CMOS RF Power Meter Calibrator (suit AD8307 power meter)

[VK6WAX]

In a bid to measure what I experiment with, I'm in the process of constructing some 50 ohm RF testing devices.

As my guide, I'm using the excellent series called "RF Workbench" on the QRP Homebuilder website (http://www.qrp.pops.net – select “Topics 2010-2013”). This is produced by Todd, VE7BPO, and written in a very readable style with liberal use of diagrams and photos aiding understanding. Todd covers a lot more than RF test devices and he also takes much better photos, so do check it out.

One of the devices planned is an RF Power meter based on the AD8307 chip. However, I decided to first build a CMOS -10dBm RF Calibrator. This served two purposes, to try out “dead bug” style construction and have a means of calibrating the subsequent RF power meter.

As I wanted to show the circuit diagram here, I contacted Todd, who very kindly gave his permission for me to reproduce his drawing. Todd also explained that his work was based on articles written by Bob Kopski, K3NHI. After contacting Bob, he also very kindly gave his permission as well as pointing out the appropriate references. They are QEX Jan/Feb 2004 and QEX May/June 2010. Many thanks to both gentlemen and please respect their ownership.

Here then is the circuit diagram annotated with the components that I used, essentially what was either to hand or available locally. Both Bob and Todd used an MX045 clock (CTS Electronic Components) but by comparing datasheets I sourced a similar device by Abracon (ACH-10.000MHZ-EK) from Element14 (code 7942278), likewise the 52.3 ohm resistor (code 1083113) but 51 ohms can be used, or if you don’t have that value, try 100 ohm and 110 ohm resistors in parallel.

The heart of the calibrator is the CMOS crystal clock with a TTL compatible output, meaning the resulting square wave swings as close as possible from zero to five volts.
The other requirements are a frequency of 10MHz and a resulting output of -10dBm. These were chosen by Bob as the most suitable to calibrate an AD8307 based power meter.
The adjustment, in a nutshell, is to terminate the 50ohm output (BNC connector) whilst adjusting the trimpot for a reading of 50mV DC on a DVM connected to the test output (RCA connector).
For more details on the circuit’s function, adjustment and application with a AD8307 power meter please refer to Todd’s website and/or Bob’s articles.

The following photo shows the results. I found it useful to have a soldering iron with variable temperature control as the copper ground plane acts as a heat sink necessitating a bit more heat for those joints. The trimpot was anchored to the pcb with a dab of “Superglue” (cyanoacrylate adhesive) to prevent it moving during adjustment. The 9v battery is held in position by the on/off switch and an upright off-cut of pcb.

Continued in Part 2 (split due to attachment limit)

[VK6WAX]

Part 2

The oscilloscope display on my Kenwood CS-1044 (40MHz) flat out at 0.2uS/div horizontal timebase looks like this. The input is through terminated with 50ohms.

Fortunately the oscilloscope has x10 magnification and with a bit more vertical sensitivity (50mV/div) it’s possible to see a fairly good representation.

Interestingly, the following composite photo shows what happens to the waveform by increasing the vertical sensitivity to 20mV/div, upper part of photo and then using only the trace position to move it to the bottom of the screen, lower part of photo. Maybe it’s a measure of the oscilloscope’s age, lack of bandwidth or both. Despite several attempts the photo is still not as sharp as it could be.

So there you have it, a CMOS 10MHz -10dBm RF Calibrator ready for a suitable Power Meter. One final photo showing the outside, it still needs suitable labelling.

Many thanks again to Todd and Bob, it’s a bit of a buzz to have these gentlemen on board. I only hope I’ve done some justice to their work. Any errors or omissions are due to me.
Suggestions or comments will be gratefully received.
Cheers
Andrew
vk6wax