Re: "Sampler??"

On Wed, 31 Jul 2024 14:46:52 -0000 (UTC), Cursitor Doom
<cd999666@notformail.com> wrote:

On Wed, 31 Jul 2024 10:54:58 GMT, Jan Panteltje wrote:
>

On a sunny day (Tue, 30 Jul 2024 15:11:35 -0000 (UTC)) it happened
Cursitor Doom <cd999666@notformail.com> wrote in
<v8avr7$13368$1@dont-email.me>:
 

On Tue, 30 Jul 2024 14:51:10 GMT, Jan Panteltje wrote:
>

On a sunny day (Tue, 30 Jul 2024 11:20:47 -0000 (UTC)) it happened
Cursitor Doom <cd999666@notformail.com> wrote in
<v8aiaf$10lqo$1@dont-email.me>:
 

On Tue, 30 Jul 2024 10:36:51 GMT, Jan Panteltje wrote:
>

On a sunny day (Tue, 30 Jul 2024 08:58:35 -0000 (UTC)) it happened
Cursitor Doom <cd999666@notformail.com> wrote in
<v8a9vr$v9r2$1@dont-email.me>:
 

On Tue, 30 Jul 2024 05:16:24 GMT, Jan Panteltje wrote:
>

On a sunny day (Mon, 29 Jul 2024 09:29:04 -0700) it happened john
larkin <jlarkin_highland_tech> wrote in
<72gfajl8ke966lsu54bpajpang9dvgmt63@4ax.com>:
 

On Mon, 29 Jul 2024 05:19:06 GMT, Jan Panteltje
<alien@comet.invalid> wrote:
>

On a sunny day (Sun, 28 Jul 2024 21:40:41 -0000 (UTC)) it
happened Cursitor Doom <cd999666@notformail.com> wrote in
<v86dsp$3n74$1@dont-email.me>:
>
>

I still can't see where the 'sampling' bit comes into it.
AFAICT, there's are two signals into this thing and one signal
out. The signal from the LHS is passed through a very fast diode
which generates harmonics from that fundamental. The signal from
the RHS is unmodified and mixed with the desired harmonic to
give the necessary output signal which is then filtered to get
rid of the unwanted mixing products. If that's all correct, as I
believe it is,
where does any *sampling* come into it?

>
A 'sampler' is fact a non-linear mixer.

>
In the audio world, a "mixer" is a linear summer.
>
In RF, a mixer is usually a multiplier, typically a 4-quadrant
diode thing or a Gilbert cell type circuit. Or a 2-quadrant thing
like a pentode. Or even a single diode. All do some flavor of
multiplying.
>
Some mixers inherently multiply a signal by a square wave, which
resembles other multiplier-type things once you lowpass filter the
output. That's a "synchronous detector", which we usually do in an
FPGA.
>
A sampler is a signal multiplier too.
>
These are all variations on the basic idea of multiplying two
signals.

 
A the word 'mixer' in RF is not the same as a multiplier,,
Multipliers are like harmonic generating stages say you have a 1
MHz distorted sine wave and then a second stage tuned at 3 MHz,..
turned to third harmonic.
 
mamamatical speaking RF mixing is indeed multiplication.
 
Like Shakepierce once said:
 "What's in a name..." CocaCola wihout coke

>
I think in this context, HP is using the term "sampler" to mean
something which 'sniffs' a small amount of a signal for testing
purposes. Like a few turns of wire around a transmission line will
'sniff' a small sample of what's going on in the line.
Where I went wrong was rigidly thinking HP meant 'sampling' in the
manner a sampling oscilloscope works. That's what threw me. I tend
to interpret things very literally. :(

 
Would be nice to see a real circuit diagram of that thing.
What is the model / type number?
I am curious.
So far I have been able to find circuits for most stuff I came
across with google.

>
It's the HP8566B - a very well known lab-grade analyser of the day
which many RF designers still use. There is a *ton* of service info on
this model online. However, I have not been able to see a *proper*
schematic for the "sampler" in the A11A5 board which is where the
issue is.

