Re: Spectrum analyzer Tektronix 7L13 on mainframe Tektronix 7603

Nenad Filipovic

On Sun, Dec 27, 2020 at 09:18 AM, Thomas Garson wrote:
However, even if it is only a "flat" white noise source, I can see the
utility it offers.
I never considered using a broadband noise source as an alternative to
the tracking generator. From your post, it appears that you have done
so, with at least acceptable results.
The eBay ad is a mess but the device is decent, well worth it but won't do wonders for $20. It's a white noise source, just a plain Zener diode hooked to a wideband amplifier. Reference (average) level is approx -30dBm, my specimen has a broad bump of +3dB at around 1GHz and then continuously drops to -5dB at 1.8GHz. However the curve is rather smooth; I mostly use it for spans of just a few MHz and in that case it's virtually flat. For wide spans I use a 7854 program to record a sample of a "flat" reference, and then subsequently apply the inverse of it to correct the actual measurements.

I also have a TR502; its generous output level adjustment is superb and it provides much better SNR in any measurement compared to the noise source. But sometimes funny things happen if you hook it to a DUT which is far from 50ohm, resulting in bumps and tilt of the bandpass curve. I empirically concluded that (this particular) noise source is more resilient to that, so I always try both devices. During Covid lockdown in spring I calibrated the IF of a B/W tube TV receiver from the 60s, and ended up using the noise source. The result was like from a textbook.

Best regards and Happy Holidays,
Nenad Filipovic

Join to automatically receive all group messages.