|
Re: SimSmith - great, not only for Measuring resonance from coax far end.
Sorry, after having rearranged my text, some not meant to send text had remained below
73, Hans DJ7BA Sorry, please ignore this remainder. Von: DJ7BA <dj7ba@...> Gesendet: Freitag, 10. Januar 2020 22:33 An: '[email protected]' <[email protected]> Betreff: AW: [nanovna-users] SimSmith - great, not only for Measuring resonance from coax far end. Hi Jim, yes, once you got started using SimSmith, you will not want to miss it anymore. It is the perfect companion of NannVNA and NanoVNA saver or any VNA. You can not only ¡°measure resonance from coax far end¡±, but so much more. The best freeware ever programmed ¨C imho ¨C for radio amateur and RF engineering use. There is, however, one little precaution that needs to be mentioned here, as you said this about SimSmith: > Like drop in an LC element between two impedances, and it automatically calculates appropriate values to match. In a sense (when finally matched) that is correct. More generally, (when mismatched) it is not. I assume you want to get best power transfer (making best use of available power). SimSmith uses the correct Gamma calculation for T R A N S M IS S I O N L I N E S by using this formula: Gamma = (Z2 - Z1) / (Z2 + Z1) (a) with Z2 being the load impedance, Z1 the characteristic wave impedance of the line. This type (a) calculation also is the mapping formula in Smith Charts between the Z (or Y) plane and the Gamma plane. So far so good, nothing wrong. But (a) is not generally adequate in your case. Let¡¯s look at this circuit: Interconnecting two identical impedances, doesn¡¯t lead to maximum power transfer as (a) could make you think. This is a common mistake resulting from using (a) only and neglecting the difference between characteristic wave impedance of the line and impedance toward the generator (or source). In your case the impedance toward the source is given. So, instead (a) we need: Gamma = (ZL - Zs*) / (ZL + Zs) (b) with ZL being impedance toward the load, Zs the impedance toward the source, and * meaning conjugate complex. (b) sometimes is called is the "generalized reflection factor". It is valid for any, including imperfectly matched, impedances. In most publications, this general formula is not mentioned, but often the special case of perfect conjugate match is. Gamma at resonance does not get real in (a), but only in (b). Using (a) for arbitrary impedances even could result in negative SWR, that SimSmith covers up in later versions by saying SWR = |SWR|, thus ironing away the bad looking negative SWR. Why? Ward Harriman, AE6TY, program Author of SimSmith, belongs to a school that does not accept the difference above. This school, instead, teaches and believes in a doctrine ¨C that (making me shake my head) even became something like a standard: (The above I found in a glossary from ATIS or former ANSI.) The difference of (a) and (b) is marginal near resonance, that is, for well-tuned narrow band antennas. But it can become important at larger mismatch, i.e. like using a 1.8 MHz resonant antenna at 2.0 MHz, or when using way out-of-resonance antennas, i.e. an electrically short antenna, with a rig side only tuner. Having mentioned this little (worth a foot note) precaution, I anyway strongly insist: SimSmith is absolutely recommended. It's all worth the reasonably small time investment it takes to get started. I Strongly recommend it. Don¡¯t panic because of the vast number of possibilities SimSmith offers. Go, find it at Download it at and use it with great benefit. Anyone interested in more detail of (b), is invited to ask me for the derivation of (b). It takes, however, some basic complex math understanding. Please don¡¯t try to bother AE6TY with suggesting (b). I did. Save his time. He just doesn¡¯t want it. 73, Hans DJ7BA But you wanted to match two impedances by an L/C network (Tuner) made of lumped L and C (but no line) in this example. Here we have no characteristic wave impedance, but we have two impedances: One, Z2, toward the generator, as above. But another one, Z1, toward the generator. That one is NOT any characteristic wave impedance of any cable. As there are no reflections (that would be caused at a mismatched, terminated cable end), but we have a simple AC serial circuit with the following: Generator (thought as made of a constant voltage source and some Th¨¦venin internal impedance), and a load impedance (having an resistive part and a reactive part) . That's all. So there is no reflection (though generally quite often people speak of a reflection factor, as if we had a misterminated line). The above (a) will do that -in case of perfect conjugate match - a situation we often want, of course. -----Urspr¨¹ngliche Nachricht----- Von: [email protected] <mailto:[email protected]> <[email protected] <mailto:[email protected]> > Im Auftrag von Jim Allyn - N7JA Gesendet: Freitag, 10. Januar 2020 05:02 An: [email protected] <mailto:[email protected]> Betreff: Re: [nanovna-users] Measuring resonance from coax far end. On 1/9/20 3:55 PM, WB2UAQ wrote: I bet SimSmith will make it even easier but right now I don't have the patience to sit still and figure out how to run it:) You won't need any patience, it's amazingly simple. Like drop in an LC element between two impedances, and it automatically calculates appropriate values to match. Tell it you want a high pass instead of a low pass, and it automatically recalculates. |
|
Re: SimSmith - great, not only for Measuring resonance from coax far end.
