Usually with the NanoVNA you need to recalibrate it every time you change the frequency range because the calibration can only hold a limited number of data points. While your point is still correct, the data displayed when you moved from 900 MHz to 420 MHz is probably completely inaccurate. Great video though, thanks!
Thanks. Good point, and something that should be checked. But with the NanoVNA version I have at least, the checks I've done indicate the unit does interpolation of correction factors - and cal-checks in the new frequency ranges show the interpolation to work well, as long as one goes from a wider span to a more narrow span within the original frequency range. I've been amazed at this NanoVNA-F unit. Unlike the HP8753 units I've used in labs that do have a cal issue if not recalibrated each time, this thing seems to do the needed interpolation. Maybe the NanoVNA-F has different firmware than others? A good test is to look at S11 of an open (or short) on a Smith chart and confirm the point is still on the far right (or left, respectively) - and use a thru-cable and verify S21 is still 0 dB. Every time I've done this using my standard saved 50 kHz to 1500 MHz cal, it has passed these tests on new frequency ranges inside the 50k to 1.5G range. Of course if the cable length is changed, S11 will be rotated and a re-cal may be needed if S11 phase is important or the new cable has significantly different loss. That said, I totally agree, in general. Calibration checks are important to be sure measurements are meaningful !
Thanks. I got it from Amazon. The board description there is: "SP1-50x50-G (Two Pack) SMTpads, Size 1, 50x50mil Pads, Unplated Holes to Ground Plane, 2 Sided PCB, Size 1 = 50 x 80mm (1.97 x 3.15in)" I really wish they had similar boards with 100x100 mil pads. That's what we used in our courses and they're easier to work with IMO. But they're hard to find and I ended up making some of my own (see the "Radio Design 101" episodes) 73's
MFJ needs to come up with a protected injection port kit that amateurs can incorporate to their tuners. Quick connect BNC connector addon afront the tuner. Probably be a big hit in the long run. Nano's Rock!
Thanks. Here's a link to a page in the companion website that has all the class stuff. The antenna hand-out is the second item on this page. ecefiles.org/rf-circuits-course-section-11/ (It took me a while to find where it was in all the uploaded materials, but I think this is what you're looking for.)
Good question. I think I agree. But I would include losses in currents induced into the case-metal due to the magnetic fields stabbing through it at the ends of the coil. I think that will induce some image currents (like proximity effect, but not with the turns proper). Hard to say which of all of these is dominant...
Hello and thank you for such good learning material. May I ask you a couple of questions about the last part of the video, when measuring different configurations for matching the resistor? The first one is: when the frequency response is broader, shouldn't the insertion loss be greater? The second one: intuition would tell me that the T network with the inductor in parallel and the capacitor in series would make up a low pass filter... But the readout of the NanoVNA suggests it acts as a bandpass filter. How could this be explained? Thank you! Miguel
Good questions. The answers come from tuned RLC circuits and Q (quality factor, which is related to bandwidth). A "high-Q" makes bandwidth less, but that results in larger currents ringing in the LC tank circuits in the T network, and higher power loss. Here's a video that goes into tuned circuits and Q in some depth: ua-cam.com/video/He0-X6FCLMo/v-deo.html . In the case of the second question, it's actually a high-pass response. High frequencies pass through the two capacitors from input to output and the impedance of the shunt L to ground is high at high frequency, so it doesn't affect things much. But its more involved. It's highpass, but also bandpass. At intermediate frequencies, the response actually peaks at the resonant frequency of the back-to-back LC circuits. Around 20:30 in the video, we can see that the circuit is actually equivalent to two resonant LC circuits back to back. These resonate at the peak response frequency and perform the matching. At lower and higher frequencies, the tuner blocks signals or passes them without matching, respectively. Hope that helps. 73's
Got them from Amazon. Search for "SP1-50x50-G (Two Pack) SMTpads". They have 50 mil pitch pads on the top and a solid ground plane on the bottom. Every 4th pad on the top has a plated-thru hole to the ground plane. They work for some things, but honestly you have to have good soldering tips, very fine solder, and magnifiers - otherwise the pads can burn off. But the concept is good for VHF+ radio work, given the ground plane. We used something very similar in our radio course designing/building FM receivers, although the pads were 80 mil on 100 mil centers. We drilled our own vias where ground connections were needed and stuffed them with a wire to the backside ground plane in lieu of a PTH. Worked well for building things with 0603 or 0805 parts, hand-wound inductors, and discrete SOT23 transistors (or old-style DIP chips with 100 mil spacings)...
@@MegawattKS thanks, they look great for smd prototyping indeed. I see that Mouser also has them, maybe I'll throw a few in for my next order there. The Amazon listing has quite a high shopping cost for ordering them to .nl
Nice video. But did you notice the error in the MFJ-945E schematic? The toroid T2 is never switched in. Typical MFJ quality control. Let's hope the schematic is wrong. LB8X
Good catch ! Definitely a typo in the schematic. I pulled the unit and on the A position, it ties the common to ground as it should. (FWIW, in the next vid on mixers I actually flagged an error in the SA602 IC's simplified internal schematic. It seems to have originated with Signetics in the early years, but still shows up in NXP's SA602 revised datasheet two decades later ;-)
Yes - looks like it is one of the handouts I put on the web for the students a while back :-) See 6th bulleted item here: ece.k-state.edu/about/people/faculty/kuhn/handouts.html
Using logmag in db instead of VSWR is almost as bad as displaying the period of the signal instead of it's frequency, or talking about the bands in feet instead of meters :) Hope to hear you in the 65.6 feet band, around 70.422 nS !
While I agree that VSWR numbers are more familiar (until S11 is embraced and becomes familiar), both S11 and SWR measure the same thing fundamentally - the amount of signal voltage (or power) reflected from the antenna relative to the forward wave. Only the numbers are different. S11 in logmag (dB) format is convenient if one just looks for the frequency and amount of dip. This is the same as for SWR where we are also looking for the dip, but SWR values are different as you noted. A simple rule of thumb is that an S11 dip of 10 dB (one division down from 0 dB full reflection off-resonance, when in logmag mode with default scaling) is usually 'good enough'. The corresponding VSWR is 2:1. While I admit that 2:1 sounds not great (since I was trained early on to want 1.1:1 or 1.2:1), it's actually pretty OK. S11= -10 dB = 0.1 power ratio, which means only 10% of the power is reflected from the antenna, so 90% is radiated. That's only a fraction of an S-unit at a distant receiver. The transmitter may prefer something like S11 = -15 or -20 dB (1.4 or 1.2 VSWR) so the impedance it sees looking into the coax is closer to 50 Ohms, but generally -10 dB is OK for most finals. Hope that helps. 73s
I've looked at a lot of training videos and when I watch yours, I understand and retain the knowledge you impart. Thank-you!
Hey now! A long wire antenna can be a good antenna! lol Thanks for the videos. Glad I found your channel.
Thanks. Yes - a long-wire was very the first antenna I used. Wish I had had a NanoVNA back then!
thanksyou for contributing to all us out here, Regards from VK3
great videos, I’m really enjoying them and learning a lot. Thanks!!
Really enjoyed your instructional style. Thanks for the great content!
Thank you !
Always excellent videos !!! Thanks !!! N2FH
Thank you ! 73
Excellent! Thank you.
Thanks ! Glad you enjoyed it .
Thank you! Great video. I just had to go out and buy a VNA.
Thanks. Enjoy the NanoVNA. It's an amazing instrument !
Usually with the NanoVNA you need to recalibrate it every time you change the frequency range because the calibration can only hold a limited number of data points. While your point is still correct, the data displayed when you moved from 900 MHz to 420 MHz is probably completely inaccurate. Great video though, thanks!
Thanks. Good point, and something that should be checked. But with the NanoVNA version I have at least, the checks I've done indicate the unit does interpolation of correction factors - and cal-checks in the new frequency ranges show the interpolation to work well, as long as one goes from a wider span to a more narrow span within the original frequency range. I've been amazed at this NanoVNA-F unit. Unlike the HP8753 units I've used in labs that do have a cal issue if not recalibrated each time, this thing seems to do the needed interpolation. Maybe the NanoVNA-F has different firmware than others? A good test is to look at S11 of an open (or short) on a Smith chart and confirm the point is still on the far right (or left, respectively) - and use a thru-cable and verify S21 is still 0 dB. Every time I've done this using my standard saved 50 kHz to 1500 MHz cal, it has passed these tests on new frequency ranges inside the 50k to 1.5G range. Of course if the cable length is changed, S11 will be rotated and a re-cal may be needed if S11 phase is important or the new cable has significantly different loss. That said, I totally agree, in general. Calibration checks are important to be sure measurements are meaningful !
How to design the HF antenna tuner bypass function to support V/U band? (for IC-705, because this model only has one BNC output.)
Sorry - don't have any good ideas for that. I'm not familiar with that ham radio. Maybe an external coax switch?
Thanks 👍🇬🇧
Again, great videos! Could you give me the detail about the board you used here? (where you got it and what the model number is, etc.) VA3GPJ
Thanks. I got it from Amazon. The board description there is: "SP1-50x50-G (Two Pack) SMTpads, Size 1, 50x50mil Pads, Unplated Holes to Ground Plane, 2 Sided PCB, Size 1 = 50 x 80mm (1.97 x 3.15in)"
I really wish they had similar boards with 100x100 mil pads. That's what we used in our courses and they're easier to work with IMO.
But they're hard to find and I ended up making some of my own (see the "Radio Design 101" episodes)
73's
MFJ needs to come up with a protected injection port kit that amateurs can incorporate to their tuners. Quick connect BNC connector addon afront the tuner. Probably be a big hit in the long run. Nano's Rock!
Great video. Where can I get a copy of that antenna hand out that you were working with?
Thanks. Here's a link to a page in the companion website that has all the class stuff. The antenna hand-out is the second item on this page. ecefiles.org/rf-circuits-course-section-11/ (It took me a while to find where it was in all the uploaded materials, but I think this is what you're looking for.)
Regarding insertion loss. Which component in the tuner is exhibiting the loss? I’m guessing inductor winding proximity and skin effect losses.
Good question. I think I agree. But I would include losses in currents induced into the case-metal due to the magnetic fields stabbing through it at the ends of the coil. I think that will induce some image currents (like proximity effect, but not with the turns proper). Hard to say which of all of these is dominant...
Hello and thank you for such good learning material.
May I ask you a couple of questions about the last part of the video, when measuring different configurations for matching the resistor?
The first one is: when the frequency response is broader, shouldn't the insertion loss be greater?
The second one: intuition would tell me that the T network with the inductor in parallel and the capacitor in series would make up a low pass filter... But the readout of the NanoVNA suggests it acts as a bandpass filter. How could this be explained?
Thank you!
Miguel
Good questions. The answers come from tuned RLC circuits and Q (quality factor, which is related to bandwidth). A "high-Q" makes bandwidth less, but that results in larger currents ringing in the LC tank circuits in the T network, and higher power loss. Here's a video that goes into tuned circuits and Q in some depth: ua-cam.com/video/He0-X6FCLMo/v-deo.html . In the case of the second question, it's actually a high-pass response. High frequencies pass through the two capacitors from input to output and the impedance of the shunt L to ground is high at high frequency, so it doesn't affect things much. But its more involved. It's highpass, but also bandpass. At intermediate frequencies, the response actually peaks at the resonant frequency of the back-to-back LC circuits. Around 20:30 in the video, we can see that the circuit is actually equivalent to two resonant LC circuits back to back. These resonate at the peak response frequency and perform the matching. At lower and higher frequencies, the tuner blocks signals or passes them without matching, respectively. Hope that helps. 73's
Interesting one. Question, what kind of breadboard PCB are you using there?
Got them from Amazon. Search for "SP1-50x50-G (Two Pack) SMTpads". They have 50 mil pitch pads on the top and a solid ground plane on the bottom. Every 4th pad on the top has a plated-thru hole to the ground plane. They work for some things, but honestly you have to have good soldering tips, very fine solder, and magnifiers - otherwise the pads can burn off. But the concept is good for VHF+ radio work, given the ground plane. We used something very similar in our radio course designing/building FM receivers, although the pads were 80 mil on 100 mil centers. We drilled our own vias where ground connections were needed and stuffed them with a wire to the backside ground plane in lieu of a PTH. Worked well for building things with 0603 or 0805 parts, hand-wound inductors, and discrete SOT23 transistors (or old-style DIP chips with 100 mil spacings)...
@@MegawattKS thanks, they look great for smd prototyping indeed. I see that Mouser also has them, maybe I'll throw a few in for my next order there. The Amazon listing has quite a high shopping cost for ordering them to .nl
Nice video. But did you notice the error in the MFJ-945E schematic? The toroid T2 is never switched in. Typical MFJ quality control. Let's hope the schematic is wrong. LB8X
Good catch ! Definitely a typo in the schematic. I pulled the unit and on the A position, it ties the common to ground as it should. (FWIW, in the next vid on mixers I actually flagged an error in the SA602 IC's simplified internal schematic. It seems to have originated with Signetics in the early years, but still shows up in NXP's SA602 revised datasheet two decades later ;-)
Is the document with the antenna types available somewhere?
Yes - looks like it is one of the handouts I put on the web for the students a while back :-) See 6th bulleted item here: ece.k-state.edu/about/people/faculty/kuhn/handouts.html
@@MegawattKS thanks, that's awesome.
Using logmag in db instead of VSWR is almost as bad as displaying the period of the signal instead of it's frequency, or talking about the bands in feet instead of meters :) Hope to hear you in the 65.6 feet band, around 70.422 nS !
While I agree that VSWR numbers are more familiar (until S11 is embraced and becomes familiar), both S11 and SWR measure the same thing fundamentally - the amount of signal voltage (or power) reflected from the antenna relative to the forward wave. Only the numbers are different. S11 in logmag (dB) format is convenient if one just looks for the frequency and amount of dip. This is the same as for SWR where we are also looking for the dip, but SWR values are different as you noted. A simple rule of thumb is that an S11 dip of 10 dB (one division down from 0 dB full reflection off-resonance, when in logmag mode with default scaling) is usually 'good enough'. The corresponding VSWR is 2:1. While I admit that 2:1 sounds not great (since I was trained early on to want 1.1:1 or 1.2:1), it's actually pretty OK. S11= -10 dB = 0.1 power ratio, which means only 10% of the power is reflected from the antenna, so 90% is radiated. That's only a fraction of an S-unit at a distant receiver. The transmitter may prefer something like S11 = -15 or -20 dB (1.4 or 1.2 VSWR) so the impedance it sees looking into the coax is closer to 50 Ohms, but generally -10 dB is OK for most finals. Hope that helps. 73s