开云体育

Mike,Look at the schematic at the end of the July user guide.?The 5v output from the inverter powers the 612 mixers and also feeds an LDO 3v regulator that powers everything else.?



On Fri, 4 Oct 2019 at 4:30 PM, mike miniver<wa7ark@...> wrote: The internal LiOn battery (if connected) is like 3.7 to 4.2V. It is normally charged from externally-supplied USB 5V. This suggests to me that the innards run on something less than 3.7V (most likely 3.3V), meaning that there is an on-board voltage regulator to drop 5V to whatever the internals run on.

My question has to do with what is the maximum voltage that the internal voltage regulator can handle. If doing this, there would be no internal LiOn battery.

Reinier,
Gabriel only pops in now and then to let the group know how things are progressing. She posted that Maxscend rf switches will be used because, "the maxscend switches do not have the shunt diode problem (most RF switch ICs have parasitic diodes from RF input to ground which will start to conduct at lower frequencies), so it has no theoretical lower frequency limit and can be applied at the IF frequency".

My guesses are the Maxscend MXD8661 or MXD8680.

Herb

33 : the mathematical expressions

hello,

Allow us, please, to emphasize that the mathematical
expressions we used for:

30 : our final report 1 - 4 Oct 2019
/g/nanovna-users/message/3989

are just those mentioned at:
16 : 27 September 2019
/g/nanovna-users/message/3161

that is the ones shown at:


sincerely,

gin&pez@arg

33

Yes, "Return Loss" s o m e t i m e s is a misnomer: If used in context with simply two different impedances meeting at a connection.
This can be understood as a complex generator's inner (serial) impedance Z2 being terminated with a (serial) complex load Z1 (parallel to
be recalculated to serial equivalent).

This can be built from lumped components.

Everybody will agree, that there is absolutely no "returning" of a "wave", and no "reflection", in what is just a complex voltage divider.
Physically it is a different (no wave) phenomenon. In such cases, all the following are "misnomers" I n t h e s t r i c t s e n s e of the wording:

Standing "Wave" Ratio,
"Reflection" Factor,
"Return" Loss.


Reflections will, however, appear on Lines: there is a forward wave and a reflected wave.
There is a (wave typical) 180° phase shift by reflection involved, that makes valid

Gamma = (Z1 - Z2) / (Z1 + Z2) (a)

But where there are no waves, there are also no reflections and thus no 180° phase shift. This is why we get a different formula.
Please note, that Z2 above is the line's w a v e impedance - that does not exist in lumped circuits.

There we need to use a different Gamma:

Gamma = (Z1 - Z2*) / (Z1 + Z2) (b) with the asterisk meaning conjugate complex.

Z2 here now means generator's inner impedance (Some call it Thévenin Impedance.)

Unfortunately so many well established papers and publications do not care for that difference between (a) and (b) , just very few do.

And the many careless publications are being quoted and repeated ever so often. That is the problem. (Just the same as with fake news ...)


So - for complex voltage dividers - that is where there are simply no waves involved -
but also if one had previously calculated equivalent impedances on places with waves, like at line ends -
thus whenever we are using a load impedance Z1 and an (equivalent) generator impedance Z2, the above (b) is the correct formula.

The problem with the Smith Chart is, that the Z (or Y) Plane is mapped to the Gamma plane only by formula (a).

It would be extremely bad to "reinvent" the Smith Chart.

So we must live with (a) - like it or not. I see no choice.

The good thing is: The same (Smith chart type) mapping formula can stay, if both (serial) reactances are combined together at the load side.
Then the (thus new) generator impedance is real without reactance and the (thus new) load includes the sum of the two (serial) reactances.
Then the conformal mapping formula (a) can be applied again as usual in the Smith Chart.


We need the voltage divider type formula (b) when dealing with power transfer.


But when will that become common knowledge?

Debating fake physics (calculating with wave formulas where there are no waves and thus no no phase shift) is like fighting against windmills ...


73, Hans
DJ7BA




-----Ursprüngliche Nachricht-----
Von: [email protected] <[email protected]> Im Auftrag von Oristo
Gesendet: Freitag, 4. Oktober 2019 17:32
An: [email protected]
Betreff: Re: [nanovna-users] Return Loss

its become the consensus that return loss is a negative number

IMO, "return loss" is a poor term.
Arguably, >>any<< return is loss, when sending power is the goal. In which case, "return loss" might be power sent (since lost from return)..




--
Diese E-Mail wurde von Avast Antivirus-Software auf Viren geprüft.

DMR_M_10KHz_sept-13-19.dfu
Such as these were for enhancements, in this case 10KHz instead of 50KHz
minimum.
I >>guess<< many enhancements are in Gen Hu Oct 2 firmware"

- Should this be installed before or after the firmware?

Any .dfu installed becomes the only firmware, replacing previously installed firmware.

- Which firmware can be used to set an impedance different than 50 Ohm?

I know of none. No real benefit, since reflection bridge components are fixed for 50 Ohms.
Firmware based on different calibration impedance would improve neither sensitivity nor accuracy.

Steven Maas if I recall in one of his little texts, mentions, I dis like the term return loss.

Its one of those EE crazy terms!

That said, I never discuss return loss with any thought whatsoever. All return loss numbers that ever roll off my lips are positive. Per the IEEE definition. I feel fine with this for the simple reason as follows:

A filter... its passive... generates no power, lousy filter, 2 dB return loss, ugh. Good filter, great match in the pass band, 20 dB return loss. I'll take it.

I just finished a power amplifier design. Output power 500 W and oh by the way, the return loss is 15 dB. Wow.... Is that 15 dB or -15 dB sir? Hey all my power amplifiers are UNCONDITIONALLY stable. I don't build oscillators here, sir!

Alan

32 : a REGION executable
-
also @Gary O'Neil :
27 September 2019 - /g/nanovna-users/message/3259
26 September 2019 - /g/nanovna-users/message/3091
26 September 2019 - /g/nanovna-users/message/3070

Hello,

Allow us, please, to inform you that we just uploaded the version 101 of
REGION executable:



- 131,584 bytes -

Sincerely,

gin&pez@arg

32

Correct.

The measurement conducted is a SPOT NOISE FIGURE measurement. The VNA or a SA may be set into the CW or fixed frequency receive mode. If you desire a swept NF measurement, than perhaps steps may be a small concern but still, that would be easy to average. Yes, conducting NF measure at BAD or bright frequencies is not easy. Try conducting a NF in the middle of the FM radio band. You need to be in a screen room. However, for the most part most of the spectrum is pretty quiet.

The LNA in front of the SA is critical. You need to have a low noise figure to eliminate the SA from contributing to the noise figure value, so called 2nd stage contribution affect and you must have sufficient gain to over come the intrinsic noise power of the SA. You must see a definitive noise power shift on the SA between firing (HOT source) and going COLD (noise source OFF) to determine the Y factor value.

Bandwidth is not an issue as long as there is sufficient noise power visible on the SA. Recall, noise power is directly linked to BW as a 10 log BW value. Our assumption here is that the BW of the DUT is larger than the BW of the measurement system. Measurement of NF of narrow band devices is difficult simply because there is insufficient noise power!

Alan

Thanks for all these info, compliments.

This interests me:
"DMR_M_10KHz_sept-13-19.dfu
Such as these were for enhancements, in this case 10KHz instead of 50KHz minimum.
I >>guess<< many enhancements are in Gen Hu Oct 2 firmware"
- Should this be installed before or after the firmware?

- Which firmware can be used to set an impedance different than 50 Ohm?

Need to place the file with the electrical circuit separately.

added to Wiki Main Menu:
/g/nanovna-users/wiki

Fit a RCA/0.1" pins/BNC/SO239/SMA (depends what freq range and what is to
hand) to either side of a pcb and the DUT (Device Under Test) in between.

I looked for BNC with 0.1" spacing, but so far found only this:

Need to place the file with the electrical circuit separately. Some do not suspect that it is in the instruction manual.
Many questions will disappear.

31 : On the Uncertainties of the "Standards" - Part II
-
23 : On the Uncertainties of the "Standards" - Part I - 30 September 2019
/g/nanovna-users/message/3517

@ Jeff Anderson
/g/nanovna-users/message/3294 - 30 September 2019
/g/nanovna-users/message/3294 - 28 September 2019

Hello Jeff.

We are terribly sorry for this delayed reply - please, accept our apologies.
Our excuses have to obviously do with the needed work for:

30 : our final report 1 :
/g/nanovna-users/message/3989 - 4 October 2019

Well, after all that said, here are our answers:

JA : "For the Open and Short standards, in what terms (or parameters)
do you define each standard's two uncertainties?"

GZ : Let's look at the [inut.txt] file contents, line-by-line:

(1st):1
(2nd):101
(3rd): -0.01 0.029 -0.01 2. 2.
(4th):1

of which their meaning is as follows:

(1st):(Fields:1):

A flag defining what will be computed and extracted:

0 : complex Gamma
1 : complex Z

(1st) : in the test file [input.txt] : 1

(2nd):(Fields:1):
Number of the above values that will be computed and extracted

(2nd) : in the test file [input.txt] : 101

(3rd):(Fields:5):(From left to Right - below as : a,b,c,d,e):

Uncertainty Data as they are given by the Manufacturer
of the "standards" :

