View Single Post
Old 6th Jul 2020, 11:09 pm   #122
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: Replacement for HP8640B signal generator

OK but the main mechanism for generating internal birdies is when harmonics of each LO can meet somewhere (non-linear) like in one of the mixers in the signal path. If the beat frequency they generate falls in the IF passband then there will be an internal birdie. The order of the LO1 and LO2 harmonics that can cause this issue can be quite high so the system needs to be able to maintain really good isolation between the mixers up to many GHz. Any cabling you add could spoil the isolation. However, with general use you would be unlucky to spot a spurious term as they are likely to be down at the -110dBm level.

The HP8568B has an inbuilt feature to try and dodge this internal spurious issue where it can fudge the LO2 cavity sideways a few MHz. I think it does this with a diode based RF switch. When the shift happens then it forces the PLL to shift LO1 to counter the shift in LO2.

This happens really quickly and it should be transparent to the user. The change in LO1 and LO2 will then cause the spurious term to go off the CRT display.

I have a lot of experience designing RF downconverters that have to achieve low internal spurious so I have been down this path many times in the past. In the early 1990s I wrote various software programs (in Quickbasic!) to try and design a good frequency plan for various downconverters that worked up to many GHz with a digital IF of up to about 20MHz bandwidth. The software then predicted where the spurious terms will be.

I had a go at using this old SW to go hunting for internal spurious in my HP8568B and was able to find one fairly easily by turning off the LO2 shift feature. This prevents the HP8568B from dodging its own internal spurious.

It should be noted that this LO2 sidestep will affect the use of a tracking generator with the HP8568B unless you disable it. Otherwise you will see 'gaps' in the tracking generator response where LO2 does the shift sideways. It could be that LO2 is more stable in one shift position than the other. My guess is that the most stable will be when the diode is biased off.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline