.

For those of you who don’t want to read all 350+ posts, here’s the summary. Every step along the way was an idea come up, tested and validated by the community.

The first step was to track down the bug which was messing with the settings. That was done with the help of Keith who worked out the original mathematics for the machine. The issue was a one line issue with the way the settings were being updated (my mistake). The core mathematics of the machine remain exactly as they were when Keith first proposed them.

The original forum post proposed creating a test shape, possibly a square and a circle which we could cut, measure, and then tweak the machine settings based on the measurements. It was decided that the test shape would be a 6 inch square with a circle inside it. Using this test shape, we were able to determine that by far the most sensitive setting on the machine is the distance between the two points where chains attach to the sled. Better yet it was found that, if the square is taller than it is wide, the settings value needs to be reduced. If it is wider than it is tall, the settings value needs to be increased. By tweaking this parameter we were able to bring the square into true as shown in the photos below.

The next step has been to explore how our calibration holds across the 4x8 sheet. Early this morning (at least it was morning in my time zone) community member Gero ran a test that I think better than anything else shows the current state of the calibration process. He wrote gcode to run the test shape at fixed intervals across the entire work area resulting in this very satisfying image.

Gero's test across the board

Gero's test across the board

After measuring all of the squares we discovered that we’re seeing about 1 mm of drift across the 4x8 area. Still room to improve, but so much better than where we were last week.

The test shape at the center of the 4x8 area

The test shape at the center of the 4x8 area

Excitingly, this center mark on the middle shape was re-drilled between each of the shapes meaning the hole was drilled 16 times after moving from each part of the work area. That it was drilled in exactly the same location (within .5mm) shows that Maslow isn’t losing any steps or drifting over time.

The next step is to automate and simplify the calibration process and to improve the feedback control system which is also certainly a factor when cutting test shapes. We’ll keep you up to date in next week’s update.

Instructions are complete:

I wrote in my last update that the calibration issues were preventing me from finishing the instructions, and that I would finish them as soon as the calibration issue was resolved enough to let me. True to my word, as soon as the calibration issue was resolved enough that I could cut parts I completed a first draft of all of the instructions. You can find them at www.MaslowCNC.com/assemblyguide. While the earlier parts of the guide have benefited from some fantastic community feedback, the later parts haven’t been read as much. I know many of you have already moved past these steps, but if you have a second to read through them any advice you have that could make them better would be much appreciated.

Fusion 360:

We talked in last week’s update about how visitors from Autodesk came see us, we’ve also mentioned in past updates that we’ve been working on making Ground Control compatible with their Autodesk Fusion 360 software. We’re excited to announce that since last week’s update we’ve been in conversation with their technical team and they are going to be giving us our own Maslow post processor option. That means that when you generate g-code in Fusion 360 there will be a “Maslow” option which will format your design just the way Maslow wants it. We’re by far the smallest and scrappiest group to get its own post processor. They were impressed with the innovative design of the machine, the fact that we are an open source project with the goal of making CNC more accessible, and the fact that we have this amazing engaged and creative community. As always, the most impressive thing about us is you.

Project of the week:

This week we’re showing off two awesome box joints made by the community this week. Box joints are a good way to test how accurately the machine is cutting and since this week’s update is all about improving accuracy it just seemed to fit (Get it?! - Oh geez).

First off is a box joint made by community member mindeye as part of a Seattle Maslow meetup. He found that one trick to getting a good cut was to cut all the parts .05 inches larger than desired, then come back in one finishing pass to clean them up. Brilliant!

We encourage you to consider meeting up with other community members in your area. There’s so much to learn from each other and so much information to share. Thanks mindeye for hosting the Seattle meet up and posting the pictures in the forums!

Our second box joint of the week comes from community member rancher who not only made this beautiful mailbox, but also made an incredible YouTube tutorial about the process. What a community!

Have a great week everyone! Thanks for reading all the way to the bottom. We’ll be back next week with more news and until then we’ll see you in the forums!

Hannah’s Mini Update:

Hi everyone! All I have to say is that this Friday is our Open House and Party!!!  For anyone who is on the fence, we encourage you to head to Portland to check us out and have some fun with us! And if you can’t make it but know anyone in the area who might be interested, please send them our way! We want to meet as many Maslow community members as possible :)

Until next week,

-Bar and Hannah