Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
- Nanovna-Users
- Messages
Search
Re: Measurement of high impedance loads
"Phase angles and imaginary parts were not in the scope of what I was checking."
Measuring resistors with a resistance bridge usually produces good results. But measuring inductors and capacitors with a resistance bridge does not work as well because measuring angles near plus/minus 90 degrees is very difficult for a resistance bridge. And these errors accumulate faster than for pure resistors measured at zero degrees. |
Re: Measurement of nonlinear (active) devices
I agree that for active devices the NanoVNA would not be acceptable. I usually start measurements at -30 dBm with my scalar network analyzers when measuring amplifiers, to avoid overload. By carefully watching the DC current into the amplifier and slowly raising? the input RF level, you can tell when it is into saturation.
toggle quoted message
Show quoted text
Trying to turn the NanoVNA, which as far as I can tell was born as an antenna analyzer for hams, into a multi-thousand dollar VNA is fraught with peril. I am sure that there is room for improvement and growth for this very nice, inexpensive piece of test equipment, but don't expect too much.? Stuart K6YAZLos Angeles, USA -----Original Message-----
From: alan victor <avictor73@...> To: [email protected] <[email protected]> Sent: Mon, Aug 12, 2019 10:55 am Subject: Re: [nanovna-users] Measurement of nonlinear (active) devices The fundamental issue in using the NanoVNA for measuring active devices is FIRST its output power level. It is specified at -9 dBm to -12 dBm. If you consider a bipolar device operating at a quiescent current of 1 mA, these power levels are too large! The small signal S parameters are just that, small signal. A bipolar transistor had a Vbe value of 26 mV/mA of emitter current. Hence, once you subject the DUT to such a Pin level, you will notice the quiescent bias level to increase! That is NOT a small signal measurement. As a rule of thumb, I would start at -30 or -40 dBm and get a S21 dB measure. So you will need a 20 or 30 dB pad. You may also pad the output of the VNA, CH0,? to achieve an appropriate level. I would also pad the CH1 of the VNA to be sure that the VNA is not over driven as the gain of the device, S21,? at audio and HF may be quite large. Also, be sure your DUT is stable! I have seen 100k$+ VNA's go up in smoke as a measurement of a power device was unstable. The device will oscillate and the VNA is destroyed. You will have to construct bias tees. More on that latter. Alas... its only a 50 buck VNA... go for it. But your warned. 73' Alan |
Re: Measurement of nonlinear (active) devices
The fundamental issue in using the NanoVNA for measuring active devices is FIRST its output power level. It is specified at -9 dBm to -12 dBm. If you consider a bipolar device operating at a quiescent current of 1 mA, these power levels are too large! The small signal S parameters are just that, small signal. A bipolar transistor had a Vbe value of 26 mV/mA of emitter current. Hence, once you subject the DUT to such a Pin level, you will notice the quiescent bias level to increase! That is NOT a small signal measurement. As a rule of thumb, I would start at -30 or -40 dBm and get a S21 dB measure. So you will need a 20 or 30 dB pad.
You may also pad the output of the VNA, CH0, to achieve an appropriate level. I would also pad the CH1 of the VNA to be sure that the VNA is not over driven as the gain of the device, S21, at audio and HF may be quite large. Also, be sure your DUT is stable! I have seen 100k$+ VNA's go up in smoke as a measurement of a power device was unstable. The device will oscillate and the VNA is destroyed. You will have to construct bias tees. More on that latter. Alas... its only a 50 buck VNA... go for it. But your warned. 73' Alan |
Re: Measurement of nonlinear (active) devices
Lapo Pieri
05:50 Mon 12 Aug 19 , RFpro wrote:
Has anyone measured the accuracy of nonlinear, active devices like semiconductors?Not yet In general, a power level of -30dBm and higher will have an effect on the measured accuracy.It depend on devices and circuits! :-) How will the harmonics of the test signal have an effect?Sure What you want to do, precisely? Lapo |
Re: Measurement of high impedance loads
I don't believe the SA612 detectors are good to 0.1dB magnitudes.SA612 are being used as mixers; do you suggest using AD8307 as mixer, perhaps connecting ENB input to local oscillator? AD8307 Upper Usable Frequency is 500MHz; ENB delay spec is 400-500 ns; |
Re: Measurement of high impedance loads
I think we are all in agreement then.Sent from my Samsung Galaxy smartphone.
