[time-nuts] high rev isolation amps

Dr. David Kirkby (Kirkby Microwave Ltd) drkirkby at kirkbymicrowave.co.uk
Sun Mar 27 07:40:33 EDT 2016


On 27 Mar 2016 03:00, "jimlux" <jimlux at earthlink.net> wrote:
>
> On 3/26/16 3:25 PM, Bruce Griffiths wrote:
>>
>> On Saturday, March 26, 2016 09:30:30 PM Rob Sherwood. wrote:
>>>
>>> You cannot add the directivity and gain.  Doesn't work that way.  Rob,
NC0B.
>>>
>>> Sent from my iPad
>>
>> Minicircuits would disagree with that and its their amplifier.
>>
>> Bruce
>>>
>>>
>
> Well, it's a bit inconsistent between gain, directivity and S parameters.
>
> at 5.4 MHz the ZFL-500LN has the following S parameters @ 15 V in the
.s2p file (there's two units' measurements in the files)
>
> S11 -35.5, -32.1
> S21  29.7,  29.5
> S12 -47.5,- 47.2
> S22 -39.5, -28.3
>
> On the other hand, the data sheet "typical data" says: at 5.4MHz 27.9 dB
gain, 26.5 dB directivity, which Minicircuits defines as isolation-gain
(implying isolation is 54.4)
> http://www.minicircuits.com/pdfs/ZFL-500LN.pdf

I think you are wrong there. From the Touchstone .s2p data you give above:

Isolation (dB) = - |S12|=+47.5 dB.

Gain (dB) = |S21|=+29.7 dB.

Directivity, a term I can't say that I have seen much applied to amplifiers
other than by Minicircuits,  is according to the Minicircuits definition,
the difference between isolation and gain. But I don't see a note of
whether its gain - isolation or isolation-gain.    Assuming the latter, the
directivity would be

47.5-28.7=18.8 dB.

That doesn't agree with your result,  nor does it agree with the
Minicircuits typical value of 26.5 dB.

But the Minicircuits typical gain in the data sheet is 27.94 dB.

I learned from a Minicrcuits sales rep that the "typical" data in their
data sheets is the result of one device taken randomly from the production
line. It is not any sort of average. I find that very odd, and so did the
sales rep, but I was told that was true.

I think it is time to confirm these numbers with Minicircuits,  but I would
not rely on "typical" data if you want a guaranteed isolation.

> Maybe there's some sort of assumption about source/load impedance (e.g.
not necessarily perfect 50 ohms)

I doubt that is so. S-parameters by definition assume that all other ports
are terminated in devices which reflect no power at the system impedance.
In the case of most devices that's 50 Ohm loads, but in the case of mixed
waveguide/coaxial devices waveguide loads would be used.

Minicircuits do a range of high isolation amplifiers.  I assume that they
guarantee figures for those although I have not checked.

Dave


More information about the time-nuts mailing list