Output impedance of Pantry Tx circuit

[6SN7WGTB]

Am tinkering with a pantry Tx and in pursuit of wringing out every possible mW (or maybe µW!) was curious as to the output impedance of the Tx - essentially so that I could tinker equally with various simple aerials which I can test to some degree using a nanoVNA.

The Tx uses a MOSFET to drive the output, which then goes through a pi network before arriving at the aerial.

I'm not experienced in this area at all, so I have some questions:
1. Is the driver FET essentially presenting minimal impedance to the pi network?
2. So it's the pi network that is setting the output impedance?
3. And if I 'match' that to the aerial impedance at my Tx frequency then I will achieve maximal power transfer?
4. Can I measure that output impedance? (I have an RF VM with a 1M input impedance which I can load up accordingly).

As an aside I modified the Tx pi to have a small VC which so far I have simply trimmed to give me greatest voltage on a nearby voltage-double RF probe connected to a DMM. (So, sort of optimised, but I'm left curious as to whether I am 'matching' well enough...)

[mickm3for]

Hi its the pi network that matches the pa to the aerial as well as some filtering. Tune for max current in the aerial to get max power out Mick

[paulsherwin]

Bear in mind that you don't really want to maximise range with a pantry transmitter. Even a very low powered design can chuck out a significant signal if well matched to a good aerial. You don't want to attract official attention or annoy your neighbours.

[6SN7WGTB]

Any suggestions as to how I might measure power?

Currently, no pun intended, I am simply using a RF probe to detect voltage at a fixed distance from the aerial.

Yes, this maximises what I can get from the pi and the aerial, but it doesn't actually help me understand if I have a true match.

i.e. if further modification might eke out a few more feet of clear signal.

I doubt I am actually transmitting more than a few µW right now.

[Radio Wrangler]

Take a reasonably sensitive receiver and verify that your signal has faded below being noticed by the boundary of your property. That will be sufficient and you do not want any more because it may bring official notice.

Knowing the power out of your transmitter might seem useful, but as you won't be using an antenna on it with any knowledge of the performance of the antenna, the power isn't a useful thing to know.

If you want to measure transmitter power at these sorts of levels, you are entering the realm of some fairly expensive test equipment and the need to get it calibrated. You need something more sensitive than a Bird 43 power meter.

This is also an open forum and anyone could be reading this.

David

[mickm3for]

Hi v pp across 50 ohms see https://www.ab4oj.com/test/pwrmeas.html also make sure any harmonics are NOT radiated ie pi tank not tuned to 2x frequency etc as harmonics could travel longer distance

[6SN7WGTB]

Yes I've looked at harmonics and the radiated area is not increased. I've used a SA to check.

All I want is reception across my own property which is not substantial!

I'm not looking to break any records but in the genuine pursuit of improving my RF knowledge I would like to make the best of what I have.

[kalee20]

Originally Posted by 6SN7WGTB

The Tx uses a MOSFET to drive the output, which then goes through a pi network before arriving at the aerial.

I'm not experienced in this area at all, so I have some questions:
1. Is the driver FET essentially presenting minimal impedance to the pi network?
2. So it's the pi network that is setting the output impedance?
3. And if I 'match' that to the aerial impedance at my Tx frequency then I will achieve maximal power transfer?
4. Can I measure that output impedance? (I have an RF VM with a 1M input impedance which I can load up accordingly).

It's not an easy question, and just like an audio amplifier, you may not want to load it with an impedance equal to its output impedance.

You could in principle measure what the output impedance is, which is likely to be a combination of resistive and reactive parts, but if you make your aerial a conjugate match, you may find the pi network is so heavily damped that you get bad harmonic rejection (I am assuming the output device is operating in class C).

Once you start adding variables, like tuning your pi network, then all bets are off anyway because you've no longer got a fixed circuit.

[murphyv310]

Originally Posted by 6SN7WGTB

Yes I've looked at harmonics and the radiated area is not increased. I've used a SA to check.

All I want is reception across my own property which is not substantial!

I'm not looking to break any records but in the genuine pursuit of improving my RF knowledge I would like to make the best of what I have.

Best idea is become a Radio Amateur, pass the foundation first then do the intermediate and you then can construct your own gear (not of course for outwith the amateur bands).

[Radio Wrangler]

The output impedance of your active device is going to vary through the phase of each cycle, so there is no possibility of designing any fixed circuit to perform a true match even at a planned spot frequency.

It comes down to what fixed impedance was the transmitter designed to feed, and then configuring the antenna and any transforming network to produce that value. While you can amalgamate things, it is easier and gives a few added freedoms to have impedance transformation and harmonic removal treated separately.

Filter design for these things is, um, 'interesting' On one hand you have antenna impedance which can go wild outside the intended band (and wild at harmonic frequencies.... it doesn't stay a nice, sane 50 Ohms) and on the other hand you have the time-variant Z of the output device seen through a fixed transformation.

David

[murphyv310]

At medium wave it will be very difficult to get a good match to an aerial. A dipole would be huge for those frequencies and would then radiate over a considerable distance even with microwatts of power.
A small whip with a base loaded coil is a good idea which can be tuned to the frequency, this will help matching.
Using a Pi network and a tuned antenna with say a base loading coil you are less likely to emit harmonics that can and do travel further.

One issue is that I find a few microwatts are simply not enough to give good results as the QRM from today's electronics exceeds the power of the pantry transmitter.

I have two LED red flicker lamps sold by a large online shopping company that I USED in my electric coal effect "fire".
When I switched them on the whole of Long and Medium wave were obliterated, I did some tests on the bulbs and sussed they were radiating nearly 300mw of RF, suitably transmitted on the mains wiring. Neither the seller or OFCOM were interested.

[6SN7WGTB]

Getting off topic a little but since I have got into restoring old MW/LW radios, I have done some detailed sweeps of the house and disposed of at least three cheap SMPS-driven devices and/or replaced their PSUs.

1. A SMPS on a Siemens Gigaset handset recharging station - broadcasting an enormous pulsed output which obliterated MW. Caps gone bad, replaced with newer model;
2. A Canon inkjet where the SMPS had gone bad, Binned the printer;
3. Another SMPS where the caps had gone bad, and aside the MOSFET running at 100 degC, was broadcasting continuous RF hash. Replaced.

In addition, ferrites on EVERY LED bulb in the house, either on the flex, or around the cable into the fitting.

Returned an LED bulb to Amazon as the RF was so invasive couldn't listen to certain FM frequencies. CE-marked but we all know they just lie and print that on anyway. Apart from the fact it's a nonsense self-certification process...

Wife has an iPal radio in her workshop room and with an LED worklight nearby was unlistenable on certain FM frequencies until I looped the iPal PSU lead into it through a chunky ferrite several times. Which did completely silence the interference. I think the iPal PSU is actually linear but it was the LV output lead that was picking up the RF from the lamp.

I now 'listen' to every new device coming in to the house with an RF sniffer, plus do periodic scans of the house with a portable SA (as this crap all goes bad eventually). Have to, because as Trevor says, no-one cares. And if it's noisy it goes back and a brutal review is placed online.

[Hybrid tellies]

On the SSTRAN 3000AM pantry transmitter they use a rectifier diode and capacitor tapped off the output pi or tank network. By putting your DMM on DC volts across this rectifier circuit with the short aerial connect you then peak the output stage. I believe this not only maximises the signal output but also tunes or matches the short aerial to the pa stage more effectively and so reduces the radiated harmonics.

This pantry transmitter gives out 100mW and when peaked running into a short wire aerial running up to the curtain rail it gives good coverage around the house but it drops off very quickly outside the house with the signal disappearing into the noise halfway down our short drive.

So I think it is a good idea to optimise the matching of the short aerial to the transmitter as it should reduce any unwanted radiation and with a careful choice of transmitter frequency to reduce any chance of interference with any remaining local AM broadcasts you should not attract any unwanted attention.

If your signal is affected by noise from the many cheap and nasty gadgets that we surround ourselves with these days you could also try different frequencies to reduce this. Where I live the lower and middle part of the Medium wave band has very high levels of noise which has got worse over the years so much so that I am considering retuning the transmitter to the HF end of the band.

[6SN7WGTB]

Thanks Simon, what you suggest appears analogous to what I've done - used my RF probe and DMM a set distance from the aerial and tuned the VC to give max mV.

Overall my house is now relatively 'quiet' across MW, but it has taken quite a lot of effort...

A friend found his "mains wifi network repeaters" a massive source of noise - no real surprise there.