|
Re: How low should a return loss at CH0 be?
On Fri, Jan 29, 2021 at 11:38 AM, joe gay wrote:
This is not the case with my example. No common mode present. It was done stand-alone on battery power with no PC connection. The feedline was 50 feet of RG-8X with a 75 ohm resistive load on the end. The reason for the wavy SWR is that the actual characteristic impedance of RG-8X is not equal to the "nominal impedance" of 50 ohms. The characteristic impedance varies considerably from 1 MHz to 30 MHz and is a complex impedance with a small reactive component that also varies with frequency. The end result is that the degree of mismatch varies with frequency at the load end and this is observed at the other end of the cable as a "wavy" SWR plot. It slopes downward due to attenuation in the transmission line. You will get a similar result with any dielectric type transmission line. Another measurement using 3M of RG-316 with SMA connectors and a 100 ohm SMD resistor load is below. A battery-powered NanoVNA-H4 with reference plane at NanoVNA SMA connectors was used. Why not try the same test conditions as I used with a piece of your coaxial cable and see what you observe. Roger |
|
Re: [nanovnav2] PROCEDURE FOR MEASURING CMCs AND OTHER 2-TERMINAL DEVICES WITH THE NANOVNAs
Reply went to the wrong group, it's not a good practice to set the
toggle quoted message
Show quoted text
"reply to" field and then add all others in the reply and CC (especially if those people in the CC do not want their email address broadcasted to the public group, add them in the BCC). There should be a note in the beginning that the Nano itself should be left in an uncalibrated state or calibrated for the desired range. Measure just the parallel line without the core to determine the loss of the line? On Fri, 29 Jan 2021 at 23:25, David Eckhardt <davearea51a@...> wrote:
With all my recent measurements of CMCs, I've been encouraged by several |
|
Re: PROCEDURE FOR MEASURING CMCs AND OTHER 2-TERMINAL DEVICES WITH THE NANOVNAs
Setups HAD to change. The initial measurements were made with the HP
8753C. Those I put out today with the procedure are made using the NANOVNA and SAVER. Of course the setups are DIFFERENT!!!!!! ENTIRELY DIFFERENT INSTRUMENTS!!!!!! Dave - W?LEV On Fri, Jan 29, 2021 at 11:02 PM Mel Farrer via groups.io <farrerfolks= [email protected]> wrote: David, thanks for the good work, but I have one question,-- *Dave - W?LEV* *Just Let Darwin Work* |
|
Re: How low should a return loss at CH0 be?
Interesting discussion, re: return loss. There is a considerable
difference between what can be obtained on the lab, or test, bench and an actual field installation. When discussing RTL measurement in actual terms, a 40dB return loss, being a ratio of incident (forward) to return (reflected) power, would represent a ratio of 10,000:1, or in this case, reflected power for a 1W transmission would be .0001W or 0.1mw. For a 20dB RTL (100:1) reflected power would be 0.01W or 10mW. 20dB RTL in most commercial transmission systems is considered pretty good. A very good basic quick note on this is from Anritsu: ,system%20limit%20for%20a%20cable%20and%20antenna%20system. Nothing wrong with going for the most precision you can get, but sometimes it can be very frustrating trying to hold very low reflection numbers (RTL >20-30dB) in the field particularly when operating in the GHz range. As Jim notes even the test instrumentation oftentimes limits the measurement capability. Just some thoughts.....Ted (KD7AQO) -----------------------------------------From: "Jim Barbour" To: [email protected] Cc: Sent: Friday January 29 2021 11:23:29AM Subject: Re: [nanovna-users] How low should a return loss at CH0 be? Another way to put it in perspective-the Anritsu FDR's bottom out at 54 dB RL. That's as far as it can measure. So 40 dB is really good, and should be enough for just about anybody. On Fri, Jan 29, 2021 at 08:52 AM, wrote: > > The required ultimate return loss measurement accuracy depends on what you > want to do. For just looking for 50 ohm antenna matching.25-30 dB return loss > measurement capability is sufficient. For measuring low or high component > impedance, the more ultimate capability of the instrument the better. The > farther away you get from 50 ohm the worse the measurement accuracy. 5 ohms > to 1000 ohms is the realistic usable impedance range for a 50 ohm reference > bridge. Look at a 50 ohm referenced Smith chart and see the return loss > difference between a 500 ohm and 1000 ohm impedance. > > The accuracy of the resistors in the nano input bridge circuit is first item > effecting accuracy. The bridge resistors shunt parasitic capacitance reduces > higher frequency accuracy. Nano PCB also contributes some accuracy > limitations. Harmonic sensing has less signal reducing ultimate capability. > Calibration nulls out some of these effects but calibration compensation > degrades the farther away from 50 ohms, short, and open calibration points the > load impedance is. > > Spend $10-$15 on a good 50 ohm calibration load from a reputable manufacturer > if you are working above HF bands. Some cheap ones are really bad at higher > frequencies. Calibration is only as good as your references. > Links: ------ [1] /g/nanovna-users/files [2] /g/nanovna-users/leave/defanged |
|
Re: PROCEDURE FOR MEASURING CMCs AND OTHER 2-TERMINAL DEVICES WITH THE NANOVNAs
Mel Farrer
David, thanks for the good work, but I have one question,?
