[time-nuts] General questions on 8510C network analyzers

[Jim Lux]

On 7/4/11 1:40 PM, Javier Herrero wrote:
> Hello, Jim,
>
>
>>
> Again for now, no extensive frequency precision, so would be suitable. I
> suppose that this is valid with the 8350B sources, but does it operates
> in the same way with an 8340 synthesized sweeper, or in that case it
> commands the frequency source for each point? This can be an important
> point to take into account (again, mostly thinking in future needs)

I think that's determined by the controller, more than the sweeper. 
That is, the controller doesn't have the software to send a command to 
the sweeper for each point.  It sends the commands to set the 
start/stop/sweep time, and then has an actual discrete "start" pulse to 
start the sweep.

>> 2) The calibration math isn't as slick as you can do today. I suppose
>> one could get raw data out, and then post process it for any sort of cal.
> Good to know :) again for now, more interested in amplitude than phase
> measurements... but I prefer an VNA since I think it is a better
> investment for future projects

you definitely want phase.  Otherwise a regular old sweeper and a 
spectrum analyzer would do.  But phase implies a synchronous receiver, 
which greatly improves the dynamic range, and that's often what you want..


>>
>> 3) It's a big beast. Better than 1/2 rack full of stuff
>> (display/controller, receivers, test set, sweeper)
>>
> I know, I know... have seen several, mostly sitting inside a 18U rack
> only for them :). But I prefer to have an external source (for
> flexibility reasons)
>> 4)The user interface for the 8510 is fairly easy and convenient,
>> although distinctly 80s-90s vintage, and the general build quality is
>> very good, so you get good quality metrology: repeatable, etc.
> I have seen that 8510s are used nowadays in lot of places, this is also
> a fact that makes me somewhat confident on them. And I feel happy with
> these 80s-90s vintage user i/fs (an age-adquired defect, I suspect ;) )


The challenge will be fixing it when it breaks. They've been "out of 
support" from Agilent for a few years now, and they're a complex beast. 
  OTOH, given the level of ad hoc support you can get from groups like 
this one, you can probably get enough info.

I think the 8510 was the first real high performance automatic VNA so 
there's tons of them out there (even at hundreds of $k each, they save 
so much time for the engineers...).

The RF sections, attenuators, detectors, etc. are really nice, I wonder 
if it would be worth trying to come up with a retrofit for modern PCs...


> Also... I like the building quality of HP in that period (do not like
> Agilent nowadays so much... you can call me nostalgic :) ). Also, I
> think that the 8510s are more 'repaireable'. I suppose that this is not
> the case for a modern Windows-base Agilent PNA.

I think the folks on this list would agree...  Those 8663s and such are 
probably never to be duplicated.



> I think I prefer the big thing ;) Another problem is that I must have it
> calibrated by an accredited laboratory, and to get a calibration
> certificate for an 8510 is not difficult since most good calibration
> laboratories already know it (not an unexpensive certificate... but not
> difficult to get). These little units, with external mixers and sources
> for the mixer, would be a bit of franken-vna to convince the customers
> on the measurement quality (you know... the -hp- logo is related to
> quality even for 20-30 year old instruments, and the customer feels more
> confident if they see that you use good quality instruments for doing
> the qualification - they don't seem to mind too much if they are not the
> latest state-of-the-art unit)


Ah, yes... then an 8510 might be what you want.