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
|
Locked
Re: Correct button?
Hi Alan,
toggle quoted message
Show quoted text
I just want to point out that the polynomial + delay + impedance method for characterising standards is itself just an approximation. In this day and age (of cheap storage) it is simpler to just sweep the standards on a calibrated VNA, get an s-param file for each one and use that data in the calcs - with some interpolation for the in-between frequencies. This gives the highest accuracy as there is no modelling error. It also means you can use any old standard you like. The polynomial models don't fit poorly made standards very well over a wide frequency range, hence the need for high quality expensive parts. However I'm not sure if this device has the storage available to do it this way, if not the old school approach might be the best way. Regards, Roger On Tue, 30 Jul 2019 at 08:49, alan victor <avictor73@...> wrote:
Thanks! That is helpful. The fact that the polynomial code to describe the |
|
Re: Possible location of latest VNA source code??
Lapo Pieri
Hi,
11:32 Mon 29 Jul 19 , DMR wrote: After compilation it works exactly the same as yours.Where are these vars? Hi hugen! Share the full working code please.I take the full code from here: Are we talking about different source code? Lapo, IK5NAX |
|
Re: nano cases...
I agree... to a point. This thing isn't a "real" VNA (in the commercial respect). I don't know if I would have used it when I schlepped up a water tower in Melbourne Australia. I had a Keysight N9914 :-)
toggle quoted message
Show quoted text
On 7/29/2019 10:19 AM, Burt K6OQK wrote:
Personally, I'd rather see this instrument in a tablet size, something I can easily see; in a real metal box with Type-N connectors and sturdy switches and push-buttons.?? My need is for an instrument that I can carry to antenna sites for broadcast antenna work.?? I understand where some folks like tiny pocket size instruments, not sure why, but I'm probably going to go to something like an HP e5061B - a totally different area of an instrument.?? I like the 8753c or ES but they are pretty big and heavy for schleping to transmitter sites. |
|
Re: Smith chart impedance printout
Lapo Pieri
Hi,
19:27 Mon 29 Jul 19 , alan victor wrote: Sure Lapo. That is great. However they are all practical representations. In fact why stop there. You could convert the result to its parallel equivalent admittance and provide that as well.I think I have to study the code a bit more to make more complex mods, stay tuned ;-) Lapo, IK5NAX |
|
Re: Smith chart impedance printout
Lapo Pieri
16:07 Mon 29 Jul 19 , alan victor wrote:
Wait a second. The display is available as R+jX. Just switch to And if you desire you can dedicate trace 0 to a Smith plot while youNo, polar format display coefficient of reflection, not impedance! Lapo, IK5NAX |
|
Re: Smith chart impedance printout
Hi Lapo,
Wait a second. The display is available as R+jX. Just switch to polar.format. And if you desire you can dedicate trace 0 to a Smith plot while you dedicate trace 1 to a polar plot. Of course in polar format you will have to multiply the polar vector by Zo of the chart/system which of course is 50 ohms. Alan |
|
Locked
Re: Correct button?
Thanks! That is helpful. The fact that the polynomial code to describe the open is present implies he was aware of the situation and at this point elected to put it aside. So for all intent and purpose, the loads are assumed to be ideal and that tends to agree with the final sweep to check the loads after cal. The check returns them to be IDEAL... And they are anything but ideal as that is the electrical-physical nature of a SMA connector.
Alan ________________________________ From: [email protected] <[email protected]> on behalf of Roger Henderson <hendorog@...> Sent: Monday, July 29, 2019 8:32 PM To: [email protected] <[email protected]> Subject: Re: [nanovna-users] Correct button? I checked the code repo's posted and this is what I think the situation is with the standards: The load is assumed to be perfect - i.e. the assumption is that nothing is reflected from the load. The open is assumed to have a single C0 capacitance term of 50e-15F The short is either assumed to be perfect, or is assumed to be 180degrees away from the open. I am not quite sure on that. There is some unused code which looks to be intended to create a C0,C1,C2,C3 model of the open, but it is not used. Roger On Mon, 29 Jul 2019 at 14:37, alan victor <avictor73@...> wrote: You are correct and I confirmed on the big box, that is the function |
|
Locked
Re: Correct button?
I checked the code repo's posted and this is what I think the situation is
toggle quoted message
Show quoted text
with the standards: The load is assumed to be perfect - i.e. the assumption is that nothing is reflected from the load. The open is assumed to have a single C0 capacitance term of 50e-15F The short is either assumed to be perfect, or is assumed to be 180degrees away from the open. I am not quite sure on that. There is some unused code which looks to be intended to create a C0,C1,C2,C3 model of the open, but it is not used. Roger On Mon, 29 Jul 2019 at 14:37, alan victor <avictor73@...> wrote:
You are correct and I confirmed on the big box, that is the function |
|
Re: Smith chart impedance printout
Sure Lapo. That is great. However they are all practical representations. In fact why stop there. You could convert the result to its parallel equivalent admittance and provide that as well.
