Re: Help Testing a Low Pass Filter
Hello GM7,
I hope the following suggestions will help.
Please send a photo of your test fixture with out the filter on top of it and some photos of how you did the SOIT terminations. This may help us figure out what your hands are doing and spot ways to help you.
I would advise you use the NanoVNA to characterize each of your components and learn the "parasitics" each have.
Parasitic components are those "free" extra elements each part we buy (or make) comes with.
Resistors can be the most ideal but all resistors come with some parasitic capacitance and inductance.
Capacitors have series resistance and inductance.
Worst of all are inductors which are "blessed" with generous amounts of resistance and capacitance. They are so "blessed" that you will want to get an estimate of the size of the parasitic elements and include them in the filter simulation to decide if they can be ignored or if you tweak another component to compensate if you can.
You approximate the parasitic elements by finding the peaks and valleys of the series and parallel measurements of the components. If you need help understanding peaks and valleys of the series and parallel measurements of the components that would be a good conversation too.
I was in my early 20s when a mentor showed me how to measure coil impedance on a network analyzer and was so baffled by what I saw. We were measuring a common mode choke for an AC line input filter and I assumed since the FCC requirements went (working from memory here) something like 450KHz to 30 MHz that the choke would be inductive over that range. Again IIRC, the choke stopped blocking way lower than 30 MHz and then had several saw tooth like shapes in the impedance versus frequency plot. It was a learning moment.
Hope this furthers the conversation.
F. Lee Erickson
toggle quoted message
Show quoted text
-----Original Message----- From: GM7 Sent: Thursday, June 04, 2020 1:17 PM To: [email protected]Subject: [nanovna-users] Help Testing a Low Pass Filter Hi All, I'm trying to test a set of Low Pass Filters for 40m, 30m, 20m & 10m. I'm getting consistent results for each of them but not the results I expected. I'm struggling to establish whether the problem is with the filters or with my test setup / calibration or even with the NanoVNA. I've repeated this with and without a test fixture and got essentially the same results. I attach the test fixture (see photo) and then calibrate (reset, followed by open, short, isolation, thru). I then check the smith chart with short, open and 50ohm loads to sanity check the calibration. The attached image shows the 30m LPF. It doesn't have the passband I would expect or the roll-off I would expect. Instead it's almost a linear roll-off over the range under test. Can anyone help? For example by either correcting my test methodology or offering any advice. Thanks in advance.
|
Re: Help Testing a Low Pass Filter
I think your calibration is off. Calibrate at the cable ends rather than at the test fixture and make sure you do a through cal as well since you are measuring the transmission response of your filter. With the capacitor values you mentioned, the test fixture won¡¯t add significant error.
Gary
W9TD
|
Re: Help Testing a Low Pass Filter
I simulated your values with RFSIM99, and I get a nice, seemingly
Butterworth, response, cutoff (-3 dB) at 9.12 MHz.
Please check your values and connections topology, something is way off.
73 de Jose, CO2JA
|
Re: Help Testing a Low Pass Filter
Hi Jos,
Thanks for the reply. The photo doesn't show the setup very clearly. The LPF is a small board with male pins at either end. The fixture I built is a small board with female sockets to match the pins of the LPF. These sockets are connected to the SMA sockets on the board.
I also built two very small plugs to use for calibration. One with the pins shorted and one with a 50ohm resistor across the two pins. To calibrate I connect the fixture without the LPF to the NanoVNA using the two short SMA cables. I then follow the normal calibration routine before fitting the LPF.
|
Re: Help Testing a Low Pass Filter
Hi GM7,
You write "I attach the test ficture (see photo) and then calibrate"? but what I see in the photo is not a testfixture but the Lowpaasfilter itself.??? You have to do a 2 port calibration just with the testcables and the calibration devices,? so the short open and load connected to the CH0 cable, and isolation with the load connected to the CH1 cable . The through you calibrate with the CH0 and? CH1 cables interconnected by the female to female adapter. After that you can connect the lowpassfilter and measure it.
Success
Jos
Op 4-6-2020 om 19:17 schreef GM7:The isolation you calibrate with the load connected to th CH1 cable
toggle quoted message
Show quoted text
Hi All,
I'm trying to test a set of Low Pass Filters for 40m, 30m, 20m & 10m. I'm getting consistent results for each of them but not the results I expected. I'm struggling to establish whether the problem is with the filters or with my test setup / calibration or even with the NanoVNA.
I've repeated this with and without a test fixture and got essentially the same results. I attach the test fixture (see photo) and then calibrate (reset, followed by open, short, isolation, thru). I then check the smith chart with short, open and 50ohm loads to sanity check the calibration.
The attached image shows the 30m LPF. It doesn't have the passband I would expect or the roll-off I would expect. Instead it's almost a linear roll-off over the range under test.
Can anyone help? For example by either correcting my test methodology or offering any advice. Thanks in advance.
|
Re: Help Testing a Low Pass Filter
It sure looks to me like the filter is for a lower frequency range than
what you're sweeping. Maybe mixed up the component sets?
toggle quoted message
Show quoted text
On Thu, Jun 4, 2020 at 3:55 PM GM7 <geeemmseven@...> wrote: I haven't measured the parts yet. I have checked the capacitors are
correct based on the markings and that the torroids have the correct number
of turns.
All 4 filters that I built show the same response so I suspected that my
test setup or methodology was likely to be wrong.
I have another kit so I could build it measuring the components as I go.
Is there a standard setup for measuring filter response with the NanoVNA?
--
Carey Fisher
careyfisher@...
|
Re: Help Testing a Low Pass Filter
I haven't measured the parts yet. I have checked the capacitors are correct based on the markings and that the torroids have the correct number of turns.
All 4 filters that I built show the same response so I suspected that my test setup or methodology was likely to be wrong.
I have another kit so I could build it measuring the components as I go.
Is there a standard setup for measuring filter response with the NanoVNA?
|
Re: Help Testing a Low Pass Filter
Gm7
double check your C's value and measure all L that are in your board
Al
CX8AT
|
Re: Help Testing a Low Pass Filter
Have you tested the parts to make sure you have the correct values?
Marco
toggle quoted message
Show quoted text
On Thu, Jun 4, 2020, 2:24 PM GM7 <geeemmseven@...> wrote: Thanks Dave,
Your simulation is the response I expected. Unfortunately i's not the
response I'm measuring. I re-calibrated and then used the Smith Chart as a
sanity check. I got the expected results for open, short and 50 ohm load.
When I built the fixture I connected the ground permanently so that the
ground is connected between CH0 and CH1 even when the LPF is not in place.
I've re-tested without this link and got the same results. In general
terms I'm not sure if this link is a good idea or if I should remove the
link.
|
Thanks for your response. I had seen that and other related questions on Stack Overflow. I tried some of the suggestions given, as mentioned in the original post.
|
Re: Help Testing a Low Pass Filter
Thanks Dave,
Your simulation is the response I expected. Unfortunately i's not the response I'm measuring. I re-calibrated and then used the Smith Chart as a sanity check. I got the expected results for open, short and 50 ohm load.
When I built the fixture I connected the ground permanently so that the ground is connected between CH0 and CH1 even when the LPF is not in place. I've re-tested without this link and got the same results. In general terms I'm not sure if this link is a good idea or if I should remove the link.
|
white screen and flashing
After a full charge (usb connector led steady) , power it on, get white screen, with the led next to the power switch flashing. Ring a bell? Haven't found where in the users the flashing is mentioned yet. Sometimes it will start normally then go white at what seems like random timing.
Looks like lots of different flavors ....nanovna-XXX. Suggestion as to how can tell them all apart? and which one this is if can tell by peek, attached
rich
|
Re: Help Testing a Low Pass Filter
I modeled your 40-Meter LPF using your values on ELSIE. Please see
attachment for the results of what you should measure for pass
characteristics (S21).
Dave - W?LEV
toggle quoted message
Show quoted text
On Thu, Jun 4, 2020 at 12:05 PM GM7 <geeemmseven@...> wrote: Hi, thanks for the replies.
It's a QRP Labs filter which uses the 7 element design shown in the
attached schematic.
I attached the wrong image in the first post. That's the response of the
40m filter, not the 30m (which was very similar). The values of the 40 m
filter are:
------ 1.3uH ------- 1.7uH ------- 1.3uH ------
| | | |
270pF 680pF 680pF 270pF
| | | |
I'll re-test the 40m filter following David's suggestions and post the
results.