toggle quoted message
Show quoted text
-------- Original message --------From: alan victor <avictor73@...> Date: 12/08/2019 17:09 (GMT+01:00) To: [email protected] Subject: Re: [nanovna-users] Measurement of high impedance loads VNA instruments maintian best accuracy when the real Z is near Zo or 50 ohms. There is a sweet? spot between 20 and 80 ohms measured in the shunt mode where all is nice. Above and below that value measurment techniques require both S11 and S21 to be recorded. The fact that decent values are achieved at higher real parts is rewarding. Further improvement is hence possible if you measured in a thru mode so called series mode or in the shunt - thru mode.See "Make accurate impedance measurements using a VNA"Alan________________________________From: [email protected] <[email protected]> on behalf of Warren Allgyer <allgyer@...>Sent: Monday, August 12, 2019 2:52 PMTo: [email protected] <[email protected]>Subject: Re: [nanovna-users] Measurement of high impedance loadsPhase angles and imaginary parts were not in the scope of what I was checking. That is why I made the measurement at 500 kHz. And I do not own a real high end VNA, HP or otherwise. So I cannot help you there.The point is, I am very sure NanoVNA will not track a $75,000 VNA with 0.1 dB precision. There are many, many users, perhaps most users, however who will derive great utility from this $55 device even if its accuracy is only 1 dB or even 5 dB. I also would expect significant variance from unit to unit so that even from unit to unit the accuracy would not track at that level.I would not discourage your intellectual exercise but would only point out that for most here it is irrelevant.WA8TOD> On Aug 12, 2019, at 10:30 AM, tuckvk3cca <tuckvk3cca@...> wrote:>> Have you got the phase angles or imaginary parts? Can you compare the readings with your HP analyzer? Comparing with the VOM values can be misleading. Reason is I don't? believe the SA612 detectors are good to 0.1dB magnitudes. The AD8307 is good to 0.3dB at best and its a much better chip. Secondly as the Return loss goes down the uncertainty in resistance increases e.g at 0dB the resistance can be either zero or infinity. To narrow this gap requires very accurate phase angle measurements. Again the smallest phase error can lead to large uncertainties.?? I don't? think its conservative to not accept any impedance values from the nano when return loss is less than 1dB.Sent from my Samsung Galaxy smartphone.> -------- Original message --------From: Warren Allgyer <allgyer@...> Date: 12/08/2019? 14:56? (GMT+01:00) To: [email protected] Subject: Re: [nanovna-users] Measurement of high impedance loads I am not sure I understand the question.The NanoVNA can certainly indicate down to 0.1 dB, so it can measure there. The accuracy of that measurement would be subject to the accuracy of the calibration and the inherent accuracy of the instrument. Personally and for my purposes, absolute measurements of RL to the 0.1 level are not very interesting so long as the instrument is capable of directional changes at that incremental level. The NanoVNA certainly indicates in 0.1 dB increments and that is more than enough for my purposes.WA8TODOn Aug 12, 2019, at 8:37 AM, tuckvk3cca <tuckvk3cca@...> wrote:I have added the vswr and the return loss. Can the nano measure down to 0.1dB return loss?Sent from my Samsung Galaxy smartphone.-------- Original message --------From: Warren Allgyer <allgyer@...> Date: 12/08/2019? 13:55? (GMT+01:00) To: [email protected] Subject: [nanovna-users] Measurement of high impedance loads Earlier there was a comment in one of the threads asking if anyone had measured the accuracy at high values of resistance. Here are four data points I measured this morning. The measurements were done at 500 kHz in order to minimize the reactance of the test fixture which was a pair of binding posts on a BNC adapter. The first value is the actual resistor value as measured on a VOM and the second is the value indicated by the Smith Chart on NanoVNA.???????????????????????????????? 810 ohms? -? 810 -16.2-1.08dB1500 ohms? -? 1500-30-0.58dB2980 ohms? -? 2880-57.6-0.3dB8100 ohms? -? 7940-159-0.11dBWA8TOD>>
|
Re: Measurement of high impedance loads
VNA instruments maintian best accuracy when the real Z is near Zo or 50 ohms. There is a sweet spot between 20 and 80 ohms measured in the shunt mode where all is nice. Above and below that value measurment techniques require both S11 and S21 to be recorded. The fact that decent values are achieved at higher real parts is rewarding. Further improvement is hence possible if you measured in a thru mode so called series mode or in the shunt - thru mode.
See "Make accurate impedance measurements using a VNA" Alan ________________________________ From: [email protected] <[email protected]> on behalf of Warren Allgyer <allgyer@...> Sent: Monday, August 12, 2019 2:52 PM To: [email protected] <[email protected]> Subject: Re: [nanovna-users] Measurement of high impedance loads Phase angles and imaginary parts were not in the scope of what I was checking. That is why I made the measurement at 500 kHz. And I do not own a real high end VNA, HP or otherwise. So I cannot help you there. The point is, I am very sure NanoVNA will not track a $75,000 VNA with 0.1 dB precision. There are many, many users, perhaps most users, however who will derive great utility from this $55 device even if its accuracy is only 1 dB or even 5 dB. I also would expect significant variance from unit to unit so that even from unit to unit the accuracy would not track at that level. I would not discourage your intellectual exercise but would only point out that for most here it is irrelevant. WA8TOD On Aug 12, 2019, at 10:30 AM, tuckvk3cca <tuckvk3cca@...> wrote: |
Re: Measurement of high impedance loads
Phase angles and imaginary parts were not in the scope of what I was checking. That is why I made the measurement at 500 kHz. And I do not own a real high end VNA, HP or otherwise. So I cannot help you there.
toggle quoted message
Show quoted text
The point is, I am very sure NanoVNA will not track a $75,000 VNA with 0.1 dB precision. There are many, many users, perhaps most users, however who will derive great utility from this $55 device even if its accuracy is only 1 dB or even 5 dB. I also would expect significant variance from unit to unit so that even from unit to unit the accuracy would not track at that level. I would not discourage your intellectual exercise but would only point out that for most here it is irrelevant. WA8TOD On Aug 12, 2019, at 10:30 AM, tuckvk3cca <tuckvk3cca@...> wrote: |
Re: Measurement of high impedance loads
Have you got the phase angles or imaginary parts? Can you compare the readings with your HP analyzer? Comparing with the VOM values can be misleading.?Reason is I don't? believe the SA612 detectors are good to 0.1dB magnitudes. The AD8307 is good to 0.3dB at best and its a much better chip. Secondly as the Return loss goes down the uncertainty in resistance increases e.g at 0dB the resistance can be either zero or infinity. To narrow this gap requires very accurate phase angle measurements. Again the smallest phase error can lead to large uncertainties.? ?I don't? think its conservative to not accept any impedance values from the nano when return loss is less than 1dB.Sent from my Samsung Galaxy smartphone.
toggle quoted message
Show quoted text
-------- Original message --------From: Warren Allgyer <allgyer@...> Date: 12/08/2019 14:56 (GMT+01:00) To: [email protected] Subject: Re: [nanovna-users] Measurement of high impedance loads I am not sure I understand the question.The NanoVNA can certainly indicate down to 0.1 dB, so it can measure there. The accuracy of that measurement would be subject to the accuracy of the calibration and the inherent accuracy of the instrument. Personally and for my purposes, absolute measurements of RL to the 0.1 level are not very interesting so long as the instrument is capable of directional changes at that incremental level. The NanoVNA certainly indicates in 0.1 dB increments and that is more than enough for my purposes.WA8TODOn Aug 12, 2019, at 8:37 AM, tuckvk3cca <tuckvk3cca@...> wrote:I have added the vswr and the return loss. Can the nano measure down to 0.1dB return loss?Sent from my Samsung Galaxy smartphone.-------- Original message --------From: Warren Allgyer <allgyer@...> Date: 12/08/2019? 13:55? (GMT+01:00) To: [email protected] Subject: [nanovna-users] Measurement of high impedance loads Earlier there was a comment in one of the threads asking if anyone had measured the accuracy at high values of resistance. Here are four data points I measured this morning. The measurements were done at 500 kHz in order to minimize the reactance of the test fixture which was a pair of binding posts on a BNC adapter. The first value is the actual resistor value as measured on a VOM and the second is the value indicated by the Smith Chart on NanoVNA.???????????????????????????????? 810 ohms? -? 810 -16.2-1.08dB1500 ohms? -? 1500-30-0.58dB2980 ohms? -? 2880-57.6-0.3dB8100 ohms? -? 7940-159-0.11dBWA8TOD
|
Re: Measurement of high impedance loads
I am not sure I understand the question.
