Maths question, voltage across a relay

[Chris Wilson]

I have always regretted not paying attention in mathematics. But there you go, we are where we are

I am told this formula will give V across an antenna change over relay's contacts:

Calculate it from impedance and power level. V = (8PR) ^ 1/2 (P= power R = resistance of the feed at the relay contacts)

P= 1000 Watts

R= 50 Ohms

I make it 500V? Wrestling with the Windows scientific calculator and a hazy understanding of whether the caret ^1/2 is the same as ^0.5

Thanks! And sorry Mr. Schofield, you did try, but some fell on stony ground

[hamster_nz]

   P = V^2 / R

   V = SQRT(P*R)

   223 V???

But then I guess you have to allow for standing waves and all that other RF trickery & magic.

[vk6zgo]

   P = V^2 / R

   V = SQRT(P*R)

   223 V???

But then I guess you have to allow for standing waves and all that other RF trickery & magic.

Not if the termination is a true resistive 50 Ohm.

[dmills]

But it nearly never is!

Also, be a bit careful of advertised relay voltage ratings, unless you are using a proper RF part they are usually measured down at wiggly DC.

Regards, Dan.

[bd139]

It's a bit more complicated than that.

You have to consider peak voltage if you're talking contact ratings.

So say you have 1KW into a 50 ohm system, that's 223 volts RMS (P=V^2/R) but it is actually around 624 volts peak to peak (Vpp = 2 * sqrt(2) * Vrms)

Now assuming a crappy SWR, assume that reflected power is all coming back, that means on a bad day your contacts are handling 1248 volts. 2.5KV isolation isn't uncommon in RF relays. You have a much bigger problem if the SWR is that bad because your little output transistors will say bye bye in record time. Reverse power protection is required.

Now imagine several tens of kW and you then have vacuum relays and all sorts of stuff.

This is one reason my TX is a more manageable 10W (still need a license before I can use it on air ) as all the protection is, is a fat zener (which brings other miseries but not those of blown PA transistors). Perhaps one day I'll build a big PA for it.

[jonovid]

I have always regretted not paying attention in mathematics. But there you go, we are where we are

so true 

[Chris Wilson]

Some very helpful replies, thanks. I have been thinking about when the disater happened. I fought intermittent PL type connectors on this thing indoors and after a few random whoopsies realised they were not making good contact or unable to handle 1kW. Cheap Ebay sourced Chinese things. Saw the light and bough proper Teflon insulated branded N type plugs and sockets. Issues ceased. until I saw a few weeks ago the gate and drain waveforms on the FET's of the amp go berserk. Luckily I was monitoring things and shut it off. Turned out one of the remaining PL plug / sockets in the box the loading coil is in outside was now in turn playing up. I should have ripped `em out and fitted more N connectors, but i didn't. I am suspecting what has happened is one of these connectors has played up again now I have added the RX / TX relay box, and instead of a 50 Ohm resistive load it's gone to God knows what and voltages have soared, maybe arcing the relay?

It's a Russian REW-15 with its proper matching connectors using RG-400. I would have thought the rlay would handle anything even a mismatched 1kW amp can throw at it, and with 500mS dead time when switching there should have been no danger of hot switching but who knows?

Relay box set up is output of 1 kW amplifier to NO connection and TS-590 to NC connection of the same relay, with COM port to antenna), then de-energized you'd always be in RX mode and energized you'd be in TX mode and there's no way the TX would ever be connected to the RX no matter what failed. The RX antenna socket on the relay box has another 100 Amp automotive relay across it to ground, and on TX that shorts the RX socket to earth, for secondary protection and reducing signal pick up. All this still seems functional, although the TS-590 having had a kW up the RX antenna port, is most definitely not.


The coaxial relay is here:

http://www.ur4ll.net/rew15.html

Thanks again.