limited lifetime of reed-switch / varistor ?

We have connected a reed-switch (max 500 mA) to a digital input of a PLC (approx 7 mA). It is switched with a frequency of approx. 100 Hz by a circulating magnet. After 2 or 3 days, initially we measure that the frequency becomes less, before it does not work anymore. Under a microscope it looks like a grain of sand between the surfaces of the reed-switch.

* what causes this effect on the reed-switch, as the current is not too high? Can higher temperatures (130°C) cause these kind of problems?

On the input side of the PLC there is a varistor (between input and ground); is it possible that a varistor acts like a capacitor (so it causes a high current-peak), although the voltage (24V) never exceeds the specified level of the varistor (30V).

Thanks,
DYV

What is the expected mechanical life of the reed switch? After one day of continuous service it has completed 8.64 million operations. This seems like a bad application for a switching mechanical device. Look for a different type of sensor.
 

----------------------------------
  Sometimes I only open my mouth to swap feet...

A hall effect sensor seems like a more ideal sensor for this application...

Dan - Owner
http://www.Hi-TecDesigns.com

@ScottyUK
The mechanical life-time is approx 10E9 operations. And isn't it true that at the end of the mechanical life-time, the contacts of the reed-switch will simply break. And the sticking effect should normally be caused by (too) high currents/power.

@macgyvers2000
Is there a Hall-sensor that is able to withstand temperatures up to approx. 130-150°C. And what will the dimensions be, as the reed-switch has a length of only 10mm.

DYV

From your observation that there is a change in the reed switch contacts, I would think that the 7ma current is the average and that the MOV capacitance results in switching current spikes.

Add a resistor in series with the reed switch - in the range of 1K to 3K Ohms. This will limit any current.

Temperature is an issue with some Hall sensors. Since there are plenty of Hall sensors in automotive use, there should be Hall sensors available.

Another possibility is a wire coil (like a minature solenoid coil) to pick up the magnet. Depending upon frequency range, you may need some conditioning on the signal generated.

As Automotive ignition systems(hall effect) are used under the hood for automotive spark triggering. 125C   ~36M triggers/year.

Another method would be to mount a mirror on the device and note a reflection with an off-the-shelf proximity light switch which could be mounted at a distance. (shiny reflected type)

Keith Cress
Flamin Systems, Inc.- http://www.flaminsystems.com

Use a magnet with a coil and comparator input.

Plesae read FAQ240-1032
My WEB: <http://geocities.com/nbucska/>

nbucska has a good idea.

On your problem you have an unlimited 24V supply driving through a switch to the PLC. Is the PLC actually rated for 24V? If not, the current would exceed the 7mA you think you have, and this would explain the problem.

If you post the type numbers of the PLC and the varistor we can give more accurate answers.

You could put a current probe around the series path through the reed switch and this will confirm if the current is too high.

Thanks for your inputs so far.

@logbook
The PLC is suitable for 24 V, and we have actually measured the current (7.3 mA).

The PLC is ED&A V5 model 1000, and we have connected the read-switch to a digital 24V input (input nrs 3,4,5). The varistor is specified as 1206K30.

@comcokid
The idea of the capacitance of the varistor has crossed my mind, but this current spike must be extremely short, as we haven't measured it.
We have tried using a resistor (1k) in series with the reed-switch, but the life-time became much shorter (??).

In general: we can not use an optical sensor, but a sensor that will be able to pick up the magnet. The dimensions should be in the range of 10 mm. The temperature can go up to approx. 150°C and as we switch with 100Hz, we are looking for a life-time of approx. 10E9 cycles. Will the solution of comcokid or nbucska work for this?

Thanks in advance.

Life-time shorter with 1k !  Puzzling.

Are you looking for an off-the-shelf, bolt-on solution, or are you solving a product design problem where you will be building from scratch?

A quick google (first random hit) found the following bolt-on Hall-effect sensor. It has a 0.5"(12mm) range when used with regular magnets. Look at the model 1101.

http://www.electro-sensors.com/documents/1101%20931%20ES-104%20Rev%20D.pdf

Pre-made coil type bolt-on devices can be found in a similar cylindrical package with leads out-of-the-ends.

If you are designing a solution for a product, Hall-effect ICs come in small packages, and there will be a number of things, both electrical and magnetic, to consider.

http://www.allegromicro.com/techpub2/an/an27701.pdf

Thanks, we have found a hall-sensor at Allegro with the correct temperature specification. We will test the samples next week.

Good luck! Let us know how it goes.

Keith Cress
Flamin Systems, Inc.- http://www.flaminsystems.com