|
I studied quite a few documents dealing with Z-Match networks, schematics, construction desrcriptions, mesurement results and the more.
I am looking for information regarding functional theory, design tools, network analysis and optimizing performance of a Z-Match.
Also it would be most usefull to have a Java Script tool like the "T Network Tuner Simulation".
I created a model in "Micro Cap 12", it gives meaningful results, but I am using a very tedious cut-and-try method.
Any contributions , be it in PSpice, OrCad and similar network simulation programs would be highly appreciated.
Just in case I am familiar with the relevant articles at .
73, Robert W4BCZ
|
After looking at several online configurations and schematics of the Z-match network for both balanced and unbalanced loads, I've come to a conclusion.? This is nothing more or less than a multi-tapped pi-network.? Where do pi-networks excel?? Transforming high-Z to low-Z.?
Please have a look at the attachment where I've redrawn the network to better communicate its function as a matching network.? And where do most of the losses occur in a matching network, a.k.a., antenna "tuner"?? In the inductor.? Personally, with the report of hi-Q matching "solutions" and touchy and influence of hand capacitance, I would not consider building or using one of these. ?
Analyze what is presented, understand it, and make an educated conclusion, not based on here-say and opinion or "everybody uses this or that". ? I would not grace my station with one of these.
Dave - W?LEV
I studied quite a few documents dealing with Z-Match networks, schematics, construction desrcriptions, mesurement results and the more.
I am looking for information regarding functional theory, design tools, network analysis and optimizing performance of a Z-Match.
Also it would be most usefull to have a Java Script tool like the "T Network Tuner Simulation".
I created a model in "Micro Cap 12", it gives meaningful results, but I am using a very tedious cut-and-try method.
Any contributions , be it in PSpice, OrCad and similar network simulation programs would be highly appreciated.
Just in case I am familiar with the relevant articles at .
73, Robert W4BCZ
--
Please join me.
Dave - W?LEV
|
Dave,
Maybe you want to have a look at this:
73, Robert W4BCZ
|
None of these canned design tools address network efficiency / losses.? Nor the fact that the L-network is the ONLY matching topology that yields a single matched adjustment of its two reactive components.? And, further, once that is established, it is operating at maximum efficiency.? No other matching network topology can make that claim.? That's why I use an L-network.? The tapped capacitor network is pretty good as well, but is not addressed.
Dave - W?LEV
Dave,
Maybe you want to have a look at this:
73, Robert W4BCZ
--
Please join me.
Dave - W?LEV
|
SimNec does a good job of addressing losses. Thus far it has not steered me
off into the Twilight Zone.
|
Dave,
I agree with you. Nothing beats the L-network as far as impedance matching/transformation goes. Reversing input and output terminals allows matching low- and high-impedance loads. Also, as you mentioned already, there is only one single L-C combination for achieving a match to any complex load. This can very easily be demonstrated on the Smith-Chart.
I have a QRP transceiver with a built-in manual Z-match and just wanted to dig-in deeper into it's operation principles.
I don't pay any attention to any "hear-and-say" reports whatsoever.
I highly appreciate your contributions to many subjects here on groups.io.
Regards,
Robert W4BCZ
|
Thanks, Robert, W4BCZ.? I just attempt to dispel some of the mountain of snake oil within the amateur beliefs and lore.? And, of course, help others.
Dave - W?LEV
Dave,
I agree with you. Nothing beats the L-network as far as impedance matching/transformation goes. Reversing input and output terminals allows matching low- and high-impedance loads. Also, as you mentioned already, there is only one single L-C combination for achieving a match to any complex load. This can very easily be demonstrated on the Smith-Chart.
I have a QRP transceiver with a built-in manual Z-match and just wanted to dig-in deeper into it's operation principles.
I don't pay any attention to any "hear-and-say" reports whatsoever.
I highly appreciate your contributions to many subjects here on .
Regards,
Robert W4BCZ
--
Please join me.
Dave - W?LEV
|
Analyzing the Z-match’s operation is complicated. ?Here are some quick comments.
?
