Need help with long distance circuit

[slick-rides]

Hello and thanks in advance for any help.

I'm trying to hook up an alarm input on a PA system. It's just an open contact input that sounds an alarm when closed. My problem is I'm trying to use a float switch that is 1400 feet away. I believe the long distance is causing it to malfunction. I've tried to run wires directly from the alarm input contact to the float, but it just alarms right away. I've tried to use a dry contact relay, but same result. I then tried to use a 110v relay which will work for up to a week at times, but it occasionally malfunctions or energizes the coil when it shouldn't. Looking for suggestions on how I can get this to work. I attached some amateur drawings so show what I've tried so far.

Also, here's a link to the dry contact relay I tried... http://www.functionaldevices.com/pdf/datasheets/RIB21CDC.pdf
and the input module on the PA system... http://www.toaelectronics.com/media/specs/s-04s_t_sa1e.pdf

 

[AnalogKid]

Given that the float switch is a dry contact, this should not be so much work. At first it sounded like the switch had failed and is closed all the time. If you're sure that's not it, then the only other option is noise pickup. Even at 1/2 mile round trip, it takes a lot of noise *current* to do anything significant like false-trip a relay.

What are the specs/ratings of the float switch? Running 110 Vac 1/4 mile could create opportunities for other problems, so if possible I think you should shift to a lower voltage DC control system. Also, DC is much easier to filter because of the availability of high value capacitors. Since the alarm module needs 24 Vdc, go with that.

When the alarm module has power but is not connected to anything else, what are the open circuit DC voltages on the two control inputs to ground? This tells things about what that module can and cannot tolerate.

ak

 

[slick-rides]

This is the float being used... https://madisonco.com/index.php/pro...t/249-m4548-industrial-tilt-float-switch.html

The float switch works correctly.

If I use a 24vdc power supply and relay, what else would I need to control noise?

I'll check the module open circuit voltages and post back.

 

[slick-rides]

I'm only getting 5 vdc on the input of the module. I'm getting ready to leave for the long holiday weekend. Will check back in next week. Thanks again and happy holidays.

 

[Bluejets]

Perhaps an opto isolator at the alarm end feeding a relay.
Only the opto current flowing in the long run via the float switch.
In the past, when running long distances (around 1 klm) between pump station and turkey nest , we used probes at the "float end" and probe relays at the "alarm end".
May be an overkill for your situation.
Control cable was 2 foil brass shielded 20 core 1.5mm2 cable( brass shield to stop termites eating the cable)

 

[GPG]

Determine the resistance of the cable, the full loop.

 

[slick-rides]

I have about 60 ohms of resistance through the full loop with the float switch in the up or closed position. Is that too high or?

 

[Alec_t]

I'm only getting 5 vdc on the input of the module.

Between which pins?
Is any of pins 1,2,3 common with the ground rail of your 24VDC module?

 

[slick-rides]

Pin # 1 is the common. Pin 2 and 3 both show 5 vdc. We use pins 1 & 2 which provide a continuous alarm, pin 3 would just be a short alarm.

 

[slick-rides]

Would a dedicated 24vdc power supply be able to overcome the 60ohm resistance?

 

[Bluejets]

If you use the attached circuit it should operate without any problem on your 60 ohm externel run and eliminate any external noise problems.

 

[Tha fios agaibh]

the only other option is noise pickup. Even at 1/2 mile round trip, it takes a lot of noise *current* to do anything significant like false-trip a relay

I agree it's likely a noise problem, but significant current is not always necessary to trip a sensitive alarm input like that.
Capacitance coupling or EMI noise could be spiking the voltage to levels that Biases the input circuit.
Perhaps just a high value resistor across the input could be enough to bleed off these noises?

 

[cjdelphi]

The first transatlantic morse code cables put across the sea bed from UK to US suffered the same fate...

But they never explained how they solved it just that the stray capacitances were to blame..

A bleed resistor (pull down) does the job so is the same applied for long cables?

 

[Tha fios agaibh]

I would imagine they countered high capacitance by adding inductors along the way.
Using twisted pair cable helps block interference, along with using shielded cable that has foil wrap and a drain wire.
I don't know what kind of noise the op is encountering, but if you were to put a scope on the input of that alarm where the 1400ft cable hooks up; you'd likely see AC noise superimposed on top of the DC signal.

 

[cjdelphi]

Inductors (allow dc but block ac) makes sense

 

[AnalogKid]

I think a few thousand miles of cable have enough inductance without adding any extra.

The first transatlantic telegraph cable was operational in 1858, but Oliver Heavyside's transmission line equations came along 20 years later. To me this says that the first cable was designed and manufactured without any real awareness of transmission line effects such as termination and characteristic impedance.

And, it weighed over 1 ton per mile.

ak

 

[Tha fios agaibh]

I believe long parallel spans of wire are more capacitive than reactive.
I would think the reactance would be relatively low unless your dealing with higher frequencies.
Telephone companies still add inductors in places to compensate for high capacitance caused by long parallel spans.

 

[AnalogKid]

Within a signal pair, wire-to-wire capacitance actually is greater with twisted pairs than with straight parallel pairs. Also, I assume that in your first sentence you meant to say resistive rather than reactive. Like inductance, capacitance is a reactance.

ak

 

[GPG]

Telephone companies still add inductors in places to compensate for high capacitance caused by long parallel spans.

Yes In the case of morse code the pulses get spread out and loading coils compensate.
https://en.wikipedia.org/wiki/Loading_coil

 

[davenn]

Yes In the case of morse code the pulses get spread out and loading coils compensate.
https://en.wikipedia.org/wiki/Loading_coil

damn that takes me back a few years, working on loading coils in phone cables ....


Dave