Minicircuits ZFL-500 directivity question

[bitrex]

How is "directivity" defined for these Minicircuits gain blocks?
Datasheet here:

<https://www.mouser.com/datasheet/2/1030/ZFL_500_2b-3305829.pdf>

Across the bandwidth of interest I'm measuring S21 as about 20-something
dB give or take a dB which seems fine, S12 is at -40.

If "directivity" in this context is just S12 then that seems within spec
but if it's |S21 - S12| then something seems wrong with my setup and/or
the amp.

Doing the measurements on a (unfortunately not my personal home-lab
machine!) Keysight 5080.

[bitrex]

On 11/4/2023 4:55 PM, bitrex wrote:
> How is "directivity" defined for these Minicircuits gain blocks?
> Datasheet here:
>
> <https://www.mouser.com/datasheet/2/1030/ZFL_500_2b-3305829.pdf>
>
> Across the bandwidth of interest I'm measuring S21 as about 20-something
> dB give or take a dB which seems fine, S12 is at -40.
>
> If "directivity" in this context is just S12 then that seems within spec
> but if it's |S21 - S12|

Or rather |S12| - |S21|

[John May]

It's just the difference in dB, from the datasheet s21 is about 25dB, directivity about 38db, so s12 is 38dB below s21, or -13dB.

[John May]

That's wrong - would be -(25+38)dB

[bitrex]

Thanks for your reply but I don't quite follow - do you know how they
calculate the ~40dB figures of the top table on page 2?

[John May]

It is just the magnitude of the amplifier isolation (reverse gain) minus the (forward) gain. If reverse gain was -80dB and forward gain 20dB direcivity would be 60dB.

[bitrex]

Gotcha, thanks!

[Phil Hobbs]

Isolation is 1/(reverse gain), and it’s a straight subtraction, so 60 dB is
correct for the wrong reason.

One way of keeping it straight is that an N-dB pad has 2N dB isolation.
(Using the magnitude would predict an isolation of 0 dB.) The idea is to
say how much a mismatch affects the input impedance.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC /
Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics

[bitrex]

Thank you, so just S21 - S12? S21 for the ZFL-500 was about 20 dB and
S21 -40, that would seem to imply 60 dB directivity while the spec says
around 40 in the mid band, seems hard to believe it's beating the spec
by that much.

But I'm unsure how much input power affects this, does the datasheet
table imply these measurements are taken at the 1dB compression point?

[bitrex]

On 11/6/2023 11:24 AM, bitrex wrote:

>> One way of keeping it straight is that an N-dB pad has 2N dB isolation.
>> (Using the magnitude would predict an isolation of 0 dB.)  The idea is to
>> say how much a mismatch affects the input impedance.
>>
>> Cheers
>>
>> Phil Hobbs
>>
>
> >
> >
>
> Thank you, so just S21 - S12? S21 for the ZFL-500 was about 20 dB and
> S21 -40

Sorry S12 was about -40, rather

[bitrex]

On 11/6/2023 11:24 AM, bitrex wrote:

>> Isolation is 1/(reverse gain), and it’s a straight subtraction, so 60
>> dB is
>> correct for the wrong reason.
>>
>> One way of keeping it straight is that an N-dB pad has 2N dB isolation.
>> (Using the magnitude would predict an isolation of 0 dB.)  The idea is to
>> say how much a mismatch affects the input impedance.
>>
>> Cheers
>>
>> Phil Hobbs
>>
>
> >
> >
>
> Thank you, so just S21 - S12?

To clarify I mean the log magnitudes here, as a VNA so conveniently can
display for us

[Phil Hobbs]

A reverse gain of-40 dB is an isolation of +40 dB, so the directivity is
40-20 = 20 dB.

If you put a short or open on the output, it will reflect the full output
power back into the output port. That’s 20 dB above the input power, so
assuming a matched input and 40 dB isolation, you’d get an input return
loss of 20 dB. That’s what the directivity expresses.

[bitrex]

On 11/6/2023 2:05 PM, Phil Hobbs wrote:
> bitrex <us...@example.net> wrote:
>> On 11/6/2023 11:24 AM, bitrex wrote:
>>
>>>> One way of keeping it straight is that an N-dB pad has 2N dB isolation.
>>>> (Using the magnitude would predict an isolation of 0 dB.)  The idea is to
>>>> say how much a mismatch affects the input impedance.
>>>>
>>>> Cheers
>>>>
>>>> Phil Hobbs
>>>>
>>>
>>>>
>>>>
>>>
>>> Thank you, so just S21 - S12? S21 for the ZFL-500 was about 20 dB and
>>> S21 -40
>>
>> Sorry S12 was about -40, rather
>>
>>
> A reverse gain of-40 dB is an isolation of +40 dB, so the directivity is
> 40-20 = 20 dB.
>
> If you put a short or open on the output, it will reflect the full output
> power back into the output port. That’s 20 dB above the input power, so
> assuming a matched input and 40 dB isolation, you’d get an input return
> loss of 20 dB. That’s what the directivity expresses.
>
> Cheers
>
> Phil Hobbs
>

I see why JL loves his time domain.

The 5080 is a nice instrument, it seems to boot to Windows 7. Crashes to
Windows 7, too..

[Simon S Aysdie]

They give S-params for 2 units. Conveniently, they are in dB/angle. The table just says 38 dB, more or less for directivity. By a quick look at the S-params and comparing to the datasheet table, I think they are just saying D = |S12| (dB), which you noted.

