Noise is presenting as spikes in analog sensor input values during setpoint modulation - switching - and occasional chirping through the 12v supply - I will refer to your diagram on this one.
I think the connection DGND to GND is necessary, but could that also be through an inductor?
On the mosfet gate side, I'm thinking inductor in series with the current limiting 100R.
Simple enough with an on-line aircore inductor calculator to wind my own.
I've used that smoothing algorithm before, but I think first, sort out the hardware. As you rightly point out, my application is flawed in places.
Attached with proposed changes/queries.
EDIT: had an hour spare so whipped up a 100nh inductor to replace the resistor. Possibly my imagination, but temperature control is more consistent - that is; less pwm hunting as it endeavours to push the temperature up or down.
Chirping is however, more consistent - resonance in the circuit I guess.
This arrangement is observably better. Setpoint regulation is smooth, with very little modulation. Setpoint LED is steady ON. I guess temp inputs are less erratic - gradual improvement with each modification.
EDIT: Suggested modifications to PCB. New images
Last edited by rcheshire; 08-04-2014 at 03:03 PM.
Reason: New images
This is all in one place now and while there is room for improvement, there is only so much time in the day. I hope to close this now and move on to other ideas. I wish to thank the people who contributed and educated me along the way. I hope this is useful to someone in part, at least.
Incorporated recent changes, except for the PCB etching pdf. The board can be modified easily, pre or post etching with reference to the attached image - no opportunity, otherwise, in due course.
Notes and drawings need updating - time permitting...
