开云体育

Mega and Giga are STANDARD abbreviations for powers of 10. Mega = 1E+6
Giga = 1E+9. Pico - 1E-12. Micro = 1E-6. These are all STANDARD
engineering notations! Oh, yes, and picoFarad = micromicroFarad as
exponents add. And long ago when SI was standardized, cps = Hz or Hertz
(for Mr. Hertz). Too bad Mr. Maxwell of the four equations that rule
electromagnetics was relagated to some nebulous unit of magnetism,
specifically magnetic flux, abbreviated (standard) Mx.

Which is more convenient in both notation and calculations to express the
speed of electromagnetic radiation in a vacuum:

c = 3E8 M/s

or

c = 300,000,000 M/s

I'll take the latter, please.

BTW: lower case c is the standard notation for the speed of
electromagnetic radiation in a vacuum.
(usually, but not always) upper case M is standard for Meters.
lower case s is standard for seconds.

If we are going to play this RF game complete with Smith Charts, please
learn and properly use the language of the discipline.

Dave - W?LEV

On Wed, Oct 7, 2020 at 7:49 PM Stephen Laurence <Gaslaurence@...>
wrote:

Yes, but there is an inconsistency, as I do not recall any scientist named
Mega or Giga, so it should be gHz and mHz........

Errrrr....

sTEVE l. (Woops Steve is my name so it should be Steve)

--
*Dave - W?LEV*
*Just Let Darwin Work*

On Wed, Oct 7, 2020 at 09:26 PM, Hasan Schiers N0AN wrote:

. ...and it is kHz
not KHz, as capital K is for Kelvin, not 1000

In the example, k is correct but K wouldn't lead to confusion because the capital rule applies to *units* (Hz in this example) and k is a *prefix* (scaling factor if you will).

Raymond

Your question has insufficient information.? Explain what you are doing and what your observations are.? Not getting any results makes no sense.? What readings don't change?? What are the parameters and what is the device?

On Wed, Oct 7, 2020 at 09:52 PM, Hasan Schiers N0AN wrote:

the capital M is to properly differentiate between lower case m = milli or
1/1000 and M Mega, one million.

Google Engineering Notation for Frequency and you will find the standards.

That is true. Engineering notation conforms to the SI unit system. Engineering notation applies to *prefixes*, not *units*. Hertz is a *unit* in the SI, milli, mega etc. are *prefixes* (to units).

Raymond

Also note:

dB deci-Bell Not DB
dBm decibels above one milliwatt Not DBM

Hasan


On Wed, Oct 7, 2020 at 2:53 PM Raymond Domp Frank <hewpatek@...>
wrote:

On Wed, Oct 7, 2020 at 09:49 PM, Stephen Laurence wrote:

Yes, but there is an inconsistency, as I do not recall any scientist

named

Mega or Giga, so it should be gHz and mHz........

Errrrr....

No, there isn't. The rule applies to *units* in the SI. Mega, giga, milli,
pico, deci etc. are *prefixes*.

Raymond

On Wed, Oct 7, 2020 at 09:49 PM, Stephen Laurence wrote:

Yes, but there is an inconsistency, as I do not recall any scientist named
Mega or Giga, so it should be gHz and mHz........

Errrrr....

No, there isn't. The rule applies to *units* in the SI. Mega, giga, milli, pico, deci etc. are *prefixes*.

Raymond

the capital M is to properly differentiate between lower case m = milli or
1/1000 and M Mega, one million.

Google Engineering Notation for Frequency and you will find the standards.

73, N0AN
Hasan


On Wed, Oct 7, 2020 at 2:49 PM Stephen Laurence <Gaslaurence@...>
wrote:

Yes, but there is an inconsistency, as I do not recall any scientist named
Mega or Giga, so it should be gHz and mHz........

Errrrr....

sTEVE l. (Woops Steve is my name so it should be Steve)

Yes, but there is an inconsistency, as I do not recall any scientist named Mega or Giga, so it should be gHz and mHz........

