SWR and Auto Tune

[NZ0I]

When pondering ways to make the transmitters as simple as possible to use, I've toyed with the idea of having the 80m transmitter antenna matching be automatic. At first blush that would appear to be outlandishly complex. I once worked for a company that built automatic antenna matching units for marine use. They were large large devices, contained 20+ relays, and a small processor devoted to the task of switching in/out the right amount of inductance and capacitance to achieve the best match, and to do it quickly, for transmitters running 100W or more output power between 3MHz and 30MHz.

But it occurred to me while modeling the SWR measurement circuit, that our tasks (both SWR meter, and matching) should be made much easier because we only have to worry about a small segment of the radio spectrum: just 500 kHz between 3.5 MHz and 4.0 MHz. In fact, for ARDF, the segment is more like 200 kHz or less. Also, the power levels are small, and only vary by ~10 dB or so: 100 mW to ~1W. Add to that, that we can choose the approximate length and type of antenna we will support (short vertical wire from 3m to 10m in length, placed above lossy ground), and perhaps the job of matching the antenna automatically becomes much easier.

I built a set of six L-network 80m antenna tuners to use with the Ohio group's transmitters a few years back. I had no idea what range of L/C combinations I might need, so I added the ability to set L and C over a large range. What I discovered in the field was that, when using the same length of antenna wire, there was never any need to adjust L at all. And C only ranged over a small arc of the tuning capacitor's tuning range. This would probably translate to switching in/out a combination of 3 or 4 capacitors by an automatic antenna tuner, to achieve an approximate match, with no inductance switching at all. We don't need to achieve a perfect match, just get to within 2:1 or 3:1.

The transmitter processor in our design will already have the ability to read the SWR, and to control the output power. So here is a strategy that might work:

1. At power up, the transmitter processor will adjust the 80m output power to its lowest level (~100mW).

2. The transmitter will take an SWR measurement at low power.

3. If the SWR is too high, the processor will set the matching capacitance to another value, and go back to the previous step.

4. If all possible values of capacitance are found inadequate for a match, the transmitter will issue an SWR error indication.

5. If match was achieved, the C value is written to EEPROM (to be tried first next time), and the transmit power is set to the competition power setting.

The matching capacitor bank will consist of (perhaps) four fixed-value capacitors, that can be added together in parallel, in any combination. This will provide 16 unique capacitance values that can be tried. If the transmitter can test one combination per second, it would take at most 16 seconds or so to achieve a first match, but only ~one second thereafter when using the same antenna.

At transmitter construction time, the user could choose the correct values of C, and value of L, from a list we provide, in order to have matching work with the antenna length to be used.

It seems to me that this can probably be made to work. And the cost might actually be less than the price of quality variable capacitors that would go into a manually-adjusted tuner.

[NZ0I]

I found this document that describes a homebrew autotuner similar to what I am proposing for the transmitters: http://www.qsl.net/cx1ddr/much/atunnerauto.pdf

The main differences: we would use only one inductor (not 8), no high/low selection relay, and probably fewer than the 8 capacitors shown. The SWR detector shown in the schematic is also essentially identical to what I've proposed.

[Gerald Boyd]

I like the hf matching idea.

We could turn off the relays when the transmitter is off cycle to save power or even better yet use latching relays so they can mechanically keep there state using no power. 
Jerry
Sent from my iPad

--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/receiver-development-platform/8f1c1070-5cda-45e8-a090-0273be923526%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Charles Scharlau]

I like your suggestions. These suitable opto-isolator switches draw about 2mA per device when turned on:

http://www.mouser.com/Search/ProductDetail.aspx?R=LH1518AABvirtualkey61370000virtualkey782-LH1518AAB

If all four capacitors were engaged (four switches turned on) they should draw less than 10mA while transmitting.

This latching relay is among the least expensive and should draw about 30mA only when changing state (while tuning):

https://www.digikey.com/products/en?keywords=399-11010-1-ND

These would draw zero current once the antenna has been tuned, and would automatically store the tuner setting. But using these relays would be more complex, because they need to be driven two different ways, one to set, and the other to reset. Not a huge deal, but a little additional complexity.

For ARDF, which has a relatively low duty cycle, the extra current required for the optical relays shouldn't be a big problem. But a latching relay solution is attractive if a reliable relay is available at a good price. The main specs to look for, whether latching or opto:

1. Coil/switch operates at 3V or less.

2. Contacts can handle at least 300mA at 100V AC. (DPDT can be half this if wired in parallel.)

Those values include a good deal of margin, I think. 5W into a tuned antenna should have a maximum amplitude of ~22V at match. It could go much higher for an unmatched condition - but we should be tuning the antenna at a much lower power level, so hopefully the devices will rarely approach the maximum voltages and currents of a 5W transmitter into a large mismatch. If it does happen from time to time, devices with the specs above should still survive... I'll see if modeling will confirm that.

To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.

[Charles Scharlau]

Oh, one more spec:

3. On resistance less than 1 ohm (ideally).

The opto-isolator relay I suggested falls short of that goal, but probably would still be ok. I think we want the on resistance to be small compared to the reactance of the capacitor being switched. For an 80pF capacitor at 4MHz the reactance would be about 500 ohms.

[NZ0I]

As Columbo used to say, oh, and just one more thing...

4. Output capacitance < 1pF.

I think the output capacitance should be small compared to the capacitance being switched, since when the relay is turned off this will appear as capacitance in series with the capacitor being switched. The smallest capacitance we might switch is 5pF. I think we'll want to larger than that on 80m, but 5pF is a worst case. So we would like the relay's output capacitance to be small compared to 5pF. We may not find that in an affordable opto-isolator relay. More likely we would have to settle for 10pF or greater output capacitance. Still, 10pF might be OK if our smallest switched capacitor is 20pF or more. But we may have to wait and see what test results we get.

So the full list of relay specs for an ideal relay would be:

1. Coil/switch operates at 3V or less.

2. Contacts can handle at least 300mA at 100V AC. (DPDT can be half this if wired in parallel.)

3. On resistance less than 1 ohm (ideally).

4. Output capacitance < 1pF (ideally)

5. At a reasonable price (e.g., < $3)

 That might rule out all but mechanical latching relays. But I'm sure some of those specs can be safely relaxed.

On Wednesday, February 22, 2017 at 10:23:56 AM UTC-5, NZ0I wrote:

[NZ0I]

Based on my searching, latching relays seem to be the best bet. If we come across a better alternative, we can always change. Below is the basic matching unit schematic. The latching relays are these: http://www.digikey.com/products/en?keywords=399-11010-1-ND. The load switching units (six of them) are the same LMN200B01 used in the receiver design, and the PCF8574AT I2C port expander is also used in the receiver design. This should give the transmitter the ability to select between 16 different capacitance values to apply for antenna matching. This could be expanded up to 64 different capacitance values with the addition of two more relays and load switches, using the same 8-bit port expander.

On Wednesday, February 22, 2017 at 10:23:56 AM UTC-5, NZ0I wrote: