Question about designing DC relays in DC application

Oh boy... I just noticed I butchered this posts title. I must have been distracted and didn't finish my thought. It should read " Question about designing WITH relays in a DC application"

Forgive me?

George

I believe you will find that exactly what happens when the contacts opens is more important than which side is which. I think if you can make sure the relay opens quickly, that will work best to prolong the life of the contacts, given your resistive load at low voltage. To that end, let the relay coil voltage fly back to as high a voltage as you reasonably can, so the coil current decays as quickly as possible. To a lesser degree, it's good to cause the relay to close as crisply as you reasonably can. So a coil driver that over-voltages the coil initially for a few milliseconds, and then goes to a lower voltage to hold the relay in, can work well.

This is the sort of thing is something the relay manufacturer should be able to help with, through either ap notes on their web site or assistance from an aps engineer.

Cheers, Tom

Unless this relay has magnetic blow outs on the contacts, the polarity of the current through the contacts probably makes no difference. My worry would be using a relay with so little extra current rating at such a high temperature. Is the coil rated to operate at 100C? Have you a derating curve for the contact current rating versus temperature?

For your application, I have never heard or seen that it matters where the hot side and the load side are connected, nor can I think of any safety issues that would make a difference (unless this is an "open" relay -- not encased in plastic).

I have read about, and used relays in other applications (i.e. low-level mercury-wetted relays) where it seems to make a difference where the "hot" side is connected to the relay contacts, because of the way the armature "wets" the stator, but this doesn't apply in your application.

It sounds like you are using the proper relay in the proper application. To increase reliability make sure the relay is rated for a temperature much higher than 100C. If you can find a relay with contacts rated for higher current that would help, also.

No magnetic blow outs (to extinguish arcs I assume). Just a simple sealed relay.

Coil is rated to operate at 125C.

I have worked with the manufacture extensively regarding this relay. It is not printed, but they guarantee performance of this relay up to

40A at 125C. I had them run a "test to failure" on 20 samples given my worse case conditions. I need about 100,000 cycles out of the product during its lifetime. Their testing revealed the first dropped out of spec (although still functional) at 230,000 cycles. The better of the batch went over 300,000 cycles before falling out of spec.

I am also running a similar test in parallel to confirm their results. I feel fairly confident I have the right device for the job.

Thanks,

George

Swapping = no problem. Why not use a mosfet? Then you don't need to worry about burning up the contacts.

Ed

We use allot of high current contactors in DC applications. I can't think of an instance that does not connect the load (to neg return) to the movable contact and the source (pos) to the stationary contact. Of course, all our contactors have magnetic blow out arc chutes and that alone depicts this contact arrangement to clear the arc upwards. If the contacts were reversed the arc would deflect downwards, possibly burning up the contacts, or coil. I'm not certain if it matters with your relay, but why not configure it that way to be safe! Also you can buy relays with magnetic blow outs for that small of current.

Three reasons:

1.) Cost - a FET may be cheaper than a relay, but with support circuitry and mechanical mods it is a more expensive solution.

2.) Heat dissipation - there is no room in my package space to add a heat sink. I could modify the package space, but the heat sink assembly and the modification adds to the cost.

3.) The potential for reverse voltage condition - current will conduct through the body diode of a FET when power wires are connected backwards (by mistake of course). This is not desireable in my application. There are ways fix this, but they add to #1 COST.

A big thank you to everyone for participating in my request for help,

George