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.
Minicircuits ZFL-500 directivity question
Started by ●November 4, 2023
Reply by ●November 4, 20232023-11-04
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|
Reply by ●November 4, 20232023-11-04
On Saturday, November 4, 2023 at 8:59:41 PM UTC, bitrex wrote:> 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|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.
Reply by ●November 4, 20232023-11-04
On Saturday, November 4, 2023 at 10:01:15 PM UTC, John May wrote:> On Saturday, November 4, 2023 at 8:59:41 PM UTC, bitrex wrote: > > 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| > 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.That's wrong - would be -(25+38)dB
Reply by ●November 5, 20232023-11-05
On 11/4/2023 6:34 PM, John May wrote:> On Saturday, November 4, 2023 at 10:01:15 PM UTC, John May wrote: >> On Saturday, November 4, 2023 at 8:59:41 PM UTC, bitrex wrote: >>> 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| >> 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. > > That's wrong - would be -(25+38)dBThanks 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?
Reply by ●November 5, 20232023-11-05
On Sunday, November 5, 2023 at 10:18:21 PM UTC, bitrex wrote:> On 11/4/2023 6:34 PM, John May wrote: > > On Saturday, November 4, 2023 at 10:01:15 PM UTC, John May wrote: > >> On Saturday, November 4, 2023 at 8:59:41 PM UTC, bitrex wrote: > >>> 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| > >> 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. > > > > That's wrong - would be -(25+38)dB > 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?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.
Reply by ●November 5, 20232023-11-05
On 11/5/2023 7:53 PM, John May wrote:> On Sunday, November 5, 2023 at 10:18:21 PM UTC, bitrex wrote: >> On 11/4/2023 6:34 PM, John May wrote: >>> On Saturday, November 4, 2023 at 10:01:15 PM UTC, John May wrote: >>>> On Saturday, November 4, 2023 at 8:59:41 PM UTC, bitrex wrote: >>>>> 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| >>>> 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. >>> >>> That's wrong - would be -(25+38)dB >> 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? > > 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.Gotcha, thanks!
Reply by ●November 6, 20232023-11-06
bitrex <user@example.net> wrote:> On 11/5/2023 7:53 PM, John May wrote: >> On Sunday, November 5, 2023 at 10:18:21 PM UTC, bitrex wrote: >>> On 11/4/2023 6:34 PM, John May wrote: >>>> On Saturday, November 4, 2023 at 10:01:15 PM UTC, John May wrote: >>>>> On Saturday, November 4, 2023 at 8:59:41 PM UTC, bitrex wrote: >>>>>> 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| >>>>> 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. >>>> >>>> That's wrong - would be -(25+38)dB >>> 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? >> >> 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. > > Gotcha, thanks! >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
Reply by ●November 6, 20232023-11-06
On 11/6/2023 8:25 AM, Phil Hobbs wrote:> bitrex <user@example.net> wrote: >> On 11/5/2023 7:53 PM, John May wrote: >>> On Sunday, November 5, 2023 at 10:18:21 PM UTC, bitrex wrote: >>>> On 11/4/2023 6:34 PM, John May wrote: >>>>> On Saturday, November 4, 2023 at 10:01:15 PM UTC, John May wrote: >>>>>> On Saturday, November 4, 2023 at 8:59:41 PM UTC, bitrex wrote: >>>>>>> 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| >>>>>> 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. >>>>> >>>>> That's wrong - would be -(25+38)dB >>>> 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? >>> >>> 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. >> >> Gotcha, thanks! >> > > 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? 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?
Reply by ●November 6, 20232023-11-06
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 -40Sorry S12 was about -40, rather
