V1 was never ironed out but lets start fresh with a V2.
Previously we had a bottle with a washer and lathe motor mounted as an agitator. The output was inverted which had the precipitate fall into the output hose which caused issues as it attached itself to the walls of the hose over time.
For whatever reason we didn’t consider using a magnetic stirrer. There are dozens of these things available for $40 on Amazon. They all have the same specs. 3L and 2000 rpm I believe. So we are going to simplify this whole process. This new approach will be about the same in price, perhaps slightly more. It will be far simpler. Less noisy. It will also remove the whole issue of the clogged output hose. So this should be a win all around. I will start the documentation in this thread.
The source of this post are my notes where the V1 was left off. The extruder motor had been turned into a rinser but that whole thing is likely no longer an issue. So that will be removed. I was going to use strike-thru to remove lines from this but it doesn’t exist.
There is that 4th pump. It is the extruder pump.
X,Y,Z each have their own calibration factor but I never did it for the extruder.
There are 2 zip files that one downloads. One made by the php and the other one has all the static manual created zip files.
Currently there are scripts for all the x,y,z but the extruder version has not been created. I did modify alldirtest.gcode but none is tested.
I have been zeroing all the origin and then I do a move where the Z is set to 0 but also there is an extruder amount. This has not been tested. example - G1 Z0 EEXTRUDERWASH FPUMPSPEED
basemix.txt is what has the gcode pre variable substitution.
Before this extruder stuff was added - it seemed that the 2 ingredients were not being pumped at the proper rates. The 2 inputs should wind up empty at the same time.
There are also many ways that this all can be calculated via what is chosen as being the constraints. Those are
Output Bucket Size Ingredient amounts No optimizing I belive output bucket size is the default that works. I have it in a screen session. - The website was meant to optimize on the various containers. Whether the constraint is the input or the output. For now we just optimize on the output bucket as that is the sanest.
Rumor has it the URL of this site has been given out to others so I’ll put a slight update.
This has been part time project for me with other things calling me. Going through various design issues. I’m no longer sure I want to open source the whole thing without trying to start a business. If I start a business - then I am screwing myself and my limited capital by giving away plans for a competitor. I’m not sure a community would ever form around this like I had hoped. I would definitely open source it after a period of time but I at least I want my investment back. With tarriffs and inflation nothing is getting cheaper.
Regardless - the short of it is I have lost my desire to document everything. Just trying to keep the project moving forward. We are into year 2 now.
We went back to a stirring method using a basic magnetic stirrer. I don’t think this has enough power in the stirring to do the smaller particles. We are likely going back to a top powered method utilizing smaller electric motors. Our attempt as simplification failed. We might have 2 separate processes to make the larger vs smaller particles. Which might mean we continue to utilize the magnetic stirrer.
Our immediate step is to make some samples of the paint using the known process and then applying this to a variety of surfaces to see how well it holds up over the course of a year and beyond. I keep learning more about the process as I churn through my learning curves and such. We will get this done though. There are other projects I find far more compelling but without the benefit to humanity of this one.
