Using a Mini-circuits ZX47 power detector in LabVIEW

I remember reading several blogs and some article in the ARRL handbook (James Hontoria, W1JGH) about the Mini-Circuits power detector series ZX47. This power detector is useful within a frequency range from 10 MHz to 8 GHZ with a typical dynamic range of 60 dB. The actual dynamic range depends on the model of the detector:

Model Number	Frequency Range (MHz)		Dynamic Range @±1 dB Error (dBm) Typ.	Output Voltage Range (V) Typ.	VSWR
click for data	Low	High
ZX47-40+	10	8000	-40 to +20	0.50-2.10	1.20
ZX47-50+	10	8000	-45 to +15	0.50-2.10	1.20
ZX47-55+	10	8000	-50 to +10	0.50-2.10	1.40
ZX47-60+	10	8000	-60 to +5	0.50-2.10	1.7

(direct link to product selection page of ZX47, opens a new window)

The cost of one of these detectors is about 90 EUR, which is still quite affordable considering the alternatives.

Unfortunately, these detectors output an analogue voltage so that it is necessary to provide some kind of voltmeter or ADC circuit to do the actual data acquisition. In the past, I just hooked the detector output up to my bench multimeter and recorded the data via the GPIB interface. Needless to say, it would be nice if I could just connect the device to a USB port and talk to it via SCPI or similar so that I can use it in LabVIEW without the need to occupy my main bench meter.

The solution I found was to use an LPC1768 embed and have LabVIEW object libraries map to the mbed API via the RPC protocol. There is a special RPC firmware readily available on the embed webpage under the following links:

http://developer.mbed.org/cookbook/Interfacing-Using-RPC
http://developer.mbed.org/cookbook/Interfacing-with-LabVIEW

After a bit of trial and error to get LabVIEW to talk to the mbed, I managed to write a nice virtual instrument that provides some basic logging functionality:

Apart from displaying the instantaneous power and some basic measurement statistics, a second ADC channel is used to display the current temperature of the power detector (it has an integrated temperature sensor). An idea for future additions would be to have automatic temperature compensation and provide means to select the frequency of interest, gate time, power offset and resolution.

For anyone interested, I wrote a short paper that shows the LabVIEW code and does some basic assessment regarding accuracy. The document is available as PDF from my Google drive. The LabVIEW *.vi files are available on request.

It's been quite a ride

Well, it really has been quite the ride during the last two years while working full time and completing my Bachelor of Engineering. When you sit down in front of a desk all day at work and then sit in front of a desk at home some more while you do your assignments and write your thesis, you really have to learn to accept that you will not always be able to accomplish everything exceptionally well. You have to learn to prioritise and learn that it sometimes is more important to get a job done and make peace with your decisions when you had to compromise.

But here I am finally, everything done and graded and looking forward to dust off some of the projects which patiently have been staring at me from their shelf.

I never would have thought that I would be able to benefit from having used older GPIB with my home lab equipment for a project at work. I thought that nowadays everything is done with more modern test racks with PXI or similar but it seems that I was mistaken. When projects have a tight budget and the majority of the instruments at your disposal have the legacy GPIB port...it seems foolish not to use that old National Instruments USB-to-GPIB controller from the dumpster and your commercial license of LabVIEW 8.6 to develop and assemble an ATE rack and use it for your production testing. I guess it's a little like those 8-bit micro controllers that certain people claim should be dead and superseded. They're just too good at what they do, they get the job done, they are cheap and they are available. They are not dead just yet.

2G Band pass

This filter was a real pain to get hold of. Since the 2G path is only single-conversion for now (I'm actually working on a front-end improvement with a 2 GHz...10 GHz YIG oscillator), the requirements were pretty demanding. The 1G path actually tunes up to 1200 MHz, so that I was looking for a band pass filter with a pass band in the region from 1.2 GHz to a little over 2 GHz.

I still need to take a picture of the amplitude response but the pass band is from 1200 MHz to just above 2000 MHz with pretty steep skirts. It apparently originates from an old Hewlett Packard spectrum analyser and I got it reasonably cheap off an US-based surplus seller.

Band pass filter for the 2G path.

Update, got screen shots of  the filter response and return loss (looks like I forgot to calibrate, ah well...):

The 3db-bandwidth is about 980 MHz from 1074...2053 MHz.

Not sure if this measurement is meaningful, I forgot to calibrate when I changed the frequency span.  The return loss is about 15 dB or better over the pass band.

Front panel connectors and cable

Inside view of the front panel. I would love to use 0.141" hard line coax to connect the front panel connectors to the slim modules but I'm breaking my fingers trying to bend the coax. I wonder if I could use the semi-rigid 0.141" coax instead, I think to remember that Sam recommended using the hard line coax for high power connections (which the input connectors kind of are). Alternatively, I also have 0.086" aluminium hard line coax.

New enclosure

400x250x150

Closeup of the 7" TFT

Rearranging the Enclosure

I only have two of these connectors so I will have to order at least another six.

Tektronix Oscilloscope

I finally found a 1 GHz analog Tektronix 7104. The technology, especially the microchannel CRT, is truly mesmerising for the age it was developed in.

It's in fantastic shape, considering its age. Current plugins are a 7A22 Differential amplifier, 7A29 1 GHz amplifier, 7B10 and 7B15 delayed time base.

Two active 900 MHz FET probes TEK 6201

Quite a few extra plugins. Mainly 7A24 and 7A26 amplifiers as well as some additional time bases.

It replaced my old 20 MHz analog scope from LG. It's still working just fine after 15 years.