--
*Dave - W?LEV*
*Just Let Darwin Work*
|
Has anyone had a similar issue yes:
|
Re: Help Testing a Low Pass Filter
Hi, thanks for the replies.
It's a QRP Labs filter which uses the 7 element design shown in the attached schematic.
I attached the wrong image in the first post. That's the response of the 40m filter, not the 30m (which was very similar). The values of the 40 m filter are:
------ 1.3uH ------- 1.7uH ------- 1.3uH ------
| | | |
270pF 680pF 680pF 270pF
| | | |
I'll re-test the 40m filter following David's suggestions and post the results.
|
Re: Help Testing a Low Pass Filter
Which schematic and values did you use? That amplitude curve is weird.
Also, VSWR is very high on marker 3 (30 m)
Jos¨¦, CO2JA
toggle quoted message
Show quoted text
On 6/4/20, GM7 <geeemmseven@...> wrote: Hi All,
I'm trying to test a set of Low Pass Filters for 40m, 30m, 20m & 10m. I'm
getting consistent results for each of them but not the results I expected.
I'm struggling to establish whether the problem is with the filters or with
my test setup / calibration or even with the NanoVNA.
I've repeated this with and without a test fixture and got essentially the
same results. I attach the test fixture (see photo) and then calibrate
(reset, followed by open, short, isolation, thru). I then check the smith
chart with short, open and 50ohm loads to sanity check the calibration.
The attached image shows the 30m LPF. It doesn't have the passband I would
expect or the roll-off I would expect. Instead it's almost a linear
roll-off over the range under test.
Can anyone help? For example by either correcting my test methodology or
offering any advice. Thanks in advance.
|
Re: Help Testing a Low Pass Filter
True. Something does not look right. Rather than running 3 through 50
MHz, try setting, calibrating, and measuring 5 through 20 MHz. Also,
before you measure, to check your cal., do a Smith chart of short, open,
load, AND through just for sanity of the cal.
Your S11 looks, 'maybe', OK, the rest doesn't look as you or I would
expect. Also, once the filter is in place for measurement, flex the cables
and all SMA connections to be sure they are tight. When on my last job in
Albuquerque before moving to N. Colorado, everything was SMA or Type N
connectors. I carried a 11/16 open end wrench on my key ring. Don't over
torque, but be sure they are all tight.
Dave - W?LEV
toggle quoted message
Show quoted text
On Thu, Jun 4, 2020 at 11:18 AM GM7 <geeemmseven@...> wrote: Hi All,
I'm trying to test a set of Low Pass Filters for 40m, 30m, 20m & 10m. I'm
getting consistent results for each of them but not the results I
expected. I'm struggling to establish whether the problem is with the
filters or with my test setup / calibration or even with the NanoVNA.
I've repeated this with and without a test fixture and got essentially the
same results. I attach the test fixture (see photo) and then calibrate
(reset, followed by open, short, isolation, thru). I then check the smith
chart with short, open and 50ohm loads to sanity check the calibration.
The attached image shows the 30m LPF. It doesn't have the passband I
would expect or the roll-off I would expect. Instead it's almost a linear
roll-off over the range under test.
Can anyone help? For example by either correcting my test methodology or
offering any advice. Thanks in advance.
--
*Dave - W?LEV*
*Just Let Darwin Work*
|
Help Testing a Low Pass Filter
Hi All,
I'm trying to test a set of Low Pass Filters for 40m, 30m, 20m & 10m. I'm getting consistent results for each of them but not the results I expected. I'm struggling to establish whether the problem is with the filters or with my test setup / calibration or even with the NanoVNA.
I've repeated this with and without a test fixture and got essentially the same results. I attach the test fixture (see photo) and then calibrate (reset, followed by open, short, isolation, thru). I then check the smith chart with short, open and 50ohm loads to sanity check the calibration.
The attached image shows the 30m LPF. It doesn't have the passband I would expect or the roll-off I would expect. Instead it's almost a linear roll-off over the range under test.
Can anyone help? For example by either correcting my test methodology or offering any advice. Thanks in advance.
|
Dear Cliff,
Thank you, my biggest problem that i am new in linux enviroment, so far I developed under windows. I can try to understand your advice...
If I have results or new question then I will come back...
Best 73
|
Kedves Gyula,
K?szi, megn¨¦zem...
Nagy 73, Zoli
|
Gary W9TD
Jose Amador
Ignilux
Oristo
Zoli