I suspect that if you are doing antenna work and wish to design an impedance matching network, YES. The usual process is to specify the LOAD, the antenna as a complex Z or Y and not as a R and C/L unit. However, a little calculator can handle the conversion. But you are right, it would just be more convenient to have the vna spit it out. I think best to provide this as a user selection menu item. What do you think? 73' Alan ________________________________ From: [email protected] <[email protected]> on behalf of Lapo Pieri <ik5nax@...> Sent: Monday, July 29, 2019 5:51 PM To: nanovna-users <[email protected]> Subject: [nanovna-users] Smith chart impedance printout Hi, I think that Smith chart presentation of impedance as <real_part>[ohm] <series cap or ind>[F|H] is not practical in many case of use. So I decide to modify the code to show impedance in <real_part>+/-j<imag_part> format, you can find the patch to plot.c to obtain this. If I found a bit of time I could split Smith chart in two, one as original and one as just proposed. Do you think it worth? Lapo, IK5NAX -------- diff --git a/plot.c b/plot.c index 10074c3..c5c92e5 100644 --- a/plot.c +++ b/plot.c @@ -576,16 +576,22 @@ gamma2imp(char *buf, int len, const float coeff[2], uint32_t frequency) // float z = sqrtf(zr*zr + zi*zi); int n; - n = string_value_with_prefix(buf, len, zr, S_OHM[0]); - buf[n++] = ' '; + n = string_value_with_prefix(buf, len, zr, '\0'); + if(zi<0) + buf[n++]='-'; + else + buf[n++]='+'; + buf[n++]='j'; + string_value_with_prefix(buf+n, len-n, fabs(zi), S_OHM[0]); - if (zi < 0) { - float c = -1 / (PI2 * frequency * zi); - string_value_with_prefix(buf+n, len-n, c, 'F'); - } else { - float l = zi / (PI2 * frequency); - string_value_with_prefix(buf+n, len-n, l, 'H'); - } + + /* if (zi < 0) { */ + /* float c = -1 / (PI2 * frequency * zi); */ + /* string_value_with_prefix(buf+n, len-n, c, 'F'); */ + /* } else { */ + /* float l = zi / (PI2 * frequency); */ + /* string_value_with_prefix(buf+n, len-n, l, 'H'); */ + /* } */ } void |
|
Re: Possible location of latest VNA source code??
Hugen has already committed his changes back to edy555's original github repository.
toggle quoted message
Show quoted text
Have a look at On Mon, Jul 29, 2019 at 02:32 PM, DMR wrote:
|
|
Re: Possible location of latest VNA source code??
DMR
Lapo Pieri,
After compilation it works exactly the same as yours. There is still a problem, an error occurs when compiling if USE_PROCESS_STACKSIZE = 0x200 USE_EXCEPTIONS_STACKSIZE = 0x200 With these values, everything compiles without errors. USE_PROCESS_STACKSIZE = 0x100 USE_EXCEPTIONS_STACKSIZE = 0x100 Hi hugen! Share the full working code please. |
|
Smith chart impedance printout
Lapo Pieri
Hi,
I think that Smith chart presentation of impedance as <real_part>[ohm] <series cap or ind>[F|H] is not practical in many case of use. So I decide to modify the code to show impedance in <real_part>+/-j<imag_part> format, you can find the patch to plot.c to obtain this. If I found a bit of time I could split Smith chart in two, one as original and one as just proposed. Do you think it worth? Lapo, IK5NAX -------- diff --git a/plot.c b/plot.c index 10074c3..c5c92e5 100644 --- a/plot.c +++ b/plot.c @@ -576,16 +576,22 @@ gamma2imp(char *buf, int len, const float coeff[2], uint32_t frequency) // float z = sqrtf(zr*zr + zi*zi); int n; - n = string_value_with_prefix(buf, len, zr, S_OHM[0]); - buf[n++] = ' '; + n = string_value_with_prefix(buf, len, zr, '\0'); + if(zi<0) + buf[n++]='-'; + else + buf[n++]='+'; + buf[n++]='j'; + string_value_with_prefix(buf+n, len-n, fabs(zi), S_OHM[0]); - if (zi < 0) { - float c = -1 / (PI2 * frequency * zi); - string_value_with_prefix(buf+n, len-n, c, 'F'); - } else { - float l = zi / (PI2 * frequency); - string_value_with_prefix(buf+n, len-n, l, 'H'); - } + + /* if (zi < 0) { */ + /* float c = -1 / (PI2 * frequency * zi); */ + /* string_value_with_prefix(buf+n, len-n, c, 'F'); */ + /* } else { */ + /* float l = zi / (PI2 * frequency); */ + /* string_value_with_prefix(buf+n, len-n, l, 'H'); */ + /* } */ } void |
|
Re: Possible location of latest VNA source code??
Lapo Pieri
Hi,
02:32 Mon 29 Jul 19 , hugen@... wrote: You may have pressed the touch screen while adc is initializing.No, I don't I'm sure It might be better to add a wait.I agree. I don't know STM32, I usually develop on NXP LPCxxxx but it seems to me good. void adc_init(void)Ok, now it works but after a full power-down. I mean: after re-build and fw load ($ dfu-util -d 0483:df11 -a 0 -s 0x08000000:leave -D build_900/ch.bin) on subsequent reset it hangs again just after start and stop frequency printout; switching off and then on rapidly (blue led remains on) it hangs again. Switching off, waiting for blue led to go down and switching on again it works. Why? Many thanks! Lapo, IK5NAX |
|
Idea for a bracket for mounting N connectors
Folks,
A number of you were commenting on maybe using female N connectors on this unit but were concerned about breaking the SMA connectors off the circuit board. I did a little 'doodle' at lunchtime about how I thought you could fabricate a steel bracket to hold two N-SMA adapters to the NanoVNA. If you are no sure of mounting at the one end of the device, you could always use a thicker plate steel and use it to replace the entire rear panel of the VNA. That would provide more rigidity as would sides to the bracket - In fact, a solid back cover could be produced that had sides to cover the edges of the VNA. Anyone who has worked with sheet metal can make this very easily, although I would not use aluminum/brass/copper - too flexible. Probably 18-20 gauge would be ideal. I have another drawing to add that shows a rear cover plate - probably Tuesday (gotta get back to work) |
|
Re: Best unit to purchase?
jim
Yep .. still had no (easier) way of compensating hugen so...
toggle quoted message
Show quoted text
and that was my original intention (stare and compare) Jim On Monday, July 29, 2019, 4:41:28 PM UTC, Hans J Albertsson <hans.j.albertsson@...> wrote:
So total was $71.90? Please, Jim, once you get the Hugen NanoVNA, compare the two units' hardware and describe the differences, as far as you can! TiA. Den m?n 29 juli 2019 17:58jim via Groups.Io <ab7vf@...> skrev: ? Same vendor I used ...Ordered july 3, recieved july 26 ... |
|
Re: Best unit to purchase?
So total was $71.90?
toggle quoted message
Show quoted text
Please, Jim, once you get the Hugen NanoVNA, compare the two units' hardware and describe the differences, as far as you can! TiA. Den m?n 29 juli 2019 17:58jim via Groups.Io <ab7vf@...> skrev: Same vendor I used ...Ordered july 3, recieved july 26 ... |
|
edy555 design notes on CALIBRATION sets
There is a wealth of good information in these notes. Thanks Larry for the translation. It is clear the developer put a great deal of effort into the project.
Reading the description of the work in the area of calibration, I see that significant effort was addressed in the algorithm addressing the coupler performance. This is clear as the majority of the CAL routine steps address the coupler tracking, directivity and frequency response. Reading carefully "Trying out the calibration process of nanoVNA with Python" you will see that the designer recognizes that the CAL kit pieces are not perfect. However, he state they are adequate to provide a reasonable agreement with measured frequency response vs. actual. The cost performance tradeoff is reasonable as the first order of business was to handle the coupler degradation to measured vs. actual response data. So at this point, create the best short, open and 50 ohm load... the supplied are not too bad. Perhaps the open as suggested in a prior post can be improved. However, KEY is the VNA is under the impression that these cal standards are IDEAL. What does this mean. Without going through extensive arithmetic and explanation, consider the fact that the measured short when CAL is complete is precisely located at the short side of the chart as a single dot. This is not the case for a true SMA short. There is clockwise (CW) rotation as frequency increases. Hence there is series inductance present with a real SMA CAL standard. As well, that rotation is removed from your DUT. So expect for example, for this case, that the actual value of you L is a bit larger than measured. At low frequnecies the difference is negligible. So if you desire to use the nanoVNA as a component measurement tool, like an LCR meter, conduct your measurments at say 1 MHz or even 50 kHz! More on this topic I am sure will be forthcoming. Alan |
|
Re: Best unit to purchase?
jim
Same vendor I used ...Ordered july 3, recieved july 26 ...
toggle quoted message
Show quoted text
In order to (maybe) compensate Hugen for his efforts, I also ordered one via "taobao focus" ( "hugen79") .....cost of unit $53.54 shipping/handling by shipper $18.36 ...shipped july 28 ..not here yet Jim On Monday, July 29, 2019, 3:12:00 PM UTC, Larry Rothman <ac293@...> wrote:
Ed, I got mine from Aliexpress vendor:? It's a very nice unit, comes in a plastice case with all the attachments/cables/cal and has a battery. It took 21 days to Toronto, Canada. Check my previous posts. Cheers, Larry On Mon, Jul 29, 2019 at 11:06 AM, kg7swp@... wrote:
|
|
Re: Best unit to purchase?
Here is an SWR meter (to 60MHz) with Bluetooth Mini60S
toggle quoted message
Show quoted text
On 7/29/2019 11:11 AM, Larry Rothman wrote:
Ed, |
to navigate to use esc to dismiss