Errrrr....

sTEVE l. (Woops Steve is my name so it should be Steve)

By SI convention, all units named after people are to be spelled with an uppercase first letter,
so Hz, not hz (the unit Hz is named after Heinrich Hertz). Also mHz, kHz, MHz, GHz, *not* mhz, khz, Mhz, GHz but cm, dm, km, ml, etc. Same for e.g. V (Volta), A (Ampère), Ohm, F (Faraday), H (Henry), B (Bell).

Raymond

As in mA as well, thanks for the reminder

and since Hertz is a name, the H should be capitalized. ...and it is kHz
not KHz, as capital K is for Kelvin, not 1000 ...although in resistance
values it is not as likely to lead to confusion.

So, it is:

GHz (Giga Hertz)
MHz (Mega Hertz)
kHz (kilo Hertz)
mA (milli-Ampere) and ..
pA (pico-Ampere)
pF (pico-Farad)


At least that's how I recall them.

73, N0AN
Hasan

Martin, I upgraded my SAA-2N FW to ver 20200926, works fine on
instrument, however I have the same issue with NanoVNA-Saver, as well
as some other anomalies (program sweep runs on 1 step but hangs on 2
steps). NanoVNA-QT works OK except that low freq stimulus won't
calibrate below 100Khz, but is fine to 3000Mhz (101 steps).
I haven't seen but one other post reporting the same issue with Saver
and the SAA-2N..

-----------------------------------------From: "Martin V via
groups.io"
To: [email protected]
Cc:
Sent: Wednesday October 7 2020 10:31:26AM
Subject: Re: [nanovna-users] nanovna saver on saa-2N

I have the same issue.
VNA 2_2 Version: git-20200617-1a9a11d
Build Time: Jun 19 2020

-=-=-=-=-=-=-=-=-=-=-=-
Groups.io Links: You receive all messages sent to this group.
View/Reply Online (#17984):

Links:
------
[1] /mt/77047112/4866111
/mt/77047112/4866111
/> -=-=-
Learn about the NanoVNA in the following forum areas:
Documentation Update Files: </br<></a><a href=
[2] /g/nanovna-users/wiki
/g/nanovna-users/wiki
/> -=-=-
Group Owner: [email protected]
Unsubscribe: </br<></a><a href=

What about a balun. If I connect two wires directly to any measuring device without a balun, I don't get any results. Readings don't change no matter what I do.

Dear Alan,

Thankyou. We are a little way down the message ladder, so this refers to several messages back.

Thankyou for an injection of sanity (to me, anyway). I will continue to play, plod on, and learn from it all. I am still in awe of the whole process:- that something costing as little as ?50/§60 can give so much information and learning potential.

In my whole personal history of self-teaching in electronics, the nanovna saga beats the occasion when I got an Army Surplus Creed teletype working on an HRO ww2 reciever through a home made terminal unit decoding Roeters News about the 1967 Czech revolt which was reported in The Times newspaper the next day (I was 17).

I should be greatful that my career does not depend on understanding all this stuff. In my work, I just put people to sleep and (usually) woke them up again (ok, a joke I was somewhat better than that in my career as an anaesthetist).

Steve L

Jeff,

interesting question that you ask. I remember, back when I had to implement TOSM (=SOLT) calibration with arbitrary standards, including a fully-known but non-ideal through standard, I was too lazy to derive individual equations for the error terms. I just hacked everything into matrix form as recommended by Prof. Schiek [1] and let the computer solve it numerically. But I'm going to have a look at what you wrote on your blog / website with respect to handling a non-ideal through and I'll come back to this discussion thread. Just give me a few days since it will require some thinking and deriving of equations...

Regards
Christian

[1] In his book: B. Schiek, "Grundlagen der Hochfrequenzmesstechnik", Springer-Verlag, 1999. I don't think there's an English edition, though.

