Talk:Aidan Williamson Log
Marcin on sensors September 2019
MJ Comment: I would not complicate the system unless you can show repeatable data that the current algorithm is inadequate. Simplicity is genius. Adding complications is easy. We have shown already that this works without failure for thousands of bricks for Microhouse 1, 2, and 4. There were no rejects here due to improper timing. Self-calibration occurs every brick cycle. Pursuing additional sensors is in my view going backwards instead of continuing to refine and simplify.
Reply October 2019
I don't like the way the drawer hydraulics spasm every cycle. Also it's slower to use the pressure switch. A limit switch system on the drawer, like Rob Beddingfield uses, makes the machine quieter, safer, leak less oil, and makes it faster since. It's faster because you save the time it takes to build pressure at the end of a stroke and because the stroke length can be shortened about an inch.
I like the timing based on fluid volumes approach. I fully understand why it's useful and I'm not going to put sensors on the main cylinder since the timing method there works. It will continue to be used. I did change the code so that we are not using delays and simultaneous operations can occur, such as LEDs or even a LCD screen downstream.
I am only including a mid position sensor so that the user can press a button to bring the machine to the mid-point from anywhere. Otherwise I would just use two drawer limit switches.
Aidan Williamson (talk) 22:07, 9 October 2019 (UTC)