Author Topic: Using a function generator to phase-lock a low-precision signal gen?  (Read 489 times)

0 Members and 1 Guest are viewing this topic.

Offline okto

  • Newbie
  • Posts: 4
I need an RF signal generator that’s good up to 500MHz, but even older HP and Fluke equipment is out of my budget. What isn’t out of my budget are lower-frequency function generators.

Would it be possible to create a precision RF signal generator by using the function generator’s signal output to modulate a lower-precision RF oscillator, and the function generator’s sync output to phase-lock the RF oscillator?
 

Offline Miti

  • Frequent Contributor
  • **
  • Posts: 422
  • Country: ca
That big spark at power up was by design!
 

Online 0culus

  • Super Contributor
  • ***
  • Posts: 1259
  • Country: us
  • Electronics, RF, and TEA Hobbyist
What is your budget? If you look around you might be able to get a deal on a real RF signal generator.

Failing that, there are tons of relatively cheap options. I have an Analog Devices ADF 4351 based dev board that you can buy for ~$24 on ebay. It's not perfect, but it's easily controlled by an Arduino (software already exists or you can roll your own). I bought it to provide a signal for a tracking generator I have for HP 8566 and 68 spectrum analyzers. Sweeping the low band on the 8566 with this TG requires a 3621.4 MHz signal, which is just above my current best RF generator (HP 8664A). [edit] if you don't use a 3.3V Arduino to control it, you'll need to design (or buy) a logic level changer. It's relatively easy to make a set of voltage dividers with some perfboard that will work well enough. I'm using an Arduino Uno so I built a level changer.

This board will get up to 4.4 GHz and you can hook up an external precision frequency reference with a little SMD soldering (need to move some 0 ohm resistors around to disable the internal xtal and connect the path to the extra SMA). Performance-wise it's pretty good for the price. However, phase noise isn't the greatest and there are a lot of harmonics down in the lower frequencies so you'll want good filters. On the other hand, below 100 MHz or so it makes a pretty damn fast edge, around 1 ns. It isn't a good permanent replacement for a proper RF signal generator, but it's better than none!
« Last Edit: May 07, 2019, 09:29:43 am by 0culus »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 9361
  • Country: us
  • DavidH
You could pull this off by using a DDS function generator as the reference to drive a phased locked frequency multiplier.  If the low precision signal generator has a voltage control input, then it can serve as the oscillator being controlled by the phase locked loop.
 

Offline VK5RC

  • Supporter
  • ****
  • Posts: 2170
  • Country: au
Some questions that may be helpful.
Specific purpose or general rf source?
Budget?
Output level?
Phase noise desired / essential ?
How frequency agile / adjustable?
Harmonic suppression?
Modulation?
-Rob
Whoah! Watch where that landed we might need it later.
 

Offline okto

  • Newbie
  • Posts: 4
Thanks for the responses, y’all.

I’m pretty new to this area of test equipment, and this is a hobby for me, not a profession, so my requirements could certainly be misinformed. :P

Budget is $200ish. Would prefer no computers be involved. I like the tactile interface of dedicated lab gear and the whole reason I need it is testing/calibrating/aligning other analog test equipment and older radio gear which I have gotten into because it’s NOT computerized or excessively automated.

Output level probably no more than 10-15dBm, don’t have a spec for phase noise, harmonic suppression as good as I can afford (I have a Tek spectrum analyzer that needs some attention).
Modulation AM/FM 1kHz tone, don’t need arbitrary waveforms or anything fancy like that. 
I’m not sure I know what frequency agility means in this context. Like sweep generation?

I have been offered an HP 3325A, which probably exceeds my requirements for modulation/phase noise/harmonic suppression but only goes up to 20MHz, so my idea was to use it to modulate and sync an external likely DDS oscillator.
 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 4657
  • Country: au
"Low budget" , 500MHz & stable don't often appear in the same package.

I would suggest to continue looking for something (or things) in the HP, Marconi, Rohde & Schwarz, etc, class.(Several sig gens for different frequency ranges will do the same job as one "all singing, all dancing" multiband one does.)

I'm not really sure how frequency stable the average function generator is, so you may be better off trying to pick up  a "GPS disciplined reference source" on eBay instead, & lock the sig gen to that.

 

Offline joeqsmith

  • Super Contributor
  • ***
  • Posts: 5365
  • Country: us
I have a very old HP8640B that's good for about 500MHz.  Maybe you could find one of these for cheap.  Specs are decent for the age.  They support AM and FM only.  It's manually tuned and has a sort of frequency lock function.   I gave away its synthesized brother.    Personally, I would have more use for this than something low end.
How electrically robust is your meter?? https://www.youtube.com/channel/UCsK99WXk9VhcghnAauTBsbg
 

Online 0culus

  • Super Contributor
  • ***
  • Posts: 1259
  • Country: us
  • Electronics, RF, and TEA Hobbyist
@ OP, I PMed you a link. I found an excellent condition 8640B with the 1 GHz and expanded AM options for sale right at $200. It would be a huge step up from trying to bodge something together with function generators.  :-+
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf