TheRadioBoard

I'm thinking I need a separate receive antenna for my station-in-progress. A larger non-rotatable balcony antenna would be used for transmit, and a smaller hand-rotatable magnetic loop antenna with deep nulls would be used for receive to null out local interference sources.

Since my FT817 rig doesn't support a separate receive antenna, I'll need to build a T/R switch. I'd prefer it to be automatic and will be focusing on CW at the start.

The simplest thing it seems to me is to wire the key to two relays, one that keys the rig and the other that switches the antenna.

I'm suspecting it may not be as simple as this. Is there anything else that needs to be considered in this situation? It seems that if the keying relay is faster than the antenna relay that there could be a tiny instant when the transmitter is transmitting into either the receive antenna or an open load, which I think should generally be avoided to protect the finals and also prevent carbonizing any circuitry in the receive antenna path (e.g. a preamp, non-high-voltage tuner, etc.). On the other hand, switching the antenna relay before the keying relay (assuming I find some way to do this) will subject the receiver, for an instant, to the noise from the transmit antenna wihch the receive antenna is designed to reduce, meaning that key-down would be accompanied by a brief burst (or click?) of static. It seems I would need to mute the receiver, switch the antenna, then key the transmitter... this is starting to sound a lot more complicated than two relays. And, will the antenna relay contacts get dirty over time, being switched at a high rate with high current flowing through the contacts?

Pointers and pitfalls would be appreciated.

I built something like this to connect an MFJ noise-canceling box to a transceiver without separate receive-antenna port. What made it more interesting was that I was receiving on a loop and using my transmit antenna for the noise antenna on transmit. Oh, yeah, and keying an amplifier, too.

Use two identical, high-speed rf-spec relays. They aren't expensive if you shop eBay. The first relay, which is the one you key, is double-pole. One pole handles antenna change-over. The other pole keys the second relay, which in turn keys the rig. The brief mechanical delay before the rig-keying relay closes is enough to ensure the antenna change-over has occurred. There are more-elegant solutions, but this is good enough for your purposes. I used a third relay to connect the transmit antenna to the MFJ box on receive. My relays are tiny, about 10-pin DIP footprint but taller, and can handle 100 W at HF.

The relays have 5-V coils, but have gold-plated contacts and were designed for rf switching. Knowing the dc coil resistance, I just stuck a resistor in series with each coil, so the box runs off 12 V. Rf-spec 12-V relays are usually more expensive; resistors are cheap.

73,

_________________

http://kr1s.kearman.com/
http://qrp.kearman.com/

Excellent idea to ensure sequential switching! Time to go shopping for some relays.

I found this relay. Does it look suitable?

http://akizukidenshi.com/download/941H-2C-12D.pdf

Also, on key up, is there a possibility of transmitting into the receive path? It seems first the antenna relay will be released, during which time it starts to travel back to connect the receive antenna. A brief instant later the rig keying relay has started to travel back to unkey the rig.

The question is, can the antenna relay return all the way back to the receive position with the rig relay still being partially energized and allowing transmitted RF to flow into the receive antenna path? Or will the rig relay break contact quickly enough before the antenna relay can return to connect the receive antenna? It still seems like there will be an instant though (between breaking contact of the antenna relay and breaking contact of the rig keying relay) when the rig is transmitting into an empty load, even if not the receive antenna.

EDIT: I think I've found a way to ensure the rig is keyed on after the antenna is switched to tx, and also ensure that the rig is keyed off before the antenna is switched to rx. Will post a schematic later today. Idea: a third relay in line with the rig transmitting relay that will release simultaneously with or, if I'm right, before the antenna relay releases.

That will work, and that relay looks like a winner. When I had a Heath SB-220 I rigged a reed relay for the input and a vacuum relay for the output. At 1000+ W I needed to keep the vacuum relay closed until I was sure the input relay had dropped. I used diodes and the RC time constant of electrolytics across the relay coils. But that was 20+ years ago and I'd have to sit down and work it out again! At 5-10 W I don't think you'll have a problem with two relays.

73,

_________________

http://kr1s.kearman.com/
http://qrp.kearman.com/

How did the diodes figure in?

My idea is the same - an electrolytic across the relay coil which - if I remember right, as I haven't played with relays in ages - will keep one relay energized longer after power is removed. Question: will an electrolytic across the relay coil also slow the energizing time of the relay when power is applied? I think not. (For an adequate power supply, it should be impossible for the charging current required by the capacitor to reduce the current flow through the relay coil at all, meaning there should be no change in the energizing time of the relay with or without a capacitor across the coil - right?) And also - I suppose at high keying rates the electrolytic across the relay coil might not have enough time to energize.

Enough talk - here's my schematic:


S1 is the operator's key. "To ant" is the transmit antenna; receive antenna (connected when S3a breaks) not shown.

Key down behavior: The rig is only keyed after both S2 and S4 close. S4 can only close after S3b closes. S3b closes basically simultaneously with S3a, the antenna. So S4 is closed and the rig is keyed only after the antenna is switched to transmit.