toggle quoted message
Show quoted text
1.? With all of the equipment and access to the testing site, did you make the measurements at different times and different calibrations?? Sort of defeats the accuracy right off the get go.? In my book any serious engineering comparisons like that, are done at one sitting, N O changes in setup PERIOD.? Sorry, my stick. Mel, K6KBE On Friday, January 29, 2021, 02:25:24 PM PST, David Eckhardt <davearea51a@...> wrote:
With all my recent measurements of CMCs, I've been encouraged by several readers to put together a procedure for the various measurements I've made with the HP 8753C, but using the NANOVNAs and SAVER.? The attachment is the first installment which presents a procedure using the NANOs and SAVER to measure the DM loss through any CMC.? It is also useful for measuring other two-terminal transmissive devices such as filters, attenuators, active stages, and....... Have a look at my first cut.? I'd invite anyone to make suggestions to make things clearer and/or correct any errors I've made in the write-up.? Have a read of the attachment. The next installment will address measuring the bulk CM impedance presented by the CMCs using the NANOs and SAVER (another day, please) which have been previously measured using the HP equipment. Dave - W?LEV |
|
PROCEDURE FOR MEASURING CMCs AND OTHER 2-TERMINAL DEVICES WITH THE NANOVNAs
With all my recent measurements of CMCs, I've been encouraged by several
readers to put together a procedure for the various measurements I've made with the HP 8753C, but using the NANOVNAs and SAVER. The attachment is the first installment which presents a procedure using the NANOs and SAVER to measure the DM loss through any CMC. It is also useful for measuring other two-terminal transmissive devices such as filters, attenuators, active stages, and....... Have a look at my first cut. I'd invite anyone to make suggestions to make things clearer and/or correct any errors I've made in the write-up. Have a read of the attachment. The next installment will address measuring the bulk CM impedance presented by the CMCs using the NANOs and SAVER (another day, please) which have been previously measured using the HP equipment. Dave - W?LEV |
|
Re: How low should a return loss at CH0 be?
Hi Roger,
I got the same pattern with both a SAA-2N & a AA-600 rigexpert over a similar range (1-30 mhz) when connected to a PC. I did both again using same no of points, same cal plane, etc but on battery power - wavy lines gone. I concluded that there was some kind interference from the PC. I also had anomalous R & X values while on the PC. As soon as you connect to a PC, one side of your DUT is automatically connected to your home power company ground system. Common mode response is likely the issue? Joe |
|
Re: How low should a return loss at CH0 be?
Another way to put it in perspective-the Anritsu FDR's bottom out at 54 dB RL. That's as far as it can measure. So 40 dB is really good, and should be enough for just about anybody.
toggle quoted message
Show quoted text
On Fri, Jan 29, 2021 at 08:52 AM, <roncraig1@...> wrote:
|
|
I'm also wondering about these. I seem to be seeing that maybe the SAA2N is a little better for things like checking cavities and the like, which is one primary use I would have. True/false? Opinions?
toggle quoted message
Show quoted text
On Tue, Jan 26, 2021 at 01:11 AM, John Gord wrote:
|
|
Re: How low should a return loss at CH0 be?
The required ultimate return loss measurement accuracy depends on what you want to do. For just looking for 50 ohm antenna matching.25-30 dB return loss measurement capability is sufficient. For measuring low or high component impedance, the more ultimate capability of the instrument the better. The farther away you get from 50 ohm the worse the measurement accuracy. 5 ohms to 1000 ohms is the realistic usable impedance range for a 50 ohm reference bridge. Look at a 50 ohm referenced Smith chart and see the return loss difference between a 500 ohm and 1000 ohm impedance.
The accuracy of the resistors in the nano input bridge circuit is first item effecting accuracy. The bridge resistors shunt parasitic capacitance reduces higher frequency accuracy. Nano PCB also contributes some accuracy limitations. Harmonic sensing has less signal reducing ultimate capability. Calibration nulls out some of these effects but calibration compensation degrades the farther away from 50 ohms, short, and open calibration points the load impedance is. Spend $10-$15 on a good 50 ohm calibration load from a reputable manufacturer if you are working above HF bands. Some cheap ones are really bad at higher frequencies. Calibration is only as good as your references. |
|
Re: Display freezes moving cursor with touchscreen
#nanovna-v2
#crash
Hello,
Where can I find latest Nanovna v2 3.2 inch firmware? On the github repository the latest is dated 20201013, while it looks more recent versions are available, like 20201122. Thanks, Simone |
|
Re: Where to order a VNA-F?
