[R-390] HP-8640B vs. HP-8654A/8655A/11710B

[Barry]

> Well, the 8654 (A or B) was never intended to be an 8640B-class 
> generator.  In terms of signal purity, it may be the dirtiest RF 
> generator HP ever made, including the earliest synthesized 
> models.  Fine for basic alignment, but not really useful for 
> twin-tone IM measurements or anything fancier.
> 
> These days, one would typically use it with a counter -- but be 
> careful, the auxiliary output on the rear panel is not well isolated, 
> and any crap that comes out the counter's input (not at all unusual) 
> can end up on the main output, further compromising signal purity.  I 
> suggest using a 20 dB pad between the rear panel output and any 
> counter to minimize this.
> 
> The 11710B works reasonably well (I use them with 8640Bs to get below 
> 400 kHz), but the internal crystal oscillators are a bit noisy so the 
> mixed output has a poorer noise floor than the 8640B and also has 
> more phase noise, if you are doing anything where phase noise is a
> concern.
> As Al noted, the "B" version makes a voltage control input available 
> on the back panel.  If you have the HP 8655A reference/phase detector 
> (or build something similar -- the originals were rare even in their 
> own day), you can use this input to lock the 8654B to an external 
> standard, at the price of quantized frequency steps.
> 
> Best regards,
> 
> Don

Okay, Don.  I'm just doing simple IF/RF alignment so I guess it's probably good enough for my purposes.  The 8655A is the counter and not sure how much extra noise it would add to the mix.

Still might just go with an 8640B.  The 11710Bs cost almost as much as a good 8640B.

Thanks for the info,

Barry - N4BUQ