Hi Jim,
yes, once you got started using SimSmith, you will not want to miss it anymore. It is the perfect companion of NannVNA and NanoVNA saver or any VNA. You can not only ¡°measure resonance from coax far end¡±, but so much more. The best freeware ever programmed ¨C imho ¨C for radio amateur and RF engineering use. There is, however, one little precaution that needs to be mentioned here, as you said this about SimSmith: > Like drop in an LC element between two impedances, and it automatically calculates appropriate values to match. In a sense (when finally matched) that is correct. More generally, (when mismatched) it is not. I assume you want to get best power transfer (making best use of available power). SimSmith uses the correct Gamma calculation for T R A N S M IS S I O N L I N E S by using this formula: Gamma = (Z2 - Z1) / (Z2 + Z1) (a) with Z2 being the load impedance, Z1 the characteristic wave impedance of the line. This type (a) calculation also is the mapping formula in Smith Charts between the Z (or Y) plane and the Gamma plane. So far so good, nothing wrong. But (a) is not generally adequate in your case. Let¡¯s look at this circuit: Interconnecting two identical impedances, doesn¡¯t lead to maximum power transfer as (a) could make you think. This is a common mistake resulting from using (a) only and neglecting the difference between characteristic wave impedance of the line and impedance toward the generator (or source). In your case the impedance toward the source is given. So, instead (a) we need: Gamma = (ZL - Zs*) / (ZL + Zs) (b) with ZL being impedance toward the load, Zs the impedance toward the source, and * meaning conjugate complex. (b) sometimes is called is the "generalized reflection factor". It is valid for any, including imperfectly matched, impedances. In most publications, this general formula is not mentioned, but often the special case of perfect conjugate match is. Gamma at resonance does not get real in (a), but only in (b). Using (a) for arbitrary impedances even could result in negative SWR, that SimSmith covers up in later versions by saying SWR = |SWR|, thus ironing away the bad looking negative SWR. Why? Ward Harriman, AE6TY, program Author of SimSmith, belongs to a school that does not accept the difference above. This school, instead, teaches and believes in a doctrine ¨C that (making me shake my head) even became something like a standard: (The above I found in a glossary from ATIS or former ANSI.) The difference of (a) and (b) is marginal near resonance, that is, for well-tuned narrow band antennas. But it can become important at larger mismatch, i.e. like using a 1.8 MHz resonant antenna at 2.0 MHz, or when using way out-of-resonance antennas, i.e. an electrically short antenna, with a rig side only tuner. Having mentioned this little (worth a foot note) precaution, I anyway strongly insist: SimSmith is absolutely recommended. It's all worth the reasonably small time investment it takes to get started. I Strongly recommend it. Don¡¯t panic because of the vast number of possibilities SimSmith offers. Go, find it at Download it at and use it with great benefit. Anyone interested in more detail of (b), is invited to ask me for the derivation of (b). It takes, however, some basic complex math understanding. Please don¡¯t try to bother AE6TY with suggesting (b). I did. Save his time. He just doesn¡¯t want it. 73, Hans DJ7BA But you wanted to match two impedances by an L/C network (Tuner) made of lumped L and C (but no line) in this example. Here we have no characteristic wave impedance, but we have two impedances: One, Z2, toward the generator, as above. But another one, Z1, toward the generator. That one is NOT any characteristic wave impedance of any cable. As there are no reflections (that would be caused at a mismatched, terminated cable end), but we have a simple AC serial circuit with the following: Generator (thought as made of a constant voltage source and some Th¨¦venin internal impedance), and a load impedance (having an resistive part and a reactive part) . That's all. So there is no reflection (though generally quite often people speak of a reflection factor, as if we had a misterminated line). The above (a) will do that -in case of perfect conjugate match - a situation we often want, of course. -----Urspr¨¹ngliche Nachricht----- Von: [email protected] <[email protected]> Im Auftrag von Jim Allyn - N7JA Gesendet: Freitag, 10. Januar 2020 05:02 An: [email protected] Betreff: Re: [nanovna-users] Measuring resonance from coax far end. On 1/9/20 3:55 PM, WB2UAQ wrote: I bet SimSmith will make it even easier but right now I don't have the patience to sit still and figure out how to run it:) You won't need any patience, it's amazingly simple. Like drop in an LC element between two impedances, and it automatically calculates appropriate values to match. Tell it you want a high pass instead of a low pass, and it automatically recalculates. |
David J Taylor
bill K7WXW