20m cable run switch input to 08M

[Eclectica]

Hi,

I need to monitor the spare contact on a Finder DP relay but the other contact is switching a 240V contactor coil. I will use 20M of 4C screened cable with the screen connected to mains protective earth. The relay coil is switched by 24Vac.

Prospective circuit diagram based around optoisolator attached. I always feel uneasy about long cable runs going direct into a PICAXE (or even short ones where any inductive switching occurs nearby), but am just wondering if this is the most effective approach?

[The PICAXE at the other end of the 20M cable run is powered from a switchmode PSU to which I have added mains filter and 375V 220n X2 capacitor. 5V is then fed through another filter and transorb.]

Any suggestions welcome - sensible of course .

Regards
E.

 

[BeanieBots]

As both sides of the opto are connected to the same supply, it isn't really doing anything for you.
Either, drive the opto from the relay side, (maybe from the 24v coil supply) or simply remove the opto.
Your cable is screened so there should be very little in the way of noise.
Using a low(ish) pull-down resistor (~1k) will also help.
A 22k resistor in series with the PICAXE input will protect it from any nasties that might be picked up. A 100nF cap (that you have already used) will also help.

 

[Grogster]

I have used PICAXE inputs directly connected without the opto's, and although I did have some problem tripping in the first prototype, once I added 100n caps across the 10k to deck on the input, all the problems vanished. This was suggested by another member here, and it worked like a charm. I fit them by default now on pretty much every design, just to be on the safe side. Have a double-check routine(or more) to confirm the contacts are indeed closed and it is not just a spike or other "Nastie", as BeanieBots calls them.

In my case the cable-runs were about the same: 20M or so, and I was not using shielded cable in the prototype, but do use shielded cable now...

 

[westaust55]

The screen itself will add capacitive effects between the active cores and the screen.

Suggest that you:
1. only earth the screen at one end to prevent earth loop currents in the screen which in turn generate noise into the active cores
2. tape up the ends for safely but do not earth the unused cables (if you local electrical regulations permit this since one end is in a panel with 240 Volts). If the unused cores are earthed they for a further capacitance to the active cores.
3. Used a twisted pair cable such as CAT5 if you can.

I have done some experiemental work with up to 40 metres of screened two core cable for 1-Wire networks to a PICAXE without problems.
I have used 100 metres of CAT5 for a 1-Wire network with slightly better results (may use a higher value pully up resistor) than with just 40 metres of the screened cable

 

[RobertN]

Suggest the spare contact operate a secondary 24VAC relay placed well away from the main contactor for isolation from high voltage, transients, and any flash over. On the Picaxe end, use a R - C low pass filter with a roll off well below 60Hz. Follow the filter with a 4.7 or 5V Zener, or TSV to handle any DC or AC components that gets through the filter.

 

[lanternfish]

Have the relay switch the 0V of the LED. The 360R resistor connects to +5V. This way your 5V supply is well seperated from the 240V.

 

[Eclectica]

Thanks everyone for the responses - much appreciated.

Based on this I will go for circuit attached.

I will probably also run the contact pair through small bead inductors at the PICAXE end to further reduce any common mode transients.

Most of the cable run will be away from power cables except for a shortish 2m run next to a steel wire armoured running a 3phase 5kw pump - hence my obsession over induced transients!

[I was trying to get away without using another interface relay at £9 a shot as the 24Vac panel transformer supply is being used (worst case) beyond 80% of capacity.]

Thanks again.

Regards
E.

 

[hippy]

I'd consider something 'analogue' ...

                 ___
            .---|___|---- 5V    10K
            |
   .--------{
   |       .|.
   O /     | |                  22K
   O/      |_|
   |        |
   `--------{    ___
            }---|___|---> ADC    1K
            |    ___
            `---|___|---- 0V    10K

 

[Eclectica]

Hmmm... not a bad idea, that way the power rails are better protected I think.

Thanks.