It seems like we start off every newsletter saying that it’s been a big couple of weeks, but once again it's been a big couple of weeks for us!
Since the last newsletter we’ve been preparing for the next batch of kits to go out, we delivered the CNC-made tiny house to its future occupant, community contributions to the open source software hit an all time high with the most notable being a major improvement to the calibration process, and someone built a boat!
February Kits Update:
There is no exciting news to report on the February kits which in itself is exciting news! We have been pushing to get everything to us on time and as long as there are no unforeseen delays we should have all the parts and be ready to start shipping near the end of February. We know you guys are excited to get them and we’re excited to send them out to you. We’ve gotten a number of emails asking for updates so we’ll be sure to do everything we can to ship them as soon as possible and we’ll keep you updated as we get closer.
Improved Calibration System:
There has been a TON of community collaboration going into the software this week. Between the firmware and Ground Control we’ve had 36 separate proposed changes merged by the community. I won’t go into all of them here, but I would like to highlight one in particular as an example of open source at it’s best. This is a summarized version of the conversation and collaboration effort, the full conversation can be read here and here.
This week community member reecej made a HUGE contribution by rewriting the way the triangular kinematics calibration process works so that now it will detect and take into account the sag in the chain. Previously we were assuming that the chain made a perfectly straight line while in reality the chain will have a slight arc to it from the force of gravity acting on the chain.
The new calibration process cuts five marks and asks for a series of measurements to be taken. It is then able to compute the rotation radius and chain sag from those marks. The process concludes in a single pass rather than asking for multiple rounds of cuts and measurements to be taken.
Because this is such a big change to the way the calibration process works, we wanted to be sure that the new system was better than the old system before adopting it. To compare the two we needed to come up with a benchmark test for how well calibrated a system is. We came up with this test pattern:
The calibration benchmark test consists of three long horizontal measurements, three long vertical measurements, and the measurements of four squares. The long horizontal and vertical measurements test for distortion in large parts cut either horizontally or vertically. The four squares test for accuracy in the far corners of the sheet where accuracy is likely to be the worst. To easily compare different tests we wanted to boil the results down to a number. Ultimately we decided on a two number result where the first number is the average absolute value of the error in the long measurements and the second number is the average absolute value of the error in the squares. A result of 5.43-2.10 would indicate that the long were off by an average of 5.43mm while the squares were off by an average of 2.10mm. For comparing calibration systems the first number is the important one.
Testing the v1.02 version of the calibration system to get a baseline, we found this calibration pattern giving us a score of 6.00-1.43 meaning that long parts would be off by an average of 6mm.
After working out some issues with the way the new calibration system works we got to the result below which shows a score of 1.00-1.31...a six fold improvement in the accuracy of large parts.
With a new calibration system and a new way to benchmark accuracy I would expect more improvements in the future. If you would like to run the test yourself to measure your machine’s current accuracy you can access it at any time by clicking “Actions -> Advanced -> Load Calibration Benchmark Test”.
One of the limiting factors to the accuracy of Maslow that we discovered this week is flex in the frame itself. My frame is of the stock design and it has a measurable flex when the sled is near the top center of the work area. The official frame design has not changed significantly since the Kickstarter while nearly every other aspect of the machine has been improved. The forums are full of examples of excellent alternate designs. During this coming week we plan to work on a new design for the official frame which pulls the best elements of all of the designs we’ve seen. If you’ve built a custom frame head over to the forums so we can learn from your experience.
On Saturday we delivered the PlyPAD to its permanent home at the Kenton Women's Village. The reception there was overwhelming, and it felt really good to finally get to deliver that project. Overall the delivery went really well. The sun came out, which is unheard of for January in Portland, and the sections which we struggled to get to bolt together in our sloped and uneven parking lot went together easily. We would like to say a huge thank you to The Rebuilding Center for letting us use their truck and forklift to move the pod. We never could have done it without them.
As people are becoming more comfortable with their machines we’re seeing more ambitious projects. This week we saw two amazing project in the forums.
First Brr built this awesome ipad/notebook stand for indoor cycling.
Just as cool as the stand itself are some of the modifications that he made to his machine.
In what must be the most amazing thing built by Maslow to date, Sonny_Lacey built a boat! (Or at least a very solid start on one). The design is for a Bolger Bobcat sailboat based off of purchased plans. The plans were then digitized using Fusion 360 and cut. Sonny_Lacey has some excellent advice on that process in his forum thread here.
We can’t wait to keep following along with Sonny_Lacey on this project. When you build amazing things (even if they are less amazing than this boat), don’t forget to show them off in the forums.
Have a great two weeks everyone!
-Bar and Hannah