Works as advertised, would benefit most from additional docs
For AC 60Hz/50Hz Dimmer/SSR Controller Board
When I decided to replace/redesign a light alarm I made in 2008, I bought this board rather than designing my own phase-control dimmer board. So far I've been chasing more problems in my software than in this board.
The general design of the zero-cross input and the triac output seem broadly similar to this schematic, though not all details are the same: http://playground.arduino.cc/main/ACPhaseControl
The details about how to control the board seem a little bit sparse. Here are some notes, based on driving a "rev 1.3" board with a Raspberry PI 2:
* I used a 5V supply and 3.3V IOs, which works fine
* The "PWM" control mode is based on measuring the *length* of the pulses rather than the duty cycle (arduino pulseIn). The documented "5%-40%" figure is appropriate for an Arduino with a 490Hz PWM (http://arduino.cc/en/Reference/analogWrite)I used a 1kHz PWM, which caused the effective range to be about 10% to 80% instead. This also seems to cause the light to turn off with a "100%" input, which I think is also a limitation of pulseIn. So I think the PWM input might better be described as 100-800us "high" time plus a minimum "low" time (200us is adequate, less is probably OK). (this bullet point updated after receiving Arduino source for the board's firmware).
* I drove the "mode" pin from another GPIO. That GPIO floats until my application on the PI is totally initialized. (I don't have a good reason for this, it's just how it turned out) To avoid turning on the light at this point because of a voltage seen on the (undriven) CTRL input pin, I tied CTRL to GND. (I see now that the product page suggests a 65k resistor here)
* Incandescent and halogen bulbs still dim better than any of the CFs or LEDs I tested
* On request, the seller supplied me with the source in .ino (arduino) format and the schematic. I haven't yet built it or flashed the device, but I may—I intend to find out whether serial or spi input is feasible.
* The LED has 3 states: off (no AC commanded), bright (100% AC commanded, or no AC zero-cross detected) and dim (somewhere in between).
Some suggestions for a future revision of the board would be:
* rotate the opto chips 90 degrees, so that they straddle the dashed line that implies isolation of the low-voltage pins and traces from the high-voltage side.
* if possible with your fab, use an internal slot under the optos to improve isolation
* document the permissible range of PWM frequencies
* Consider including an RC network (possibly jumpered?) at the CTRL input pin so that PWM can be applied at CTRL, and document the minimum PWM frequency to avoid eccessive ripple
* Or at least add a pull-down at the CTRL input pin so "no analog input" leaves the lamp off
* provide an accurate rating for permissible current through the triac Q1 (250W halogen led to an unreasonbly hot heatsink but no apparent damage)
* A fuse or PTC would be another great addition in case of mishap. I need to add an inline fuse to my setup.
Response from RF Simplicity | April 5, 2015
Thanks for the review Jeff, I agree with all of the suggestions. I will implement these with the next revision.
