Wanted - opinions on the HP 8647A RF signal generator

[philpem]

Hi there,

I'm toying with the idea of picking up a used HP 8647A signal generator (I've found one for what I think is a reasonably good price). Obviously it's HP so should be built fairly solid, but I'd like to hear from someone who's actually used one.

I'm planning to use it for radio receiver testing - HF (40m) up to VHF/UHF, occasionally modulated externally with data signalling. The spec-an and TG are good for aligning filters, but sometimes you need to feed the thing a real test signal!

I've heard the 8647A isn't well regarded for phase noise performance, but how likely is that to be a problem?

How easy are they to repair? Does anyone know if a CLIP (component-level repair manual) has made it out into the wild?

Cheers,
Phil.

[larry42]

I think this series was originally conceived for production floor testing of pagers and 2 way land mobile, so for your application it should be fine.

Sent from my One S using Tapatalk

[G0HZU]

I've used the 4GHz HP8648D plenty of times over the years at work and it's a close relative to the 1GHz HP8647A.

Basically, these are economy class signal generators with very poor noise performance and 'adequate' amplitude level performance.

The good points:
Relatively small and light. Easy to lug around.

I quite like the general appearance of the generator user interface. Very touchy/feely because of the two separate rotary controls for frequency and amplitude. So it can be very nice to use if you find yourself wanting to (constantly) manually trim the frequency and amplitude. Other generators can be quite fiddly by comparison.

I would expect you can buy these H8647A generators quite cheaply now.

The bad points:
Poor noise floor when used at levels around -10 to -20dBm. Notably worse than other types of generator.

Poor close to carrier phase noise compared to other generators. The method of signal generation is very basic and means that the phase noise performance down at the bottom end (eg the 2-30MHz HF band) is limited to the phase noise performance of the generator up at UHF. i.e. the phase noise at LF-HF is typically no better than the phase noise up at 500MHz.

This means it is going to be woeful for phase noise across LF-VHF when compared to a decent old school lab grade generator from the 80s or 90s.

I'd expect the quality/versatility of the internal modulation to be more than adequate for most users but notably inferior to other signal generators.

Down at the HF bands I'd expect the close in phase noise of the HP8647A to be about 30-50dB worse than a fairly decent generator (or a receiver with a decent low noise LO). i.e. in the carrier offset range of 1kHz through to 10kHz.

How much this affects you depends on what receiver tests you want to do. For basic sensitivity testing and for setting up the S-meter and aligning the dial of a homebrew HF receiver this generator will be fine.

But the problems will start if you want to do anything like IF rejection tests or image rejection tests or blocking tests on a fairly decent HF receiver. Ideally, the phase noise of the generator should be much better than the phase noise of the LO of the receiver under test.

[edpalmer42]

I have an 8647A.  One other point to remember about the entire series of generators is that detailled service information is NOT available.  There is a service manual, but it's pretty trivial.  If you have a faulty unit your chances of repair are slim.

In my unit, the battery backup has failed.  Instead of drawing microamps (or less) of current, it draws milliamps.  I decided not to try to fix it because of the lack of a schematic and the fact that this failure has no effect on calibration.  It just forgets the state it was in when you powered off and powers up to a default configuration.

I didn't realize how bad the phase noise was until recently.  The difference between the 8647A and my other synthesized generator was shocking! 

Ed