开云体育

Rudi,

Glad to hear you got the results you expected

A couple of questions...

1. Your concern was that S11 was different when you hooked up the NanoVNA instead of the 50 ohm load. Could you please post your graph comparing S11 with the trace cut on the NooElec.

2. Your V1 NooElec balun is different from mine and the pictures submitted by others on this thread. See the pictures below. Does yours have a track on the back to cut like my previous picture?

3. Did you try my suggestion of measuring S11 with the coax and 50 ohm resistor attached to the SMA connector? I wonder how it compares to S11 when the coax is attached to the VNA which is not 50 ohms impedance at all frequencies.

4. You show S21 being compared with 2 different methods. I understand how you did it with the VNA and your 400/50 method but how did you get the other result? With an oscilloscope and manual calculation or a spectrum analyzer?

Roger

I find using a soft stylus made a capacitance touch display or the eraser end of a mechanical pencil works better than my finger. See attachment.

Don't use the hard pointed end of the stylus which is made of a resistive display

I would also recommend the modification of placing thin plastic washers between the display and the top of the case. I cut my washers out of a piece of plastic.

Mike N2MS

That is typical of screen cal.
Use the jog sw to select the screen cal function, cal and test but remember to save the cal through the menu.

I tried to cal the touch screen. The menu bar is erratic. Touch it and it
usually disappears. I wonder if this is a screen cal issue or if it is
something else (softwre?).

On Wed, May 27, 2020 at 8:37 PM Simon Eatough <simon.eatough@...>
wrote:

Have similar issues with my new H4. The touch screen interface is
inconsistent. Sometimes pressing a menu action works with the correct
option highlighted. Other times not either with the wrong button selected
or nothing happens at all. I have calibrated the LCD.
Simon
ZL2BRG

--
Paul

W8SBH
?? ?? ????
(this too shall pass)

Thanks for the compliment and the advice. Attenuators are a design feature, component values are on he schematic for 3, 6, and 15 dB. I discovered a similar comment while researching various ideas in an attempt to make this as widely applicable as possible.
I plan to post design files and a source for the pc board by this weekend.

Hi Mark,

at first I didn't understand your question, then it started to dawn, maybe you have the touch screen doing this thing too. The touch screen is tensioned or pressed somewhere, or the cover plate is just touching it, causing the touch screen to be pressed by hand grip.
Please remove the cover on the screen side and perform the operation. If the error does not occur, it is caused by what has been described above.
If this causes it, place 4 pcs plastic washer between the pcb and the cover plate and screw it back on.

I hope it helps
73, Gyula HA3HZ

--
*** nothing is permanent only change ( ) ***

My suggestion is to remove the screen-side cover plate and see if the error phenomenon still exists.
The touch screen is particularly sensitive anywhere there is tension or pressure on the surface.
See mobile phones, mostly the side is grasped by the cover plate, it is not pressed anywhere.
All further operations should then be performed. If the fault persists, an electrical fault can be found.

73, Gyula HA3HZ

--
*** nothing is permanent only change ( ) ***

On Thu, May 28, 2020 at 04:34 AM, <david.hostetler@...> wrote:

The high impedance output of this balun is center tapped and the center tap is
grounded back to the input SMA (unless you cut the small trace on the back of
the PCB). When you connect one side of the output to the coax shield going to
the VNA, you short the winding!

Hello David and Roger,
This was the solution. Thank you very much for the hint.

I have the schematic of the balun, but I did not consider the R1 jumper.
II cut the R1 jumper on the lower side of the PCB and calibrated from 50 KHz to 100 MHz.

Now it works as expected, see NooElelc_Balun_9-1_S21_Loss.png.
And the reference from NooElect: balun_one_nine_insertion_loss_comparison_to_70mhz.png
And the setup: RF_Balun_1-9_S11_DSC08237.jpg

So, it is proved, that with a 400 Ohm resistor and a SMA plug you can measure the *Insertion Loss*.

73, Rudi DL5FA

David had a good suggestion on cutting the trace on that NanoVNA balun. I have done that on the ones I own. Here is a circuit drawing of the balun and a photo showing how to cut the track.

Roger

The high impedance output of this balun is center tapped and the center tap is grounded back to the input SMA (unless you cut the small trace on the back of the PCB). When you connect one side of the output to the coax shield going to the VNA, you short the winding!

On Wed, May 27, 2020 at 01:11 PM, <reuterr@...> wrote:

The problem arised when I exchanged this 50 Ohm resistor with a 50 cm RG-315
cable,
connected to the CH1 port of the nanoVNA.
The calibration with that 50 cm RG-316 cable was made before.
So, how can I improve the situation?

Calibrate at the end of the 50cm of cable! USe a small SMT resistor
49.9 ohm 1% works well enough or you can get exact 50 ohm in 1%
SMT.

Short and open are easy.

Did this kind of stuff for measuring antennas feed point Z (non connector) into
the ghz region.

Allison



--
-----------------
No direct email, it goes to bit bucket due address harvesting in groups.IO

Have similar issues with my new H4. The touch screen interface is inconsistent. Sometimes pressing a menu action works with the correct option highlighted. Other times not either with the wrong button selected or nothing happens at all. I have calibrated the LCD.
Simon
ZL2BRG

Have similar issues with my new H4. The touch screen interface is
inconsistent. Sometimes pressing a menu action works with the correct
option highlighted. Other times not. I have calibrated the LCD.
Simon
ZL2BRG

If the touch screen still fails to work after recalibration, the touch screen may be damaged and you can contact your dealer for replacement.

Possible Need run touch screen calibrate

Push leveler button
Push right while not select CONFIG menu
Push leveler button
Push left or right to select TOUCH CAL
Push leveler button

On screen use stylus tap on top left corner
Then on bottom right corner

Check touch screen work, run Touch test (drag on screen and see line)
If ok select SAVE

Rudi,

Here is a suggestion

- Try the test again by putting the 50 ohm load at the end of the cable instead of connecting it to CH1. Now you just have the coaxial cable connected with the same load on the end. Do you get similar S11 results for the NooElec when you put the 50 ohm load on the SMA connector?

- Connect the 50 cm cable from CH0 to CH1 and then use NanoVNA Saver to measure R+jX at the input to the cable. Are you close to 50 + j0 ohms? My results on a NanoVNA (2.8" ver 3.3) using 10 cm of RG-316 below.

Roger

I have the -H model, HW ver. 3.4, SN 20011259 build Dec 26, 2019. I have had this for about 2 weeks and it had been working nicely. Purchased from GigaParts.

1. The unit turns on and the identification screen appears.

2,Touching the screen brings up the menu bar on the right..

3. Touching the menu bar brings up the user screen with the grid. The menu remains on.

4. Touching the menu bar to selection an action turns the menu off. Nothing is selected.

5. The unit is not functioning.

Any suggestions please?

Paul, W8SBH

I ordered the updated, 3 button "F" model on February 29, 2020. I was delivered yesterday, May 26th, by the USPS. Tracking consistently showed that the parcel had been shipped some time in early March but that it was stuck in either outgoing customs or incoming customs in the U.S. The little training kit that I ordered at the same time was delivered in a timely manner and I have had it for almost 3 months now.

Hang in there, things will work out!

KM5XK

Thanks, Herb! Crystals are one of my interests.

I tried running it with Python2.x and 3.x, it apparently requires Python3.7. Haven't tried actually measuring a crystal with it yet, but this works, anyway:



jim@jim-P55A-UD3P:~/crystalweb$ python3.7 ./crystalweb.py --help
usage: crystalweb.py [-h] [--fixture] [--loss] [--theta THETA] [--stray STRAY]
[--repeat REPEAT] [--load LOAD] [--title TITLE]
[--device DEVICE] [--start START] [--stop STOP]
[--capture]

optional arguments:
-h, --help show this help message and exit
--fixture measure test fixture stray capacitance (default: False)
--loss measure test fixture loss (default: False)
--theta THETA phase angle for measuring bandwidth (default: 45)
--stray STRAY test fixture stray capacitance in pF, affects Co (default:
None)
--repeat REPEAT number of times to repeat measurements (default: 10)
--load LOAD test fixture source and load resistance (default: 50)
--title TITLE title of measurement (default: )
--device DEVICE name of serial port device (default: None)
--start START starting frequency of initial sweep (default: None)
--stop STOP stopping frequency of initial sweep (default: None)
--capture capture screenshots of the measurements as movie_xx.png
(default: False)
jim@jim-P55A-UD3P:~/crystalweb$

Hello Roger,
Thank you very much for your quick answer.
My answers are in your text:

On Wed, May 27, 2020 at 08:30 PM, Roger Need wrote:

When you replace a 50 ohm resistor load with the RG-316 cable connected to the
nanoVNA you no longer have a 50 ohm resistive load. This is due to two
factors:
- RG-316 cable is only specified as a "nominal" 50 ohm characteristic
impedance. Its characteristic impedance varies considerably with frequency as
shown below.
- The NanoVNA CH1 port is also not 50 ohms as reported by others on this site.
You can measure it yourself and see what I mean.

The net effect of these two effects is that you will see a complex impedance
at the end of the RG316 that differs from 50 ohms so the balun will not be
seeing 450 ohms and the return loss will be different.

I was my understanding, that the calibration will give at the end of the 50 cm RG-316 cable
nominal 50 Ohm impedance (reference point).

One other point. The style of 50 ohm SMA load you are using is not a good 50
ohm load. They are usually made with a 51 ohm resistor with a long lead
(inductive) and have a poor contact with the load case. So this will affect
your initial Return Loss measurement

With that 50 Ohm resistor in the picture I got a good S11 Return Loss measurement.
So, this was not my point.

The problem arised when I exchanged this 50 Ohm resistor with a 50 cm RG-315 cable,
connected to the CH1 port of the nanoVNA.
The calibration with that 50 cm RG-316 cable was made before.

So, how can I improve the situation?

73, Rudi DL5FA