After a year of lusting after an SG 504, but never being willing to spend what was being asked on eBay, and then again spend the same amount for a separate leveling head, I splashed out a much more reasonable amount of money for the stand-alone 067-0532-00 Calibration Fixture, Constant Amplitude Signal Generator (which happens to have the leveling head permanently attached). I'm actually quite pleased with the purchase, as the instrument came with a manual, is working, and seems to be the same instrument pictured on the TekWiki page. It's also a lot more satisfying to use than a TM500 plug-in (the controls are nice and beefy!).
The instrument has a hand made plastic label on the tuner housing that reads "300 MHZ MAX" but I have verified that it produces a signal all the way up to the maximum dial setting of over 500 MHz. What I can't so easily verify is whether or not constant amplitude is maintained all the way to 500 MHz, as my fastest scope is a 2465 which is only specced to 300 MHz (though it doesn't seem to roll off until around 425 MHz).
The peak to peak detector calibration fixture (067-0625-00) can be had on eBay for almost what I paid for the '0532' but I'm wondering if I can effectively roll my own, or otherwise get a cheaper solution. There are ICs that might do the job (AD8317 or LTC5507), and complete RF power meters based on these devices are available on Amazon and eBay for $20-$50, but I'm interested in rolling my own from discrete parts (because why not?).
My current plan is to construct a very simple peak detector using a Schottky diode and a small capacitor (1 pF) with a small value resistor (200 ¦¸) on the input to limit current through the diode (~10 mA), and a large value resistor across the output (1 M¦¸) to bleed the capacitor down slowly. The Schottky diodes that I have on hand include BAT41s and HSMS-282xs, which may just be able to handle rectification of a 500 MHz signal (I don't fully understand Trr of Schottky diodes: I watched a video by W2EAW about Trr and it looks like it's a different behavior for Schottkys than for regular diodes).
Beyond building this simple peak to peak detector, my plan is simply to verify that the output amplitude is flat all the way up to 500 MHz (or to do the required adjustments to get it there). As a second goal I would like to verify that the output amplitude is correct for the selected amplitude (it looks very close on my scope, at lower frequencies, but I'd like to be sure). I figure that I can just add the Vf of the diode to the measured rectified voltage across the cap to get the actual output amplitude.
My question is, will the BAT41s work for this application? Do I need to try using the HSMS-282x diodes (which will need to be mounted on an adapter board), or are they even suitable for this?
I've also got questions about some of the calibration steps, but I will ask those separately when I get around to doing the calibration.
As an aside, I'm curious about the history of this calibration fixture: it is SN 000366, but other than two plastic labels (one, red, on the tuning housing, previously noted, and one, green, on the face plate echoing the serial number) there are no other markings. The power supply has been recapped, but nothing else seems to have been modified or replaced. There is some damage to the plastic housing of the leveling head. If anybody knows anything about this unit's history I would be happy to hear it.
-- Jeff Dutky
