Selective voltmeter wave analyzer project?
I want to alter it to a "selective voltmeter" or "wave analyzer", for the HF radio amateur bands, so that I can monitor the different harmonics levels.
I believe this can be done with a set of band pass filters in the input. Is there any other way this can be done?
Hm... searching around I have found an excellent filter for the job http://www.af4z.com/Contest_Filter.htm
It uses band-reject filters, which do not need to use power intuctors and capacitors, based on the fact that the harmonics of a transmitter cannot usually be that large to cause components burn-out. The good thing is that this filter if properly switched will have a much smaller attenuation of the wanted signal, whereas all band-pass filters have some attenuation. This will hopefully allow for more accurate measurements.
What is your oppinion on that approach?
As you do not indicate the required frequency range and RF bandwidth, I would prefer to make a simple spectrum analyzer instead. There are nice designs on the Web, some use old TV tuners, some modern "digital radio" concept.
By the way, a "wave analyzer" is generally an oscilloscope. "Signal analyzer" may be a distortion meter. The reference from Sweden describes only a rough power-level indicator with a quite wideband input.
The important question here is what you need to measure?
I thought this as a very simple scheme for measuring the actual power of the carrier and ignoring any harmonics. Most higher ham bands are multiples of the lower ones, so the band reject filters only need to be on these bands, they do not need to be tuned.
Now compare the simplicity of this thing with the complex circuits of the spectrum analyzers. It is not a substitute, but it could do a good job tor decent measurement, without the need tor an SA.
Not to mention that it would be nice to see on a scope, the effect in the waveform distortion by cutting out the harmonics you want not all the higher ones (which would normally be done with a BPF).
They say that the band reject filters of this type may be too narrow (their deepest notch), so maybe a small variable cap could be used instead of the fixed ones.
I also winder if I could use an LDR and a led instead of this expensive optocoupler.
The Swedish power indicator is a very rough device that can indicate short-wave band power on a log scale. For real measurement of e.g. harmonic content in a RF signal, one needs a spectrum analyzer or a calibrated selective voltmeter (wideband tunable receiver).
You can replace these instruments with a set of filters and a power indicator but it is a cheap solution with corresponding results.
You can use AD8313 series MMICs for UHF. I used it from 10 kHz up to 3 GHz. It has a log response, too, but its response is very wideband starting from -60 dBm. Any filter would have to be designed with reject level so high.
Nowadays you can buy a SDR dongle for 35 USD, and program it as a spectrum analyzer. I think this is a good solution for the future, no filters needed.
Or try to find on e-bay a good receiver covering 1 through 50 MHz. With calibration they can be used as good test instruments to measure harmonics and more.
Hardly if you want a similar transfer function. But surely an inexpensive standard optocoupler.
Is there any other passive way one could achieve a logarithmic scale?
Will I better use Germanium diodes?
I am thinking that this circuit might not be able to measure any harmonics, if they are low enough.
Theoretically it should, by switching the relevant notch filters, but the meter, being able to show only 5mW or so, means that the low amplitude harmonics might not be able to be measured with this simple circuit.
The problem with home brewing something like this will be calibration, and knowing with some certainty what you are actually seeing.
Even building filters, how are you going to know the insertion loss and rejection figures, which would be the entire basis of relative harmonic amplitude measurement.
Even with a fairly crude spectrum analyser, at lest you will be able to see the whole HF band across the screen, and can then make some very rough judgements about relative amplitudes.
Thanks Tony,
However a spectrum analyser it a more complex project. It can be made more simple but not that simple. I have started such a project please see http://www.qrp.gr and navigate at the link "Poor man's HF spectrum analyzer"
This is done without any VHF oscillator by splitting the 0-30MHz into two and using the image of the mixer. Hopefully this will finish one day.
However the simplicity of the filter/meter aproach is attractive. On simulation I find the deepest notch point of each filter section (I design my own filters) to be quite deep (-50db or more in some cases) in properly chosen components values, but also quite narrow and in some cases it cannot cover the whole band. I believe that a simple means of varying the inductance or the capacitance in each filter section, could tune the filter a bit to eliminate the harmonics no matter where they exist inside tha band.
Note that I would not use this filter in receiving as the site suggests, because it cannot cut ALL other frequencies, but just the harmonics, which means that RX overload might happen from powerful stations out of the filter notches. But for the TX harmonics use, this is fine and allows for greater flexibility (for example to see on a scope the effect of each individual harmonic on a signal, or measure only one harmonic and cut the others)
I would not expect very accurate measurements, just enough to judge the output signal quality and power of a transmitter. Calibration is only a matter of setting the notch of the filters and indeed it may not be so easy in some cases. Maybe a low frequency square wave VCO/VFO or a combo generator can be used, that is used as a marker for all frequencies and one of it's harmonics to be set on the same frequency as the transmitter frequency. Then one can calibrate the filters. Anyway, it is more complex.