The pi-network analogy suggested by W0LEV doesn’t "work" because the two caps within the usual pi network are independently adjustable, as is the inductor (either with taps or a variable inductor). ?In the Z-match, ?by contrast, the caps are ganged and the inductor itself is fixed (but not the effective inductance within the circuit). ?In addition, you can replace the multiband tank circuit in the usual Z-match configuration with a single band tank circuit (for a narrower operating range). ?That leaves only one tuning cap within the tank circuit. ?(As an irrelevant aside, despite what Dave indicates, the pi-network will match either high to to low impedances, or the reverse.)
?
The best analogy I’ve run across (and used in my 1996 article on the “single-coil” Z-match) is to see the Z-match as an L-network, but with link-coupled output. ?(I borrowed this analysis from a Z-match article back in the mid-1950s in the Radio Amateur’s Handbook.) ?The series X (the input cap) is on the input side of the L-network. ?The parallel X (the tank circuit) is on the output side, detuned from resonance enough to produce the right amount of XL (or, I suppose, rarely XC). ?Of course, the multiband tank circuit also tunes two frequency ranges simultaneously. ?But it’s more complicated than that. The parallel X within the L-network is affected by where the input cap is tapped into the inductor within the tank circuit, and the output impedance of the Z-match is affected by the size of the output link and the degree of its coupling to the tank circuit. ?Also, in my version of the Z-match the two output links increase the matching range. ?(But in his variation on my circuit some years ago in the ARRL Handbook, Phil Salas, AD5X, used a single 8-turn link.)
?
You can get my fuller analysis from my article in volume 5 of ?the ARRL Antenna Compendium (1996), which I’ve attached in pdf format, and more from my article in volume 3 of the Antenna Compendium (1992). ?
?
Some additional notes on the 1996 article appear online after an earlier draft version of the article, here: ? ? ? But forget the comments in the draft on measuring efficiency! ?The method devised by Frank Witt, AI1H, mentioned in the published version, is much better.
?
The fullest attempt I’ve seen to analyze the Z-match was in an article by Claude Franz, DJ0OT, that I ran into 10 or 15 years ago. ?Judging by dates in its footnotes, it was done in the early 2000s. I don’t find it now on the Internet, so I’ll attach a pdf of it. ?It focuses on a slightly different version ?of the “single-coil” Z-match, but I don’t think that affects the analysis.
?
Charlie, W6JJZ
|
You are correct.? The pi-network analogy is incorrect.? I like your single-band Z-Match schematic much more than the original:
It appears to me that C2 resonants L1 based on the variable input coupling through C1.? The ultimate impedance across L1 is influenced by the presented source impedance to the left of C1 through the value of C1.? The center tap on L1 forces the system to be differential wrt the left side for link coupling.? This is the purpose of the "ground" connection of the center of L1.?
I'm not going to plug this into LT-Spice or other simulation applications, but that could be easily done.? It's an interesting network.
Dave - W?LEV
On Wed, Jul 31, 2024 at 9:52?PM Charlie Lofgren W6JJZ via <clofgren= [email protected]> wrote:
Analyzing the Z-match’s operation is complicated.? Here are some quick comments.
?
The pi-network analogy suggested by W0LEV doesn’t "work" because the two caps within the usual pi network are independently adjustable, as is the inductor (either with taps or a variable inductor).? In the Z-match, ?by contrast, the caps are ganged and the inductor itself is fixed (but not the effective inductance within the circuit).? In addition, you can replace the multiband tank circuit in the usual Z-match configuration with a single band tank circuit (for a narrower operating range).? That leaves only one tuning cap within the tank circuit. ?(As an irrelevant aside, despite what Dave indicates, the pi-network will match either high to to low impedances, or the reverse.)
?
The best analogy I’ve run across (and used in my 1996 article on the “single-coil” Z-match) is to see the Z-match as an L-network, but with link-coupled output. ?(I borrowed this analysis from a Z-match article back in the mid-1950s in the Radio Amateur’s Handbook.) ?The series X (the input cap) is on the input side of the L-network.? The parallel X (the tank circuit) is on the output side, detuned from resonance enough to produce the right amount of XL (or, I suppose, rarely XC).? Of course, the multiband tank circuit also tunes two frequency ranges simultaneously.? But it’s more complicated than that. The parallel X within the L-network is affected by where the input cap is tapped into the inductor within the tank circuit, and the output impedance of the Z-match is affected by the size of the output link and the degree of its coupling to the tank circuit.? Also, in my version of the Z-match the two output links increase the matching range. ?(But in his variation on my circuit some years ago in the ARRL Handbook, Phil Salas, AD5X, used a single 8-turn link.)
?
You can get my fuller analysis from my article in volume 5 of ?the ARRL Antenna Compendium (1996), which I’ve attached in pdf format, and more from my article in volume 3 of the Antenna Compendium (1992). ?
?
Some additional notes on the 1996 article appear online after an earlier draft version of the article, here: ? ? ? But forget the comments in the draft on measuring efficiency!? The method devised by Frank Witt, AI1H, mentioned in the published version, is much better.
?
The fullest attempt I’ve seen to analyze the Z-match was in an article by Claude Franz, DJ0OT, that I ran into 10 or 15 years ago.? Judging by dates in its footnotes, it was done in the early 2000s. I don’t find it now on the Internet, so I’ll attach a pdf of it.? It focuses on a slightly different version ?of the “single-coil” Z-match, but I don’t think that affects the analysis.
?
Charlie, W6JJZ
--
Please join me.
Dave - W?LEV
|
Re Dave's follow-up email:? The circuit in Fig. 2 (excerpted by Dave) is the one I used way back when the NorCal QRP club asked me to design a Z-match for 40 meters , to sell as a kit.? It turned out that the tuning range extended much higher than 40 (thru 20 or 15 meters and sometimes 10, depending on the magnitude and composition of the load).? This is the Z-match now available in kit form from Pacific Antennas (the "BLT Plus") and QRPGuys (the "Multi Z tuner 40-10 meters).
?
I'd analyze it similarly to the wider range Z-Match described in my article--in effect an L-network, but with link-coupled output.? C1 provides the series input XC, with C2/L1 (the detuned tank circuit) providing L-network's parallel output XL.? (Other assessments/analyzes??? Effect of output link's size?? Coupling coefficient?)? One difference here is the tank circuit doesn't tune two frequency bands simultaneously.? Also, of you compare the two schematics, you'll also see that the output balance in the single-range model is probably better.? In both versions the output links are centered around the ground tap of the tank-circuit inductor, but in the dual-range version the full inductor itself is not centered around the ground tap.? If there were only inductive coupling between the tank-circuit inductor and the links, that would make no difference, but in reality there's also a bit of capacitive coupling.? ?As with most tuners, the common-mode impedance through the whole tuner circuit (coax input thru the output circuit) is quite variable.? ?In any case when I've checked the balanced line to the antenna for common-mode current (with various feedline and antenna configurations over the years), I get decent results.
?
This all takes me back in time.? To put numbers on these comments, I'd have to dig through some old files.
?
Charlie, W6JJZ
?
?
|
A balanced, ganged inductor tuner, with switched capacitors can achieve a match with two adjustments, at max efficiency. Simsmith can be used to calculate the system efficiency.
None of these canned design tools address network efficiency / losses.? Nor the fact that the L-network is the ONLY matching topology that yields a single matched adjustment of its two reactive components.? And, further, once that is established, it is operating at maximum efficiency.? No other matching network topology can make that claim.? That's why I use an L-network.? The tapped capacitor network is pretty good as well, but is not addressed.
Dave - W?LEV
Dave,
Maybe you want to have a look at this:
73, Robert W4BCZ
--
--
Please join me.
Dave - W?LEV
|
Charlie (W6JJZ),
Thank you for the interesting info and the pdf files re. the Z-Match.
By now I have collected and studied quite a number of articles on the
subject, which in IMO is far too complex for a quantitative analysis.
My goal would be to find the best (compromise) configuration of the
Z-Match for any point on the Smith-Chart.
I believe this is wishful thinking and there is no such thing on paper or in reality.
My appologies if I am wrong. The answer is - as we all know - an L.-Network.
Robert - W4BCZ
|