Key up behavior: C1 holds S3a/S3b energized briefly whereas S2 breaks contact more quickly. S2 breaking contact unkeys the transmitter (even though S4 is still energized). Some time later C1 discharges, S3a (and S3b/S4, but we don't care about that) breaks and disconnects the transmit antenna (and reconnects the receive antenna, not shown).

Comments?

One other thing I realized - my rig has a built-in CW keyer (for paddle use), but with this relaying setup I can't use it. Oh, well.

You're in danger of over-engineering here. I mentioned the switchbox for the MFJ noise-canceling device, where I have to move the transmit antenna over to the noise-canceling box on receive, etc. My rig is a Kenwood TS-450S. It doesn't have a separate receive-only port, but it does have two internal relays. The normally-open (NO) contacts of one relay go to a rear-panel RCA connector. I use that set of contacts to switch the relays in a power amp. Another set of SPDT contacts comes out to a DIN connector, and that's the set I use to manage the switchbox.

In Receive, the transmit antenna is shunted over to the Noise Antenna jack on the MFJ box, and the processed output of the MFJ box is connected to the TS-450S antenna terminal. When I key the radio, the antenna jack is switched back to the amplifier input and the MFJ box Noise Antenna line is shorted. I'm not using any caps to delay relays, and nothing has let out its smoke.

I use this setup on 80 M. On 40, noise and TV crud is less of a problem, so I turn off the switchbox and MFJ device. It takes about 80 W from the transceiver to drive the amp to rated output, substantially more than the 10 W your rig can produce. If the FT-817 has an amp-control connector you should be able to use that, and retain its internal keyer function.

Your rig should have a QSK-delay setting, which determines how long it stays in Transmit mode after you let up on the key, or stop talking when using VOX. If you set that so the rig stays in Transmit between dits, or even a bit longer, all rf output from the transmitter will cease before the relays disengage.

You asked about the diodes. I wanted to delay the closing of the input relay by charging a cap, but I didn't want the cap's discharge time to hold the relay engaged. I put resistors in parallel with the diode-controlled caps to discharge them.

73,

_________________

http://kr1s.kearman.com/
http://qrp.kearman.com/

Well, there are some reports (whether accurate or not seems disputed) that the FT817 is unusually sensitive to mismatched loads and blowing its finals. I'd prefer to over-engineer a solution if it reduces the amount of stress I subject the rig to.

For your amusement, here's another variant that occurred to me. It should ensure the proper "activate key-connect antenna-key rig" and "deactivate key-unkey rig-disconnect antenna" sequence.

I'll check my rig manual about your suggestion of QSK timing delays. But that might reduce the QSK usability of the rig (my ability to hear my frequency in between my CW elements) once my speed develops. On the other hand, is QSK that important in practice?

P.S. It was unexpectedly interesting to sit down and think through a firing sequence of relays that would produce the desired result. I suppose its digital behavior makes it easier to reason about than analog circuit behavior.

I'm sorry I got you started on this path. Relays are so 20th Century. Zack Lau, W1VT, of ARRL, published a 40-M transceiver project in QEX several years ago, that was picked up in one of their QRP compendia. In it, he used comparators to sequence T/R switching, muting and sidetone. Simply charging caps and setting thresholds for comparators is probably easier, and better practice than ganging up relays. You could use PIN diodes for rf switching, but for that part, surplus relays may be less expensive. The comparator method would require fewer relays, though you might enjoy all that ticking while you're sending.

73,

_________________

http://kr1s.kearman.com/
http://qrp.kearman.com/

Well it looks like I may need to actually implement this project, which started out more as a thought experiment. On my balcony I have strung up 2 dipoles, one for 10m and one for 6m, but whenever I listen either (a) the band is dead (at night), or (2) deafening broadband noise plagues these bands (during the day, though 20m and down seem OK). I did barely manage to hear a CW station calling CQ through the noise on 6m, so at least I know there is some life out there on 10 and 6.

Next step will be to try a small receiving loop for 10 and 6, as small loops have two deep nulls that can cancel isolated noise sources. If that works, then I'll make a T/R relay to swap between the dipole for transmit and the loop for receive.

Let's hope the noise is coming from an isolated direction, not everywhere...

Looks like I'll need every trick in the book (like separate T/R antennas) to get my ham station on the air. It's certainly a very educational experience though!

you just need one relay.

Connect a coax from the behind or front antenna connector to a relay (central connection) and operate this relay through the ACC connector. You need the "TX GND" connections,..this is the PTT switch from the ACC connector to activate a transverter or amplifier.
The other two "outputs" from the external relay go to your seperate antenna`s. Or build yourself a pin diode switch instead of the relay.

There`s something on it on the net,..not sure if it is level operated or just pulled to gnd.

How it reacts on CW I do not know if it is fast enough.

EDIT:
see:
http://www.microwaves.pwp.blueyonder.co.uk/FT817%20interface%20v%202.pdf

And look at "the design" to operate the transverter. This can also operate a normal relay od pin-diode switch IMHO.

73
Frank

_________________
www.pa2mrx.nl