 
OK, thank you, I will google a bit..
I got the user manual now:
 https://elektrotanya.com/

hp-8566b_spectrum_analyzer_100hz-2.5ghz_2-22ghz_usr-
programming_1984_sm.pdf/download.html#dl

Then trouble shooting manual part 1
 https://xdevs.com/doc/HP_Agilent_Keysight/

HP%208566B%20Troubleshooting%20&%20Repair%20Vol.%201.pdf

Seems to have at least some circuits (just a quick look)
and part 2
 https://xdevs.com/doc/HP_Agilent_Keysight/HP%208566B,

%208568B%20Troubleshooting%20&%20Repair%20Vol.%202.pdf

also has some circuits I will have a look at the hundreds of pages,
will take some time.
I presume you have these?
Took 10 seconds to find with google

>
Yes, I have all of those and more. I think altogether there are 7 very
large manuals for this analyser. However, I've not been able to find a
single actual schematic which gives component level detail of what they
call the "sampler" - and that's unusual, as they've given a great deal
of attention to providing very full service details in all other
respects.

 
Will read some of it and let you know what I come up with ..
 
OK have flipped through all circuit diagrams in the thing.
 
It looks to me like a heterodyne down converter followed by a narrow
filter IF stage tuned over the whole frequency range.
No sampling on the input signal at all.
Just a RF mixer.
PLLs in the tuned oscillator that are used to scan the band.
 
Like an old AM shortwave radio where you turn the big tuning knob really
fast mechanically to see the amplitude of each station on the cats eye
as you tune over the whole receive spectrum.
spectrum width set by the 450 kHz or so IF filter stages in those radio
days....
To get the IF narrow enough I did notice some crystal filter?, multi
stage RC filters?
Scanning the band takes time that way... horizontal deflection scope,
amplitude on vertical deflection scope..
Does it have a long persistence screen?
That way you can point at any small section of the input like long wave
/ medium wave / shortwave in the old radios.
Also because I do not see any Fourier transform / computah /
computation..
 
That thing is from 1987? enough computah power back then for Fourier
transform?
 
I could be wrong though, more than a 1000 pages read in... lemme see...
an hour.
Need to sleep on it and look again.
Rest is just scope trace drive and power supplies..
Its all so easy:
 https://panteltje.nl/panteltje/scope_tv/index.html
HP should have added a teefee mode, but alas, analog teefee is dead in
most of the world these days.
 
This is more fun:

OTOH my 1.4 GHz RTL_SDR sticks cost 30 dollies($$) or so,
 https://www.ebay.com/p/22032165355?iid=335019807057

 
Use a Raspberry Pi or a PC ..
 
I remember when I build a frequency counter with nixy tubes and TTL and
100 kHz crystal took it to work (tube days) it blew the guys away..
Then demoed it at the local ham group connected to the local oscillator
of a shortwave receiver Explained 'subtract the IF frequency from the
indicated frequency to get the real frequency'
Most radios have that now with nice LCD screens and PLL local
oscillators.... Or now even direct sampling like that SDR_RTL stuff.
 
 
I personally would dump the HP thing.. or maybe donate it to some
museum..

>
Haha! No, neither of those things are going to happen, Jan. I'm still
suffering from PTSD at having to toss a Tek 555 into a dumpster 20 years
ago.
Plus...... I've ordered *another* HP boat anchor to assist me to fix this
one! You can never have too many boat anchors (although of course, I
have). :)

Those old monsters were big, heavy, and immensely complex. Not so
reliable.

Nowadays you can get a nice color digital SA with tracking generator
for around $1300 from Amazon. The little handheld toys start around
$60.

I think I'll order one for my new lab. Does anyone have suggestions?
There's a 1.5 GHz Rigol for $1300 and a 3.2g Siglent for $2600.

One key point will be front-end preselection/image rejection. I've
found that even the classic HPs had terrible spruious spikes, at
harmonics of the input frequency. Tek made one SA plugin that had *no*
image rejection. You had to figure it out yourself.