It is awesome that you respond to our comments and make new videos to answer our questions. Bravo! I hope that everyone (like myself) intends to eventually watch ALL of your "old" videos to learn from, and eventually follow your instructions forward to today 's videos. I love when you answer our questions that are older, and not just last weeks questions because there are months of subscriber questions that would be un-answered if you did not take the time to resolve video viewers' uncertainties from last months or years ago.
Try this... with the cable open, find the lowest frequency where the cable appears shorted. This will be where the cable is 1/4 wavelength. Attach a 50 ohm load to the cable and read the resistance. The cable impedance is then Zo = sqrt(Zl*Zmeasured). In the case of 75 ohm cable (but you don't know that), you should read about 112.5 ohms with a 5o ohm load at the frequency where the cable is 1/4 wavelength. So Zo = sqrt(50*112) or 75 ohms. For your 88 ohm cable you should get a reading of 154 ohms using the procedure as above. Hence, Zo = sqrt(154*50) or 88 ohms. I hope this makes sense. This way you do not need a test fixture. Just a sweep to find the 1/4 wave frequency and then measure again (at that same frequency) with a 50 ohm load attached then apply simple math. Cheers, John
One thing that would help would be to pick a pot with a smaller range and make sure the pot is a linear taper (not an audio taper). For this cable, he could use a 50 ohm resistor in series with a 100 ohm pot. That way you get a range of 50 to 150 ohms. Or, use a multi-turn pot, one that has a standard carbon base to minimize inductance.
Between a NanoVNA and a multimeter, once you figure out how to use them you can do practically anything. And what little you can't do with just those 2, you can use them to build what you do need.
Another way is to remember that a 45 degree transmission line shows a reactance value in ohms which is equal to the characteristic impedance; with cable open on the far end, find the lowest frequency where it shows shorted. Let's say 50 MHz - the length is a quarter wave. At 100 MHz it will be 1/8 wave or 45 degrees.
Might get some better resolution if your max frequency is higher than 30 MHz. Problem however may be the trim pot stops looking like a pure resistance the higher in frequency you go. I think the cable is probably 91 Ohms as that is a standard impedance you see around some places.
Funny coincidences.I just ordered yesterday from Mouser 3 meters of HUBER & SUHNER 316 Enviroflex cable at 5.5€/m (ouch) and some SMA right angle connectors 10€/piece (ouch,ouch,ouch) for my nanovna.Anyway,buy once,cry once!
is the far left on the smith chart where it is shown as short is always at ¼λ ? meaning at the far right where it is open is 0° and far left where it is short at 90° ?
yes you you start at 0 on the right, then 90 deg on left (going clockwise) then all the way back at 180 then right at 270 then back at 360. one wave goes twice around the chart
I thought with the help of your video I just found how to make phasing lines for vhf dipole feed element but I am way way far from that 90°... I don't understand how to measure phasing lines on the smith chart... how can you measure a piece of coax cable to have 90° ¼λ to feed a 145MHz dipole for example ?
@@IMSAIGuy hmm... there is something I am not doing right then... if I need a piece of coax to be ¼λ at 145MHz, I calibrate my nanovna between 144-148, select the smith chart and only connect that piece of coaxial cable with open end to the nanovna ch0 right ? if it is the case, it is not working (as the plan I am looking at state a particular lenght of .66 velocity factor) and I don't know what am I doing wrong.... but thanks for the replying and the precisions !
@@IMSAIGuy thanks I will try this again with larger frequency range... I was expecting to be right on it since I already had some measurement according to a antenna plan. Thanks for the hints !
You can also do it by measuring it on a LCR meter - you measure the short circuit inductance (so short the far end of the twisted pair) and the open circuit capacitance, and zo = sqrt(L/C). It works good on twisted pair cables. The longer the piece of cable you measure, the more accurate the measurement will be
I don't think that this is correct. What you measured at 9:45 is the impedance seen by the VNA when the phase is zero (and thus reactance is zero). This will occur when your cable length is an odd integral multiple of a quarter wavelength. At this point your measured impedance is equal to (Z0)^2/ZL (where ZO is characteristic impedance and and ZL is the load impedance). Solving for Z0 we have Z0 = sqrt(Z measured X ZL). You should have measured the load impedance at this point multiplied it by the measured impedance (VNA) and taken the square root. AT 5:10 you were misled into thinking that, because the measured impedance was 50 ohm on a known 50 ohm cable, that meant that the measured impedance on the VNA is equal to the characteristic impedance. Had you measured your load impedance at this time you would have discovered that it, too was 50 ohm and thus that Z0 = sqrt(50X50) = 50 !!. Try placing a 100 ohm termination on a known 50 ohm coax and repeating your methodology. What you will find is that your measured impedance (VNA) at the zero phase point would have been 50^2/100 = 25 ohms. Edited post for clarity (I hope).
It depends on the multi turn pot. Many are made with the same resistive material but a screw or screw and gear mechanism to move the slider. The other thing that would help would be to pick a pot with a smaller range, make sure the pot is a linear taper (not an audio taper). For this cable, he could use a 50 ohm resistor in series with a 100 ohm pot. That way you get a range of 50 to 150 ohms.
yes and no. The system i calibrated to 50 ohms and the cable will be the only piece that does not match. the termination zeros this out at the correct value. You can do a similar measurement sending a pulse down the cable and looking for the return on an oscilloscope. if terminated correctly there will be no return reflection.
I'm also going and watch many of the old videos - they are a fascinating trove of goodness - I swear I learn more practical EE from these videos than I did in an undergrad EE education where everything is presented from a theoretical / mathematical perspective, not a real world practical perspective.
It is awesome that you respond to our comments and make new videos to answer our questions. Bravo!
I hope that everyone (like myself) intends to eventually watch ALL of your "old" videos to learn from, and eventually follow your instructions forward to today 's videos. I love when you answer our questions that are older, and not just last weeks questions because there are months of subscriber questions that would be un-answered if you did not take the time to resolve video viewers' uncertainties from last months or years ago.
I have used this trick to determine the characteristic impedance of PCBs. Love your videos !
Hadn't thought of that. Thanks.
Try this... with the cable open, find the lowest frequency where the cable appears shorted. This will be where the cable is 1/4 wavelength. Attach a 50 ohm load to the cable and read the resistance. The cable impedance is then Zo = sqrt(Zl*Zmeasured). In the case of 75 ohm cable (but you don't know that), you should read about 112.5 ohms with a 5o ohm load at the frequency where the cable is 1/4 wavelength. So Zo = sqrt(50*112) or 75 ohms. For your 88 ohm cable you should get a reading of 154 ohms using the procedure as above. Hence, Zo = sqrt(154*50) or 88 ohms. I hope this makes sense.
This way you do not need a test fixture. Just a sweep to find the 1/4 wave frequency and then measure again (at that same frequency) with a 50 ohm load attached then apply simple math.
Cheers,
John
Best comment of the thread.
You are Amazing😮😊
I enjoy your videos like this one where you clarify and answer comment questions much more than you soldering together Yet Another kit again
One thing that would help would be to pick a pot with a smaller range and make sure the pot is a linear taper (not an audio taper). For this cable, he could use a 50 ohm resistor in series with a 100 ohm pot. That way you get a range of 50 to 150 ohms. Or, use a multi-turn pot, one that has a standard carbon base to minimize inductance.
I've had a nano-VNA for months and I really need to learn how to use it because this is neat. I'm going to have to watch your other videos.
Between a NanoVNA and a multimeter, once you figure out how to use them you can do practically anything. And what little you can't do with just those 2, you can use them to build what you do need.
Another way is to remember that a 45 degree transmission line shows a reactance value in ohms which is equal to the characteristic impedance; with cable open on the far end, find the lowest frequency where it shows shorted. Let's say 50 MHz - the length is a quarter wave. At 100 MHz it will be 1/8 wave or 45 degrees.
Might get some better resolution if your max frequency is higher than 30 MHz. Problem however may be the trim pot stops looking like a pure resistance the higher in frequency you go. I think the cable is probably 91 Ohms as that is a standard impedance you see around some places.
You should use a multiturn variable resistor to get a more accurate reading
Now that was very interesting. Thanks
Built one and it works - yay :) Why did you limit the span to 30MHz rather than the full span (ie to 3GHz).
9:40 get yourself a precision multiturn pot....Way more precise and easy to adjust.
Funny coincidences.I just ordered yesterday from Mouser 3 meters of HUBER & SUHNER 316 Enviroflex cable at 5.5€/m (ouch) and some SMA right angle connectors 10€/piece (ouch,ouch,ouch) for my nanovna.Anyway,buy once,cry once!
The only change I would makke would be a decent 10 turn pot (or selection of them_.
More About Where One Can Get The Right Connections to the Nano VNA.😮😮😮😮❤😊?
Inductance on top.....
Is there a part number on that tuning tool? I lost mine.
www.amazon.com/SPECTROL-008T000-TRIMMER-ADJUSTMENT-TRIMMERS/dp/B072Q3CZV3
@@IMSAIGuy thanks!
is the far left on the smith chart where it is shown as short is always at ¼λ ? meaning at the far right where it is open is 0° and far left where it is short at 90° ?
yes you you start at 0 on the right, then 90 deg on left (going clockwise) then all the way back at 180 then right at 270 then back at 360. one wave goes twice around the chart
@@IMSAIGuy awesome thank you for the precisions !
I thought with the help of your video I just found how to make phasing lines for vhf dipole feed element but I am way way far from that 90°... I don't understand how to measure phasing lines on the smith chart... how can you measure a piece of coax cable to have 90° ¼λ to feed a 145MHz dipole for example ?
a 1/4wave length of coax will go 180 degrees around a smith chart. 360 degrees of phase travels around the smith twice
@@IMSAIGuy hmm... there is something I am not doing right then... if I need a piece of coax to be ¼λ at 145MHz, I calibrate my nanovna between 144-148, select the smith chart and only connect that piece of coaxial cable with open end to the nanovna ch0 right ? if it is the case, it is not working (as the plan I am looking at state a particular lenght of .66 velocity factor) and I don't know what am I doing wrong.... but thanks for the replying and the precisions !
@@Aleziss you don't need such a small frequency range. sweep from 50 to 300 and see where you are. too short or too long.
@@IMSAIGuy thanks I will try this again with larger frequency range... I was expecting to be right on it since I already had some measurement according to a antenna plan. Thanks for the hints !
How to measure differential pair cables, such at cat5/6?
you would need a balun. this would be also 50 to 100 ohm impedance change. The spec on cat cables is 15% so don't expect a perfect 100 ohms.
You can also do it by measuring it on a LCR meter - you measure the short circuit inductance (so short the far end of the twisted pair) and the open circuit capacitance, and zo = sqrt(L/C). It works good on twisted pair cables. The longer the piece of cable you measure, the more accurate the measurement will be
I don't think that this is correct. What you measured at 9:45 is the impedance seen by the VNA when the phase is zero (and thus reactance is zero). This will occur when your cable length is an odd integral multiple of a quarter wavelength. At this point your measured impedance is equal to (Z0)^2/ZL (where ZO is characteristic impedance and and ZL is the load impedance). Solving for Z0 we have Z0 = sqrt(Z measured X ZL). You should have measured the load impedance at this point multiplied it by the measured impedance (VNA) and taken the square root. AT 5:10 you were misled into thinking that, because the measured impedance was 50 ohm on a known 50 ohm cable, that meant that the measured impedance on the VNA is equal to the characteristic impedance. Had you measured your load impedance at this time you would have discovered that it, too was 50 ohm and thus that Z0 = sqrt(50X50) = 50 !!. Try placing a 100 ohm termination on a known 50 ohm coax and repeating your methodology. What you will find is that your measured impedance (VNA) at the zero phase point would have been 50^2/100 = 25 ohms. Edited post for clarity (I hope).
would a multi turn trimmer have helped in this case?
no, need to keep inductance low
@@IMSAIGuy are all the little screw drive trimpots made with resistance wire?
@@nickcaruso I've always thought so
It depends on the multi turn pot. Many are made with the same resistive material but a screw or screw and gear mechanism to move the slider. The other thing that would help would be to pick a pot with a smaller range, make sure the pot is a linear taper (not an audio taper). For this cable, he could use a 50 ohm resistor in series with a 100 ohm pot. That way you get a range of 50 to 150 ohms.
Treat yourself to some museum putty!
How does the 50 Ohm connectors come into play? I would think that would effect your results somehow.
yes and no. The system i calibrated to 50 ohms and the cable will be the only piece that does not match. the termination zeros this out at the correct value. You can do a similar measurement sending a pulse down the cable and looking for the return on an oscilloscope. if terminated correctly there will be no return reflection.
You could so sell these! LoL
I’m running through old vids, but I’m an unbelievably weird person, so I wouldn’t base any statistical significance from my behavior.
😊 You Are Good.....😊
I'm also going and watch many of the old videos - they are a fascinating trove of goodness - I swear I learn more practical EE from these videos than I did in an undergrad EE education where everything is presented from a theoretical / mathematical perspective, not a real world practical perspective.