MK1 Steel router build log 1500x800x200

[Doddy]

Aha, the same style that I have also.

Inputs:
[I'll correct the 101 mistakes in the following para, below... ignore the stuff in italics - included as a reminder for the crap I typed this morning]
So, that addresses the onto-isolated inputs. However, the design of that board presents the cathode of the onto-isolator via a 1k resistor to the input pin. The associated anode is wired to the on-board +5V supply. So, shorting the input to ground will activate the otto-isolator. Similarly, having an NPN drive to the pin *should* activate the opto-isolator, however, you have to consider a couple of issues: With the NPN drive OFF, you need a pull-up (or the NPN sensor must have a pull-up). If you intend to use 24V signalling this means that, with the sensor off, and the sensor output pulled to 24V, that you're reverse-biasing the LED fragment in the opto-isolator. My board has EL817 onto-isolators, which have a typical maximum reverse voltage of 6V. Assuming that your 24V and 5V supplies have a common ground that gives you a reverse-bias of 19V across the LED, which exceeds the data sheet value substantially.

5V switching (or 6V if that's the lowest supported by the sensor) is completely do-able.Above 11V is giving you problems. There are ways around all this - let me know if you want to investigate these options.


So, that addresses the opto-isolated inputs. However, the design of that board presents the cathode of the opto-isolator via a 1k resistor to the respective input pin. The associated anode is wired to the on-board regulated 10V supply used for the PWM output (and fed from the 12-24V input). So, shorting the input to ground will activate the opto-isolator. Similarly, having an NPN drive to the pin *should* activate the opto-isolator, however, if the sensor has a pull-up (or you've added a pull-up) you have to consider a one issue: With the NPN drive OFF and with a pull-up resistor and if you intend to use 24V signalling this means that, with the sensor off, and the sensor output pulled to 24V, that you're reverse-biasing the LED fragment in the opto-isolator. My board has Liteon LTV-817B opto-isolators, which have a maximum reverse voltage of 6V. Assuming that your 24V supply for the sensor is the same as the feed into the BoB, or otherwise have a common ground that gives you a reverse-bias of 14V across the LED, which exceeds the data sheet value substantially.

This is only an issue if you have a pull-up as part of the design (or part of the sensor). If not, then it's not an issue, but be aware although you're driving the BoB at 12-24V, the actual switching is regulated to 10V. Don't inject 24V into the inputs of this board (worst case scenario: you'll fry the opto-isolator, and possibly the onboard regulator - but protect the UCx00 controller).

It also means that the inputs are dependent on the 12-24V supply, even if you don't intend to use the PWM output. The logic on the board is dependent on the 5V supply, as are the stepper motor outputs.



Outputs:
Just remember the resistor-bank that you asked about - your drive to the stepper drivers is still 5V signalling and requires no additional resistors for current limiting. You need to source a 5V supply for the BoB, as well.



The more that I look at BoBs, the more I'm inclined to design my own extension BoB boards for the UCx00 range of controllers that give complete galvanic isolation to the input circuitry.


EDIT: Since the visual inspection, I've now metered the board and I'm happy with the info above.

I'm in the same position of trying to understand the reliable interfacing to a UCx00, although my own apathy is slowing my build.