Any difference in response

[Hood]

I was wondering if anyone would know if there would be a difference in response time seen on an input depending on where the LED was in the circuits as below.

The Input to the CNC control is sourcing so it outputs 24v and connects to 0v. When the connection is broken the CNC sees that as an input.

I am just wondering if the LED being at the start of the circuit may introduce a slight delay due to the back feed of the diode across the LED when the circuit is broken at one of the switches?

 

[Seadog]

I can't see that it makes any difference.

 

[voipio]

It doesn't matter where the LED is in this circuit, as long as it's not across any of the switches. It will need a current limiting resistor, though. Not sure what the other diode is doing. Is it supposed to be protecting the LED against some sort of reverse polarity?

 

[Hood]

The Input limits the current as far as I am aware.
I think the diode may be to protect the contacts of the switches, having now said that I am wondering if the LED and diode do need to be as per the first sketch.

 

[mike os]

Nothing you will notice

 

[Hood]

Nothing you will notice

Ah but it is not me that needs to notice, it is the CNC control where a few mS can make all the difference to accuracy

 

[Hopefuldave]

Unlikely to make a difference re delay, an LED will switch in a few nanoseconds (they're used as fibre optic transmitters to launch gigabit/second signals) - what may make a difference is the current it's passing (probably need to keep it below 20 mA or it'll play at being a fuse) and the voltage drop, between 1.5 and 3.3v depending on which colour.
The anti-parallel diode is there because LEDs can't stand much reverse voltage, so it's there to conduct at 0.6v reverse and limit the reverse voltage on the LED.
So... does the controller specify things like minimum logic level (it's 10v on most ABB stuff), the input impedance you're feeding it into, the max current the 24v source can supply? If the input resistance is too high the LED may not light well and just give a glimmer. It might be better to put the LED from the controller input to 0 volts, so switch closure puts 24v on it, and put a current limiting resistor in series with it.

If the input (labelled 24v) is active low (want the LED lit when all the switches are closed and the input's at 24v inactive, 0v active) the LED can go between the input and a 24v (or less) source, closing the switches will put both the input and the -ve end of the LED at 0v

Dave H. (the other one)

 

[mtt.tr]

Unlikely to make a difference re delay, an LED will switch in a few nanoseconds (they're used as fibre optic transmitters to launch gigabit/second signals) - what may make a difference is the current it's passing (probably need to keep it below 20 mA or it'll play at being a fuse) and the voltage drop, between 1.5 and 3.3v depending on which colour.
The anti-parallel diode is there because LEDs can't stand much reverse voltage, so it's there to conduct at 0.6v reverse and limit the reverse voltage on the LED.
So... does the controller specify things like minimum logic level (it's 10v on most ABB stuff), the input impedance you're feeding it into, the max current the 24v source can supply? If the input resistance is too high the LED may not light well and just give a glimmer. It might be better to put the LED from the controller input to 0 volts, so switch closure puts 24v on it, and put a current limiting resistor in series with it.

If the input (labelled 24v) is active low (want the LED lit when all the switches are closed and the input's at 24v inactive, 0v active) the LED can go between the input and a 24v (or less) source, closing the switches will put both the input and the -ve end of the LED at 0v

Dave H. (the other one)

Perfect write up

 

[Hood]

Unlikely to make a difference re delay, an LED will switch in a few nanoseconds (they're used as fibre optic transmitters to launch gigabit/second signals) - what may make a difference is the current it's passing (probably need to keep it below 20 mA or it'll play at being a fuse) and the voltage drop, between 1.5 and 3.3v depending on which colour.
The anti-parallel diode is there because LEDs can't stand much reverse voltage, so it's there to conduct at 0.6v reverse and limit the reverse voltage on the LED.
So... does the controller specify things like minimum logic level (it's 10v on most ABB stuff), the input impedance you're feeding it into, the max current the 24v source can supply? If the input resistance is too high the LED may not light well and just give a glimmer. It might be better to put the LED from the controller input to 0 volts, so switch closure puts 24v on it, and put a current limiting resistor in series with it.

If the input (labelled 24v) is active low (want the LED lit when all the switches are closed and the input's at 24v inactive, 0v active) the LED can go between the input and a 24v (or less) source, closing the switches will put both the input and the -ve end of the LED at 0v

Dave H. (the other one)

Thanks, sort of understand most of that
I already have a tool sensor on the lathe and I wired it as per the first sketch and it has been working fine for months now so it would seem the control does the current limiting.
Specifics about the I/O on the control are minimal, I am sure I saw mention of currents in the manual but I can't find them at all now.
Only thing I can find is the following.

 

[minimutly]

Well the question was a simple one - will the circuit change make the led go off sooner or later?
No, it wont.
Simple answer...

 

[Hood]

Well the question was a simple one - will the circuit change make the led go off sooner or later?
No, it wont.
Simple answer...

Well more or less, it was actually will the input be seen by the control sooner or later or no difference and it seems from all replies it is no difference, so thanks all

 

[mike os]

Ah but it is not me that needs to notice, it is the CNC control where a few mS can make all the difference to accuracy

Nothing it will notice either.... how fast does it flow?

 

[Agroshield]

Well more or less, it was actually will the input be seen by the control sooner or later or no difference and it seems from all replies it is no difference, so thanks all

I went looking for 'high speed switching' and got lost in motherboard nanosecond stuff, and then particle accelerators, but one thing I did see were oscilloscope traces. If you have a fast scope, you could try both configurations and see which gives the best rise time.

And just to throw in a real curve ball, your 'start of circuit' is predicated on the pixies running from plus to minus, whereas they really go minus to plus.

 

[Hood]

Have a scope but probably not fast enough, might try just to see but swapping around the LED is a pain so probably won't bother doing that.

 

[Hopefuldave]

Nothing it will notice either.... how fast does it flow?

About 5 nanoseconds per metre for the cable, a couple of nanoseconds (if that) for the LED...

Dave H. (the other one)

 

[minimutly]

It's not water, the led will extinguish just as fast nomatter which end the switch(es) are...

 

[Hood]

It's not water, the led will extinguish just as fast nomatter which end the switch(es) are...

I realise that but it was not how fast the LED extinguished that I was concerned about but how fast the Input would be seen by the control.
I was wondering whether there would be a slight delay due to the diode across the LED but it seems that is of no concern from all replies

 

[voipio]

Not a concern at all. The switch opening and wire length and circuit capacitance, together with pull-up resistance, will determine how fast the voltage will rise across the "0V" and "24V" terminals when the switch opens. The rise time will be in the nanosecond time range. If the CNC moves at 1 metre/second, that will be a few nanometres. The intertia and braking of the axis positioners will be the biggest factor.

 

[Hood]

The axes are travelling much slower than that anyway, well whilst doing the tool sensing operation.

 

[m_c]

In practice, it'll make no discernible difference.


But it reminds me off a question a mate asked me years ago, regarding a pressure relief valve on the sea bed.
IIRC it was something like what pressure would a 200bar relief valve vent at to surface pressure, when submerged to a depth where sea pressure is 50bar.

I thought it was a trick question at first, because the obvious answer was, as it was surface pressure vented, it would still blow at 200bar. The sea pressure didn't affect the valve, as it was totally external, and the pressure differential was still 200 bar.
Turned out earlier that week, somebody within in his team of engineers had asked the question, and it had taken an entire morning involving multiple engineers from various suppliers to come back with the same answer I gave in about 30 seconds. He worked out that the answer had cost him and his suppliers a couple thousand pounds to answer what was a basic physics question, and was asking random people to see just how stupid they'd all been in not being able to answer the question themselves.