Christoph,
toggle quoted message
Show quoted text
Aliexpress, vendor is Deepelec, this is the official vendor, everyone else is a clone on Aliexpress or anywhere else. Stay safe. John VE7KKQ On Thu, Jan 28, 2021 at 2:54 PM Christoph Ratzer <ratzer@...> wrote:
As a newcomer, is it permissible to ask the question, which has certainly |
|
Re: Finally getting into my Gecko, only two traces
#newbie
Use menu to go to DISPLAY then TRACES then select the other two which are
toggle quoted message
Show quoted text
off. On Thu, 28 Jan 2021, 00:09 Bear Albrecht, <W5VZB.NM@...> wrote:
I finally started learning about this thing and all the videos I've seen |
|
Where to order a VNA-F?
Christoph Ratzer
As a newcomer, is it permissible to ask the question, which has certainly been asked many times, where - from a European perspective - it makes most sense to order a VNA-F? There are so many offers from Ali, ebay and other platforms, you have to know your way around...
Many thanks and greetings from Austria 73 Christoph, OE2CRM ¡ª |
|
Re: How low should a return loss at CH0 be?
The following numbers are from
toggle quoted message
Show quoted text
Return Reflection VSWR Transmitted Loss Coefficient Power 10 dB 0.316 1.92 90% 20 0.1 1.22 99% 30 0.032 1.07 99.9% Twenty dB return loss means 99% of the incident power is transmitted, practically all of it. The payback for working for higher return loss is very small. Dave WA8YWQ On 2021-01-28 14:00, Mel Farrer via groups.io wrote:
Consider the magnitude of that measurement in terms of %. For practical uses, anything over -40 dB is meaningless unless you are comparing a part specification. Like it is supposed to be >-45 dB. Yes, my cal gives me >-60 dB, but I don't ever need it. Please don't get blinded by specifications. |
|
Re: How low should a return loss at CH0 be?
On Thu, Jan 28, 2021 at 01:27 PM, Leif M wrote:
I just noticed that I can't get return loss better than about 40dB even with a40 dB return loss is an SWR of 1.02. If you calibrated with a load that was exactly 50 ohms and the DUT was purely resistive it would be 50.1 ohms at this SWR. However the loads supplied with the NanoVNA's are not near 50ohms. The one supplied with my NanoVNA-H is 49.81 at DC and the nanoVNA-H4 (also supplied with SAA-2) is 50.86 DC. The impedance varies with frequency. So if I calibrate with one of those and then measure a 30 dB attenuator 40 dB of Return Loss would not be much of a surprise. I get the wavy Return Loss on my RigExpert and NanoVNA's when using typical "nominal impedance" 50 ohm cable. These ripples in the SWR will have a minimum at multiples of 1/4 wavelength and maximum at multiples of 1/2 wavelength. Here is an example with 50 feet of RG-8X with a 75 ohm resistive load at the end. Roger |
|
Re: How low should a return loss at CH0 be?
Mel Farrer
Consider the magnitude of that measurement in terms of %.? For practical uses, anything over -40 dB is meaningless unless you are comparing a part specification.? Like it is supposed to be >-45 dB.? ?Yes, my cal gives me >-60 dB, but I don't ever need it.??Please don't get blinded by specifications.
toggle quoted message
Show quoted text
Mel, K6KBE On Thursday, January 28, 2021, 01:53:38 PM PST, Dragan Milivojevic <d.milivojevic@...> wrote:
In theory down to 90dB. If I calibrate and measure with my DIY standard I get those figures in the HF range. But that's mostly meaningless. You have to remember that these are relative measurements, you can't measure over X dB RL if your calibration standard RL is less than X. On Thu, 28 Jan 2021 at 22:28, Leif M <leif.michaelsson@...> wrote: I just noticed that I can't get return loss better than about 40dB even |
|
Re: How low should a return loss at CH0 be?
In theory down to 90dB. If I calibrate and measure with my DIY standard
toggle quoted message
Show quoted text
I get those figures in the HF range. But that's mostly meaningless. You have to remember that these are relative measurements, you can't measure over X dB RL if your calibration standard RL is less than X. On Thu, 28 Jan 2021 at 22:28, Leif M <leif.michaelsson@...> wrote:
I just noticed that I can't get return loss better than about 40dB even |
Roger Need