Example:
For "unit 1" and at 100 MHz, forward gain is +24.47859 dB.
For "unit 1" and at 100 MHz, reverse gain is -38.36183 dB.

Because the "directivity" looks to be about same as |S12|, it is safe to say they mean D = |S12| (dB)

!Freq S11:SOLT2(ON) S21:SOLT2(ON) S12:SOLT2(ON) S22:SOLT2(ON)
# Hz S dB R 50
100000000 -3.318152e+001 -2.120138e+001 2.447859e+001 -1.975411e+001 -3.836183e+001 -1.134106e+001 -2.672768e+001 -1.570933e+002

[Simon S Aysdie]

I should say:
D = -|S12| (dB)

[bitrex]

Thanks Simon, yes that seems right, this amp was bang on within spec
otherwise so I'd have been surprised if it was way off there.

They're nice lil modules, I didn't have this purpose in mind but I
wonder if they make a LNA suitable to replace the shit ass front end on
my DTV antenna, I live in a fringe area so reception varies by the day.
and I'm in a condo so mounting stuff outdoors not really feasible

[Fred Bloggs]

Your reception is probably being degraded by suboptimal antenna pointing, the sun, and unpredictable multi-path signal impingement [ coastal is littoral causes no end of propagation anomalies ]. It doesn't cost the manufacturer anything to provide a high performance front end amplifier- all of which intended for satellite reception are LN-A's. The customers pay for everything, do you think it's wise then for them to provide junk performance parts making customers into ex-customers?

[bitrex]

It doesn't?

all of which intended for satellite reception are LN-A's. The
customers pay for everything, do you think it's wise then for them to
provide junk performance parts making customers into ex-customers?

Could apply the same argument to many other industries...and conclude
all products are high-quality because otherwise the manufacturer might
lose customers.

Hey, I have a cool idea, I'll see if I can put it on the 5080 next time
I have a chance and see how it compares to a Minicircuits gain block for
similar bandwidth, I think I can get a hold of one..

[upsid...@downunder.com]

On Wed, 8 Nov 2023 11:11:31 -0500, bitrex <us...@example.net> wrote:

>They're nice lil modules, I didn't have this purpose in mind but I
>wonder if they make a LNA suitable to replace the shit ass front end on
>my DTV antenna, I live in a fringe area so reception varies by the day.
>and I'm in a condo so mounting stuff outdoors not really feasible

That ZFL-500 is not very suitable as antenna amplifier for DTV. First
of all it is quite noisy, noise figure NF = 4.2 dB (typical). These
days a good antenna amplifier the NF should be below 3 dB. That
amplifier is specified only up to 500 MHz, thus most UHF channels are
out of range. However, it might still work since the specs are good at
500 MHz, wonder why the specs ends at 500 MHz.

An antenna amplifier should be installed _at_ the antenna, thus the
cable between amplifier and DTV receiver can be long.

You seem to suffer from tropospheric fading and the signal nulls can
be quite deep and the signal is still momentary lost no matter what
antenna amplifier you use. Only a proper outdoor antenna will really
help.

[bitrex]

On 11/8/2023 5:02 PM, upsid...@downunder.com wrote:
> On Wed, 8 Nov 2023 11:11:31 -0500, bitrex <us...@example.net> wrote:
>
>> They're nice lil modules, I didn't have this purpose in mind but I
>> wonder if they make a LNA suitable to replace the shit ass front end on
>> my DTV antenna, I live in a fringe area so reception varies by the day.
>> and I'm in a condo so mounting stuff outdoors not really feasible
>
> That ZFL-500 is not very suitable as antenna amplifier for DTV. First
> of all it is quite noisy, noise figure NF = 4.2 dB (typical). These
> days a good antenna amplifier the NF should be below 3 dB. That
> amplifier is specified only up to 500 MHz, thus most UHF channels are
> out of range. However, it might still work since the specs are good at
> 500 MHz, wonder why the specs ends at 500 MHz.

Yeah it has gain for days. S21 is above 20dB well past 1 GHz.

> An antenna amplifier should be installed _at_ the antenna, thus the
> cable between amplifier and DTV receiver can be long.
>
> You seem to suffer from tropospheric fading and the signal nulls can
> be quite deep and the signal is still momentary lost no matter what
> antenna amplifier you use. Only a proper outdoor antenna will really
> help.

I'll see if high pressure systems seem to make it better...atm with a
planar antenna and associated amp I'm getting a station with a purported
field strength of -70 dBm at 36 miles no problem if I kinda point it in
that direction, but a station at 16 miles -50 dBm doesn't come in at all
/shrug

[Phil Hobbs]

(Supernews doesn't seem to be registering my 3 PM EST reply just
now--apologies for duplication if it shows up eventually.)

It makes no sense to take the absolute value of a quantity expressed in
decibels--that's saying that there's no difference between, say, a 100X
voltage amp (+40 dB) and a 40-dB attenuator (-40 dB).

The directivity is just |S12 * S21|, i.e. the effects at one port of a
short-circuit load reflecting all of the signal backwards through the
network. In decibels, it's

D = 20 log(|S12 * S21|). Symmetrical. No big mystery.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics

Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

[Fred Bloggs]

The business is going to provide the best part suited for the application and price range. It doesn't take state-of-the-art.

Your reception is problematic because of the location and other factors. Nothing by way of more expensive parts is going to fix it.

[Phil Hobbs]

It isn't the absolute value in decibels. The directivity is just the
magnitude of S21*S12. Otherwise you get the wrong answer for attenuators.