Hi,
Passionate woodworker, I work on a small project to automate the dust collector. The PCB I'm designing will be located at each tool, in series with the power cord. The goal is to "measure" current flow, then trig two signals to open a blastgate and stat/stop the dust collector. I made a proto on a breadboard and it works just fine. I started laying out and routing my components. 45mm x 68mm (It has two fit inside a single gang wall outlet).
The signal gets picked up by a 30A current transformer, goes to a precision rectifier using an op-amp then to a PIC16F18426 for current levels analysis. Then 2 outputs will drive 2 relays using 2 single NPN transistors.
My main concern is about my layout not really emitting EMI but more being susceptible on picking some heavy EMF from the mains power supply of the tool at about 12mm from the top of the PCB.
I will go with a 4 layer design having top and bottom as signal and layer 2 and 3 as ground. By looking at the pictures attached, does anyone sees a monstrous evidence of a bad design here ?
Let me be very clear : It is not meant to go on the moon. We're not talking about high precision triple digit resolution current reading. I'm only worried on strong signals getting picked up by the board and creating such a disturbance that it would lead to erratic behavior. As for en example, the high inrush current of starting a 4HP motor on 240V.
All advices are really appreciated.
Thanks you so much
I'm sorry. Perhaps a list that is too offtopic, but it can be useful for those looking for how to improve their layouts.
...
My main concern is about my layout not really emitting EMI but more being susceptible on picking some heavy EMF from the mains power supply of the tool at about 12mm from the top of the PCB.
...
One thing that comes to mind is that fast moving dust (and other particulates) is very good at generating ESD discharges, as I'm sure you know. Be very careful of your ground routing in the system as a whole, so that there is no danger of ESD currents passing through the board ground. Extra filtering of inputs and outputs is probably advisable too, especially as you have a microcontroller on the board.
Unrelated to the actual circuit, but one potential problem is your bottom left mounting hole may risk a short if its two sides are at different net. Make it a pad instead, with pad size large enough to cover the fastener head. There's also question whether you want the fastener to be isolated or for it to connect the ground.
It sounds like a bit of overengineering to me. Why do you need a PIC at all? Maybe no need for the current transformer, either.
If I were to design something like this, I'd probably use a simple wire shunt followed by a differential (or instrumentation) amplifier followed by a rectifier, an RC filter, and a comparator (with hysteresis), or, if the voltage drop on the shunt is sufficient, only filter+comparator. No need for a precision rectifier then, either (as long as you have, say, at least a volt or two across the shunt), just clip the negative half of the wave using a Schottky diode from ground to shunt output, and adjust the RC filter accordingly to have a decent level on the output suitable to feed into the comparator.
As far as EMI-induced false positives, it won't be an issue with the shunt approach, at least if you keep the signal traces short. Probably not even with the CT. What voltages are you getting on the CT's output?