To avoid confusion, please note (conform) to the following *standard*
notation in emails:

mHz or mhz = *MILLI*hertz = 1E-3 Hz (subsonic)

MHz or Mhz = *MEGA*hertz = 1E+6 Hz

GHz or Ghz = *GIGA*hertz = 1E+9 Hz

This notation is standard throughout the communications, RF and microwave,
EMC/RFI, and RCS..... disciplines. Please adhere to this notation to avoid
confusion.

None of the NANOVNA's cover into the millihertz (subsonic / sudio) range!!!

Dave - W?LEV

Thanks very much, Christian. That is a great help.

Pivoting slightly...I have a question for the group regarding SOLT errors and calibration, but this has to do with the THRU standard, rather than the Short, Open, or Load standards...

Specifically, my question is: what is the math to de-embed the effects of a THRU standard with non-zero delay (e.g. a barrel connector) from DUT measurements made after the THRU has been removed at the completion of SOLT calibration?

Keysights' and Ryttings' papers were invaluable regarding my understanding of 12-term error correction, but...they typically assume that the THRU standard is ideal -- that is, it is assumed to be of zero length. (If I am mischaracterizing their math, please correct me!).

Rune Broberg (in conversations with him regarding his THRU de-embedding math in his nanoVNA-Saver software) pointed me to the reference he had used, which was the user manual for a GUI application called "DeEmbed," by Schreuder Electronics ().

When I reviewed the math in this manual, it looked to me like there were some errors in its correction formulas.

I describe my take on these errors here: , but I don't know if my "corrected" math is, indeed, correct. I'm having a difficult time finding sources that are detailed enough (or that I can understand) regarding this topic. Dunsmore's first edition of "Handbook of Microwave Component Measurements," for example, wasn't as much of a help as I was hoping it would be.

Anyway, I'm hoping that someone here might have some insight into this particular topic of SOLT calibration.

Thanks,

- Jeff, k6jca

I have the same issue.
VNA 2_2 Version: git-20200617-1a9a11d
Build Time: Jun 19 2020

Be careful with large plugs. I already broke. I had to temporarily rewire from another port.

I m

The The nand oand noonioonno noo ? in

73,
KB5WCK

________________________________
From: [email protected] <[email protected]> on behalf of Rainer via groups.io <rsfoto@...>
Sent: Tuesday, October 6, 2020 3:38:17 PM
To: [email protected] <[email protected]>
Subject: [nanovna-users] Recomendation

Hi,

As this group discusses the Nano VNA I would like to ask if somebody would recommend me a group for Antenna building.

Thanks Rainer

John's and Jeff's posts make me reconsider my previously posted plot. Unfortunately, it contained an error - or rather something I forgot about my old MATLAB script. For that plot I calculated an average error over several phase differences between load standard and DUT. However, for this investigation it's of course more interesting to see the *maximum* (and not average) error that can occur. I was falsely reassured by seeing the ca. 1 dB (as error) that was mentioned (as uncertainty) in a previous post - and I didn't check my script further. Sorry about that!

Fortunately, as John points out for this rather simple case of just considering the error caused by load mismatch it's possible to find that the maximum error (in linear magnitude) happens when the phase difference between load and DUT is either 0° or 180°. Therefore, I re-simulated; this time only for these two phase relations.

My updated plots are attached. In those you can see the errors already calculated by John for a 30 dB return loss standard and a 20 dB return loss DUT: Ca. 2.4 dB (-17.6 dB instead of -20 dB S11) and ca. -3.3 dB (-23.3 dB instead -20 dB).

Jeff, as you point out, when the load has the same reflection as the DUT and they have the same phase, this is in effect the same as measuring the load standard again after calibration. From that, in theory, you would measure zero reflection, i.e. -Inf dB. (This is also indicated in the plot.) In practice, of course, noise etc. will prevent you from measuring a reflection of exactly 0. In contrast, if the DUT has the same reflection magnitude but the opposite phase, you measure twice (in linear scale) the actual reflection of the DUT, i.e., 6 dB too much.

I hope that resolves the confusion that my previous post has unfortunately caused.

Regards
Christian