Quote:
Originally Posted by Radio Wrangler
Spectrum analyser crystal filter design is somewhat contorted to making programmable bandwidths and synchronous tuning shape (not actually Gaussian, though close enough up top)
David Ford's gang did the 3585A over at Loveland, but he'd been at the 'ferry for some time. I think my write-up of the simpler spectrum analyser filter set (in the 3724A) is in the HPJ article (I no longer look like a hairier Harry Potter).
The higher freq analysers use PIN diodes to control stage Q and hence make variable bandwidth.
With so many resistors floating around, loss == noise floor takes a hit.
David
|
HP did not openly publish the schematics in the service manuals of their spectrum analyzer. I was wondering how variable RBW could be done with crystals.
Conventional filters with Butterworth or Chebyshev filter topology have a rather long settling time in the transient response. The attached example(source:
https://2n3904blog.com/butterworth-filter/)
illustrates the time response of a Butterworth filter of nth order. There is overshoot in the damped oscillation making them a deal breaker for spectrum analyzer application in particular RBW.
The bandwidth of a crystal can be affected by the input and output impedance, and the coupling coefficient k (capacitive or inductive). The variable k can be achieved by the use of variable coupling transformers or varactor diodes.The 2nd attachment shows the variable bandwidth crystal filter developed by Telefubken in the 1930s. The crystals are isolated by buffer amplifiers and the tuning is achieved by a pair of LC tanks at the input and output of crystal:
http://lucafusari.altervista.org/pag...er/Filter.html
The third attachment shows a variable bandwidth crystal filter for a homebrew spectrum analyzer. The load impedance is adjusted by a PIN diode which acts like a variable resistor:
http://lea.hamradio.si/~s53mv/spectana/sa.html
Quote:
I had a rummage in an old boneyard box of engineering samples, and I still have one of the early YIG filters. See below. This must be about 30 years old. This should be a 500-2000MHz YIG filter and I think it was about 15MHz wide. I'm not sure if it works though, it might be a dud.
The HP 8566A/B has a PRESEL PEAK button on it that allows the user to optimise the 2-22GHz YIG tuning at a given frequency. However, in my experience pressing this button can cause problems as it can cause dips in the YIG tuning at other frequencies. There is a button sequence that restores the factory default YIG calibration, and the default YIG cal works really well on mine. I try and avoid using the PRESEL PEAK button.
I've dug out my HP 8568B today and switched it on. It must be the first time I've used it in a few years. With the internal correction turned off, the various crystal RBW filters show a slight frequency error and also some amplitude errors (+/- 1.2dB between them), and some are not symmetric anymore. I do have the calibration supplement to the service manual, and this describes how to correct the LC and crystal RBW filters. The last cal sticker on it is dated 2006 and that must be about the year I bought it from the company. I'll have a go at aligning the filters today as I have some free time. It's too cold to go outside!
|
We are posting at the same time again.
Wish you luck with the calibration. The process looks complicated in the service manuals!
If you find the datasheet any of the YIG filters, you may be able to power it up. I saw this guy powered up a HP YIG tuned oscillator:
https://youtu.be/NTtPTJQGLLY
It seems the YIG tunable oscillators are used by the military often. There are tons of them on ebay.