Relays for Switching Sense Leads on Power Supplies

Mercury-wetted reeds should do the trick, but there are other kinds of relays rated for dry circuit use that don't use mercury (eg. inert-gas filled with precious metal contacts, regular reeds etc.).

Best regards, Spehro Pefhany

Most reed switches have wiping contacts. And do not have the ROHS problems of mercury.

Curiously, my Pathfinder has this problem. You have to blow the horn before turning on the cruise control.

I seem to recall reading an EU directive that stated that Mercury-wetted relays are RoHS exempt, as there is "no technical alternative" There's several exemptions that have been declared in that category.

Barry

Exemption or not, mercury can get you (or your client I guess) into trouble in some markets because you never know when those bureaucrats change their minds. Can't it be done via electronic multiplexing? Or just little FETs?

As far as I know...... most power supplies maintain a permanent connection between the main output terminals and the sense terminals using low value,

The assumption is made that if the end user wishes to use remote sensing then they will have installed the wires to do so and that automagically shorts out the resistors.

You do have to consider the relative DC error introduced by those resistors when remote sensing is not used but it's sort of in the order of 0.2% so, compared to the error introduced by the power leads, it might be ignored.

DNA

Ed, Mercury wetted relays are good if you are concerned with "contact bounce". Otherwise, any relay with gold or gold/palladium contacts should do the job nicely.

I've used them (the palladium gold contacts, IIRC) quite successfully to switch thermocouples, which are about as dry as things get*. JL, OTOH, reported a less cheerful experience with them.

• Come to think of it, if there was an open, the break protection circuitry would put a substantial voltage (volts, not mV) across the open contact at

My experience with dry reed switches is that the contact resistance will go quite high (~ 5 ohms) if they sit around doing nothing for long periods of time. Indeed, I have measured such resistance in reed elements I have in stock. As long as they are operated frequently the wiping action of the contacts will keep the contact resistance down in the milliohms range.

Due to the need for consistently reliable readings over the lifetime of the equipment indicated by the OP, I would hesitate to use dry reed switches. I would lean towards a mercury wetted type such as

Take a look at Fujitsu type FTR-B3 relays. We use them all the time for low-level switching. Several mfrs have drop-in equivalents.

Hey, I just googled "Fujitsu type FTR-B3" and got a very interesting first hit! Very strange.

John

One of the big problems of switching sense-leads is that any fault or glitch causes the supply to go to an uncontrollable output.... perhaps causing damage to A.N Other's equipment. It is really a designed-in accident waiting to happen.

DNA's suggestion looks the safer way to go.

Surely they'd run a few nA or uA through the leads to avoid that situation?

Best regards, Spehro Pefhany

Thanks for the suggestion I order a bunch of the SMT types that Mouser had in stock. ExpressPCB.com here I come.

Ed V.

I have recommended the needed data acquistion resources be used to ensure that the proper sense lead configuration is in place prior to switching the power supply output on. Whether or not my collegues will do this is a matter of debate.

All of the equipment that will powered from these power supplies are equipped to withstand a wide variety of minsconnections and report faults over an RS 485 or other com to the test system. Operators are protected from exposure to all electrical connections by interlocked shields. Which is all great when it all works but oh boy the one time it doesn't.

Thanks for keeping me on the safety track.

Ed V.

Most supplies have internal resistors across the sense paths, so they won't go open-loop if the remote sense circuit opens up.

John

:-)

I've just looked at a 600W proprietry system supply. There is a permanent resistive connection between Vout and +Sense, but it is 10k. DNA's 470 ohm or so looked much more reasonable.

Vout -----O--+--------+--------->Load | | \\ | 470/ _ o \\ /| | / +Sense ----O--+-----P o------>Remote sense.