Yeah, I was thinking that, but I wanted to keep the extra uC optional…
Plus, if the extra uC is going to be mandatory, there’s not much reason not to use RS485.
Oh, read-up on 1-wire a little more, it’s possible to implement 1-wire on a UART using a circuit like that one, but the process used to emulate the 1-wire protocol with that can only be described as byte-banging.
What is wrong with byte banging? You still need to read in a byte at a time in the interrupt.
If a board has components on it then the incremental cost of adding a uC is very small.
As for RS485 I have never used it so I can’t comment on it. I will ask around though.
Trying to reduce the potential for cross-talk. As @F_Malpartida said, I don’t like having PWM_FAN adjacent to the SCL pins, but it has to go somewhere. Think it would be better to interleave the thermistor pins with I2C? Then I could throw PWM_FAN next to HEATER and use GND and PROBE as a buffer. Does that sound like a good arrangement?
I think it would, the i2c even though hw driven the chances of having a simultaneous transmission while doing a conversion are much smaller and will get filtered out. If I remember correctly, for the time being adc conversions are not interrupt driven and will most likely happen in sequence (unless there are long block transfers, which is not likely for regular devices).
Why is there 3 supply and 1 return ground. Your full loop should support equal current forwards and reverse.
That’s a very good point, but I’m guessing that the load switching is going to happen at the controller side. That way with the +5V getting to the pins would provide a balanced config. Also the fan pwm would oblige to hace a small current fet on the breakout.
There is a 4 pin block of pins for “HEATER”, which is switched by an N-channel MOSFET, the PWM_FAN pin, which is intended to be switched by a lower-power version of the same, and one GND pin for 5 pins of supply voltage. If those were all being used at maximum capacity at once, one more voltage supply pin would be needed to match the capacity, but that should never happen (none of them should reach maximum capacity at all, and the heater’s duty cycle should be reduced for holding temperature before the fan is turned on), so I figure it’s pretty balanced.
Ok, new version that I think solves all the problems here: https://plus.google.com/105535247347788377245/posts/6gPeQmmvWCC