The first 3 fields are for the Magnitude -pure number- of
SHORT, LOAD, and OPEN in this order, and

the last 2 fields are for the Argument -in Degrees-
of SHORT and OPEN, in this order

- there is no need for the Argument of LOAD, because
this is and Undefined one -

(3rd)(a): For the Magnitude of SHORT : this value is its
Lower Error Bound and thus a Negative number must
be given - There is no need for an Upper Error Bound,
because it is always 0, in an attempt to keep the its
Computed Values as close as possible to the the
Boundary of the Unit Circle

(3rd)(a) : in the test file [input.txt] : -0.01

(3rd)(b) : For the magnitude of LOAD : this value is its
Upper Error Bound and thus a Positive number must
be given - There is no need for a Lower Error Bound,
because it is always 0, in order to be the Center of
the Unit Circle which lies at the Origin

(3rd)(b) : in the test file [input.txt] : 0.029

(3rd)(c) : For the magnitude of OPEN : exactly the
same as in (3rd)(a)

(3rd)(c) : in the test file [input.txt] : -0.01

- But, the values in (a) and (c) may differ -

(3rd)(d) : For the argument of SHORT : this value
is the absolute value of its Symmetrical, LOWER
and UPPER bounds and thus a Positive number
must be given

(3rd)(d) : in the test file [input.txt] : 2

(3rd)(e) : For the argument of SHORT : exactly
the same as in (3rd)(d)

(3rd)(e) : in the test file [input.txt] : 2

- the values in (d) and (e) may differ -

(4th) : Inaccuracy of all the readings as
number of questionable units in the Last
Significant Digit

(4th) : in the test file [input.txt] : 1

Best regards,

gin&pez@arg

31

Thanks for the response!

Remember the Electrical delay assumes and ideal 50 transmission line

So, if I
* properly calibrate at the VNA port
* plug in a candidate cable terminated with calibration load
* observe no appreciable reflection

.. then that cable unterminated should be satisfactory for delay testing..

Hello Oristo
Yes just by using the Electrical delay, but there is always a but, the delay in a transmission line is frequency dependent and if the impedance is not precise 50 ohm we might have an impact too. You may also use a phase trace and observe at what frequencies the delay is constant for which the it goes horizontal along thee 0 degree phase line. Changing the delay setting you may find at which specific frequency it crosses 0 degree thus determining the delay at that frequency. Remember the Electrical delay assumes and ideal 50 transmission line so....
Kind regards
Kurt

-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af Oristo
Sendt: 4. oktober 2019 15:30
Til: [email protected]
Emne: Re: [nanovna-users] Cal-Kit Standards' Definitions

every user would need to compute a value for the delay, which would be
different for the value in the cal kit.

If I correctly interpret Kurt, then with appropriate software:
* attach an unterminated length of balanced transmission line
* interactively dial that delay so that Smith plot does not spiral inward

Hi Rune,
I just tested a SFZ450F ceramic filter. First I had set a span of 50kHz and I set 3 markers. Than I changed to 50kHz span and all the markers were gone. If you try to set a new marker in the S21 window the data of the marker appears but you won't see the marker in any of the plots.
Very strange.

Kind regards
Norbert, DG1KPN

What "a can of worms" this subject opens up.?I noticed the Hewlett Packard rarely if ever referred to return loss as a hardware specification; it was always specified in terms of SWR. Personally I prefer using return loss, with a negative sign. I know this results in misunderstandings, such as referring to a low SWR wideband antenna as having a low return loss. IEEE uses the negative return loss as a standard for their publications. I suppose that correct or not, the most used sign (negative) has became the standard. As long as we are all on the same page, it tends to work out.?All of my network analyzers will output either SWR or return loss, and an iPhone app will convert from one standard to the other.

Stuart K6YAZLos Angeles, USA

Another option, If you want to plot smith charts from Touchstone .s1p .s2p files, Simsmith also does this. It is Free. <>
I am eagerly hoping someone integrates this new nanoVNA to output document format like Simsmith, and Zplots type functionality. <> Right now, all we have is screen captures.

Thanks,
David
W0IM

The internal LiOn battery (if connected) is like 3.7 to 4.2V. It is normally charged from externally-supplied USB 5V. This suggests to me that the innards run on something less than 3.7V (most likely 3.3V), meaning that there is an on-board voltage regulator to drop 5V to whatever the internals run on.

My question has to do with what is the maximum voltage that the internal voltage regulator can handle. If doing this, there would be no internal LiOn battery.

You must have a very cheap source of the ADF4350 since MAX2871 is just around 8 USD.
Which RF switch do you use to switch between the si5351 and the ADF4350? Most switches are not specified below a few tens of MHz.
I'm looking for a cheap one!
Thanks, Reinier

Op 4-10-2019 om 17:41 schreef Gabriel Tenma White: