Thanks. So what frequencies are we talking about for signal vs noise? I’ve seen mentions of a 500Hz cutoff, and these values which yield, if my math is correct, a almost 160kHz cutoff.
Also, what sort of response times are ideal for something like a limit switch or probe?
Thanks again!
I see that you have 3 cooling fans in your cabinet design, did you find that you actually needed 3 fans? How hot do the drivers get in practice?
I’m in the process of designing up a version of your cabinet and have gone for 80mm 33CFM fans. What size did you end up using?
Thanks.
1 Like
Yes. Gives you a 6.25 uSec cutoff. Anything faster than 6.25 uSec would be attenuated.
Ideally, you would want an instantaneous response to a switch closure (assuming NO switches).
As to the requirement for a pulse on the wire to be considered as noise, that depends on the application. You don’t want to penalize switches that turn on fast, but you want to eliminate parasitic pulses on the wire.
It’s a trade off.
1 Like
So I went overkill? Cuz there is enough airflow through the fins that I was gonna use them as main air intakes possibly. But I just got my hands on some juicy 110v case fans that may blow my mind once I plug them in so we shall see.
Would NC switches be better?
Do the fins get hot during long carves?
You are a special case. 
I’ve never encountered another situation where there was a noise problem worse than yours.
1 Like
He is a big man. Usually they are louder. 
Even when soft spoken. 
But still a brutal Typer. Hmmmm now u got me wondering. Something about carrying big sticks.
1 Like
There are trade offs. Each configuration has it’s strengths and weakness. Pick your poison.
What are the side effects of each poison?
I use the TB6560´s which have smaller and lighter cooling fins, they get warm/hot during long runs but dont struggle.
With larger fins and active airflow I would´nt worry 
Just make sure there is open vent on the other side
1 Like
Just a couple of points.
Normally Open
Good:
does not draw any current when inactive
does not stress the return spring when inactive
Bad:
Contact bounce when activated
wire to switch can be broken which makes the switch appear to be open even when the switch is closed
Normally Closed
Good:
No contact bounce when activated
broken wire will appear as an open switch
Bad:
Draws current when inactive
Stresses the return spring when inactive
2 Likes
Phil, I corrected an auto-correct in previous post, it was meant to read “would´nt” lol…! Sorry 
2 Likes
Yes.
Which design are you going to use?
Phil’s design shown on the Arduino picture, or the Inventables one?
No idea. Whatever works. Reliable. Or easier.
Use the Inventables one then.
Note: you need a 0.01 uF capacitor and a 100 ohm resistor for each homing switch and the probe input.
Phil had really bad noise so we had to use the big hammer at the expensive of responsiveness.
2 Likes
Got it. Thanks. Does the capacitor have a handing?
Some do and some don’t. I would recommend ceramic capacitors for this application. They are not polarized.
Most electrolytic capacitors are polarized and if you hook them up backwards, they tend to explode.
1 Like
What would be the largest sources of “noise” in the system?
Any order?
VFD?
Power supply?
Spindle?
Dust extractor?
All of the above can produce interference. If you get a good power supply it is probably the least worry.
Don’t forget about static electricity on dust collector hoses. The plastic ones make really good Van DeGrafff generators (20 to 30 thousand volts, easy).
1 Like