toggle quoted message
Show quoted text
The NanoVNA can certainly indicate down to 0.1 dB, so it can measure there. The accuracy of that measurement would be subject to the accuracy of the calibration and the inherent accuracy of the instrument. Personally and for my purposes, absolute measurements of RL to the 0.1 level are not very interesting so long as the instrument is capable of directional changes at that incremental level. The NanoVNA certainly indicates in 0.1 dB increments and that is more than enough for my purposes. WA8TOD On Aug 12, 2019, at 8:37 AM, tuckvk3cca <tuckvk3cca@...> wrote:
I have added the vswr and the return loss. Can the nano measure down to 0.1dB return loss?Sent from my Samsung Galaxy smartphone. -------- Original message --------From: Warren Allgyer <allgyer@...> Date: 12/08/2019 13:55 (GMT+01:00) To: [email protected] Subject: [nanovna-users] Measurement of high impedance loads Earlier there was a comment in one of the threads asking if anyone had measured the accuracy at high values of resistance. Here are four data points I measured this morning. The measurements were done at 500 kHz in order to minimize the reactance of the test fixture which was a pair of binding posts on a BNC adapter. The first value is the actual resistor value as measured on a VOM and the second is the value indicated by the Smith Chart on NanoVNA. 810 ohms - 810 -16.2-1.08dB1500 ohms - 1500-30-0.58dB2980 ohms - 2880-57.6-0.3dB8100 ohms - 7940-159-0.11dBWA8TOD |
Re: Measurement of high impedance loads
I have added the vswr and the return loss. Can the nano measure down to 0.1dB return loss?Sent from my Samsung Galaxy smartphone.
toggle quoted message
Show quoted text
-------- Original message --------From: Warren Allgyer <allgyer@...> Date: 12/08/2019 13:55 (GMT+01:00) To: [email protected] Subject: [nanovna-users] Measurement of high impedance loads Earlier there was a comment in one of the threads asking if anyone had measured the accuracy at high values of resistance. Here are four data points I measured this morning. The measurements were done at 500 kHz in order to minimize the reactance of the test fixture which was a pair of binding posts on a BNC adapter. The first value is the actual resistor value as measured on a VOM and the second is the value indicated by the Smith Chart on NanoVNA.? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?810 ohms? -? 810 -16.2-1.08dB1500 ohms? -? 1500-30-0.58dB2980 ohms? -? 2880-57.6-0.3dB8100 ohms? -? 7940-159-0.11dBWA8TOD
|
Measurement of high impedance loads
Earlier there was a comment in one of the threads asking if anyone had measured the accuracy at high values of resistance. Here are four data points I measured this morning. The measurements were done at 500 kHz in order to minimize the reactance of the test fixture which was a pair of binding posts on a BNC adapter. The first value is the actual resistor value as measured on a VOM and the second is the value indicated by the Smith Chart on NanoVNA.
810 ohms - 810 1500 ohms - 1500 2980 ohms - 2880 8100 ohms - 7940 WA8TOD |
Re: Can the factory set C0 values be stored?
Hello Bob,
I seem to recall reading on one of the eBay vendor's auction pages for a NanoVNA, some wording that claimed the following: "The buyer should not worry that the picture shows a NanoVNA with a range of 50 kHz to 300 MHz. The unit that will be shipped will have a firmware upgrade to allow operation from 50 kHz. to 900 MHz. and it will be calibrated to factory standards using the same included calibration devices." Of course, when I went back and tried to find that listing - I couldn't. In my opinion, I don't think that any seller would be using "better" standards to calibrate the unit than the standards they were shipping with the unit. None of these manufacturers or sellers are trying to compete with HP/Agilent/Keysight or any of the other VNA manufactures out there. They all just want to build and ship a unit that meets the advertised specifications and do that as inexpensively as possible to maximize their profit. I see nothing in it for the seller. Certainly the end user can take the supplied C0 data and save it in one the other Memory locations (SAVE1 thru SAVE4) and then load C0 with some other data so that it is always readily available at power-up. This would allow the initial data to be reloaded into C0 at some other time. But I have not seen or read any reason to believe that the factory furnished C0 calibration data is any better than the data that can be recreated using the furnished calibration tools. I offer all of this IMHO. Larry, AE5CZ |
Re: Accuracy of the nanoVNA units
There are good designs for 9:1 or higher ratio baluns out there using the right ferrite cores e.g. types 77 or 2401. My previous remarks about the 800 ohms max could be extended using such a balun. Any comments?
toggle quoted message
Show quoted text
-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Bob Shaw Sent: Sunday, August 11, 2019 1:31 AM To: [email protected] Subject: [nanovna-users] Accuracy of the nanoVNA units Has anyone measured some high resistive and inductive/capacitative loads and compared the resultant reported impedance with the results of the same loads on a known more accurate VNA? This would be interesting and relatively easily done if you have access to a professional grade VNA (which I don't). |
Re: New file uploaded to [email protected]
No the program compiles and run alright. It cannot connect. Downloaded v1.4 from STMmicro website. It has x64 or x86 drivers. Installed x86 win 7 version? and now working!? Take note guys.Sent from my Samsung Galaxy smartphone.
toggle quoted message
Show quoted text
-------- Original message --------From: alan victor <avictor73@...> Date: 11/08/2019 12:19 (GMT+01:00) To: [email protected] Subject: Re: [nanovna-users] New file uploaded to [email protected] I suspect the issue is the one dll (dynamic library link) present. The windows compiler does give you the option at compilation to select 32/64 or BOTH. Get the original code and re compile for 32.________________________________From: [email protected] <[email protected]> on behalf of tuckvk3cca <tuckvk3cca@...>Sent: Sunday, August 11, 2019 7:20 AMTo: [email protected] <[email protected]>Subject: Re: [nanovna-users] New file uploaded to nanovna-users@...? Alan. I have home brewed several vnas by now, everyone is a lesson. And we get a better instrument each time. On another note, I couldn't? get my nano to run on an old pc 32 bits with Win 7. The supplied driver for it is 64bit for some reason and failed. Where can I find a 32 bit driver?Sent from my Samsung Galaxy smartphone.-------- Original message --------From: alan victor <avictor73@...> Date: 11/08/2019? 01:34? (GMT+01:00) To: [email protected] Subject: Re: [nanovna-users] New file uploaded to [email protected] Hey guys, take it easy.... I was simply putting forth a very basic question. No more complicated than asking a ham... amateur radio op.... What's the sensitivity of that voltmeter your using to measure that voltage. Gee... it's just a fundamental question. Nothing more, nothing less. I said nothing about comparing it to anything.
|
Re: New file uploaded to [email protected]
I suspect the issue is the one dll (dynamic library link) present. The windows compiler does give you the option at compilation to select 32/64 or BOTH. Get the original code and re compile for 32.
________________________________ From: [email protected] <[email protected]> on behalf of tuckvk3cca <tuckvk3cca@...> Sent: Sunday, August 11, 2019 7:20 AM To: [email protected] <[email protected]> Subject: Re: [nanovna-users] New file uploaded to [email protected] Correct Alan. I have home brewed several vnas by now, everyone is a lesson. And we get a better instrument each time. On another note, I couldn't get my nano to run on an old pc 32 bits with Win 7. The supplied driver for it is 64bit for some reason and failed. Where can I find a 32 bit driver?Sent from my Samsung Galaxy smartphone. -------- Original message --------From: alan victor <avictor73@...> Date: 11/08/2019 01:34 (GMT+01:00) To: [email protected] Subject: Re: [nanovna-users] New file uploaded to [email protected] Hey guys, take it easy.... I was simply putting forth a very basic question. No more complicated than asking a ham... amateur radio op.... What's the sensitivity of that voltmeter your using to measure that voltage. Gee... it's just a fundamental question. Nothing more, nothing less. I said nothing about comparing it to anything. |
Re: New file uploaded to [email protected]
Correct? Alan. I have home brewed several vnas by now, everyone is a lesson. And we get a better instrument each time.?On another note, I couldn't? get my nano to run on an old pc 32 bits with Win 7. The supplied driver for it is 64bit for some reason and failed. Where can I find a 32 bit driver?Sent from my Samsung Galaxy smartphone.
toggle quoted message
Show quoted text
-------- Original message --------From: alan victor <avictor73@...> Date: 11/08/2019 01:34 (GMT+01:00) To: [email protected] Subject: Re: [nanovna-users] New file uploaded to [email protected] Hey guys, take it easy.... I was simply putting forth a very basic question. No more complicated than asking a ham... amateur radio op.... What's the sensitivity of that voltmeter your using to measure that voltage. Gee... it's just a fundamental question. Nothing more, nothing less. I said nothing about comparing it to anything.
|
to navigate to use esc to dismiss