I agree with Chuck. Also keep in mind that not only does the A4 perform the function of matching the internal CBT microwave cavity with the outside world with a high degree of temperature compensation, the input side of the A4 used a SRD to generate a hash of sidebands from the input 90 MHz, from which one and only one sideband was selected. thus the name "Harmonic Generator".
It's a fascinating piece of machined metal and electronics, and is a conversation piece when employed as a paperweight. Dave Carlson ----- Original Message ----- From: "Frequency Standards & Services" <[EMAIL PROTECTED]> To: "Discussion of precise time and frequency measurement" <time-nuts@febo.com> Sent: Wednesday, 01 June, 2005 12:31 PM Subject: RE: [time-nuts] hp5061A questions Hi David: It isn't so much that the equipment is rare or hard to find, it is just knowing how to tune them and what to look for when you do. One example of getting one tuned wrong, you will still get a good signal out of it but the temperature range that the unit will operate in will be drastically reduced. You will find that small room changes in the room temperature will cause serious changes in your stability and sometimes even make them break lock. I have attended the class that HP used to give on tuning them and I have a set up for completely rebuilding them. Unfortunately, anymore, it is almost cheaper to just find someone with a cesium with a bad tube and buy the Harmonic Generator from them. On the other hand, if you just want to experiment with one and see what happens, as long as you have a known good unit to put back in when you are done, it probably wouldn't hurt anything. Just don't experiment with the only one you have. As you probably recall from working with wave guide screws, a tiny movement can make a huge difference in output, frequency, etc.. Chuck Norton Frequency Standards & Services 2727 E. Palmer Park, Ste. 100 Colorado Springs, CO. 80909-3032 719-228-0540 voice 719-228-9009 fax -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of David Kirkby Sent: Wednesday, June 01, 2005 1:15 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] hp5061A questions Dave Carlson wrote: > Absolutely. Especially don't mess with any of the mechanical adjustments on > the A4 harmonic generator that is attached to the CBT. These are aligned to > resonate exactly to 9.192 GHz and it would be almost impossible to realign > one without using a special waveguide tuning fixture. Only a few of the > extrusions are intended for matching the A4 to each individual CBT. > > If you've already turned some of the knurled screws, nuts, knobs, and hex > screws, you may never be able to capture the correct resonant peak, even if > you can get the cesium oven to start working. > > The best remedy to a mis-aligned harmonic generator is to find another one > that has not been adjusted and attach it to your CBT, or send it to someone > with the correct waveguide fixture. > > Dave Carlson As a matter of interest, from someone who knows next to zero about cesium sources, but a reasonable amount about radio frequency and microwave engineering, what does the waveguide tuning fixture do? What is so special about it? Could an amateur fabricate one? X-bade rectangular waveguide suitable for 9.192GHz is easy to come by (known in the UK as WG16, but also as WR80 and RG52). I used that very waveguide as a child in a 10GHz amateur radio system with a Gunn diode. Tuning of waveguide components can be achieved by inserting screws in the side - again not too hard to fabricate. My guess (and note I have never seen a cesium source) is that the waveguide fixture would temporarily be part of the cavity, so if the dimensions of that fixture were known, it would not be hard to make it oneself with access to a milling machine, or get a small engineering company to do it. I appreciate from reading what you have said that not touching the assembly is the best idea, but if someone has "tweaked" one, I would not have thought getting the cavity back on resonance would be hard, given a few physical dimensions, and a microwave power meter. Waveguide to N adapters are pretty easy to fabricate too, although again I suspect they can be bought used for very little. And microwave power sensors with N connectors are easy to come by, or again could be made easily. Just more curious than anything else, as I don't have a cesium source, and don't feel I will ever become a real "time-nut" and get one. Soon I should have my bit of kit built which will have both a 10811-60111 and a Standford PRS10 both locked to GPS. I'm sure the two oscillators will drift relative to each other, so I know I'm going to be questioning which is the most stable, but I think I will stop before buying a cesium source. -- David Kirkby, G8WRB Please check out http://www.g8wrb.org/ of if you live in Essex http://www.southminster-branch-line.org.uk/ _______________________________________________ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts _______________________________________________ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts _______________________________________________ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts