Hi,
2 i am using individual limit switches on each axis
can you not therefore identify which limit switch is activating? This would cut down on the amount of work trying to find it. You
may find for instance that is the Z+ limit switch which just happens to be in the same cable chain as the spindle motor wires
and if you separated them the problem would go away.
I have read somebody used capacitors on theres but they didnt explain how to or what to use really,...any info on this method would be appreciated
Small capacitors (0.01-0.1 nF) between the 57CNC input pin and 0V add another layer of filtering. The software filter provided by the 57CNC does a very
similar job but there are a few technical and very minor differences between a software filter and a physical RC filter using a capacitor. Given that capacitors
are so cheap its worth trying.
Most people when confronted with an electrical noise issue do as you have done, shielded cables, passive and programmed filtering etc.
It is in fact the wrong place to start. You should try and identify the device or devices that are producing the electrical hash and quiet those BEFORE trying
to improve noise immunity of things like limit switch circuits.
Usually its the spindle motor which produces the most noise in any CNC machine. May I suggest try running some code but without the spindle running.
If the fault goes away then you'll know that the spindle is responsible for the noise. You might find that a line reactor is required....not many hobbyists
bother with line reactors and as a consequence hobbyists experience a great deal more trouble than well designed commercial equipment which does.
You may have to put EMC filters on the output of your VFD......be warned follow the VFD manufacturers instructions to the letter if you do so, just any old
EMC filter is likely to result in clouds of smoke.
The next step after ensuring that all practical steps have been taken to reduce the noise output of the offending device (spindle/servo drive etc)
is to separate the wiring between the high energy circuits, especially the identified noisy ones, from the low energy sensitive ones like limits switches.
Another tip is where a low energy circuit crosses a high energy circuit have the cables cross at right angles to each other, it reduces the magnetic coupling
between the two.
The next thing to consider 'is a particular input circuit sensitive because of the very low current in the circuit'. If you have a high impedance limit switch
for instance then very low currents will exist in the limit switch circuit and as a consequence it will be easily swamped by a noise signal. If however
the limit switch has a moderate standing current, say 5-10mA, then a much stronger noise signal would be required to swamp it.
Then you start with shielded cables and filters. Its a bit like 'trying to shut the gate after the horse has bolted' though. Try and reduce the noise at source
THEN reduce the sensitivity to the noise.
Craig
Guest
