richö butts

I think I'm done with the hardware mods for this part of the project. I spent today mostly rewriting the firmware and I'm really happy with how this ended up.

Next I'm contemplating a float switch to refill the res automatically, and gently mulling over my options to control brew pressure

Mar 7, 2023 at 8:01amIvory for iOS
Show threadGraham Sutherland / PolynomialMar 7, 2023
@richo something I've seen mentioned with electronically controlling pumps is to drive them at 100% initially to build up the static pressure, then ease off (using a motor driver) to drop the pressure to the desired level. it seems like most pumps struggle to get going if you don't give them 0.5s+ of initial on time.
Show threadrichö buttsMar 7, 2023
@gsuberland Oh interesting. Right now I can only turn it on or off, no way to modulate it. My profile only runs for .2, then .3, then .4 with a short delay and seems to be ok but this is great info. Profile is here if you're curious https://github.com/richo/things/blob/main/silvia-controller/src/brews/richo.rs#L18
things/richo.rs at main · richo/things

3d printed stuff I made. Contribute to richo/things development by creating an account on GitHub.

GitHub
Show threadrichö buttsMar 7, 2023
@gsuberland I'm interested in a way to modulate pressure at the end of the shot to taper it, I'm contemplating a needle valve with a stepper motor
Show threadGraham Sutherland / PolynomialMar 7, 2023

@richo that's a pretty good option - you'll have tighter pressure control for sure.

a digital pressure sensor might be worth investigating. I bet there are tons of them designed for boilers. with actual pressure readouts you should be able to turn the whole control scheme into a PID with a moving setpoint to match your profile.

Show threadrichö buttsMar 7, 2023
@gsuberland that's kinda what I'm hoping for. I will probably do some spread sheeting to figure out how much it's actually worth to me, but I think I'm willing to invest a little more in it. I'm borrowing a pressure gauge from a friend soon so I can at least test it what I'm doing now yields consistent ramp up.