Examining the Johnson Matchbox ATU

<< < (2/5) > >>

k4kyv:
IIRC, capacitor C2 is built so that the inner capacitors are fully meshed when the outer ones are fully unmeshed and vice versa. Each half of C2 serves as a differential capacitor that the manufacturer intended to be a variable capacitive voltage divider, serving the same function as tapping the load down on the coil, while the primary resonance is determined by C1.

I tried tapping down on the coil of a balanced link-coupled tuner in order to feed my 80m dipole as a short dipole for 160.  The flat top is about 135', and the feed line is close to the same length.  The line-input impedance on 80m is low and mostly resistive, so the 80m tuner uses series feed formed by splitting the coil at the mid point and attaching the feeders across the gap.  But on 160, the feed point is highly reactive, since each leg (feed line + 1 leg of the dipole) is 3/8λ, placing the line input point exactly midway between a voltage loop and a current loop.  I threw together a tuner using a BC-610 plate  capacitor, 150/150pf @ 7 kv, and a link coupled coil.  I tapped the feed line  down on the coil until I  could get close to a 1:1 match between the link and the coax line back to the transmitter. When I tried modulating the rig, I couldn't run over about 100 watts before the tuning cap would arc over.  I tried  several LC combinations with other capacitors and adjusting  the number of turns of the coil, always with the same result.  I finally just added an extra 60 ft. section of OWL that I automatically switch in with that tuner, and it has worked perfectly for nearly 30 years, arcing over only when bugs or trash get between the plates of the tuning cap.  I suppose I could have added either a series capacitor or inductor to each side of the OWL making it appear to the tuner as purely voltage or current feed, but I decided the extra section of OWL would be simpler and there would be no series variable capacitors or inductors to adjust when moving from one end of the band to the other.

Regarding the Johnson Matchbox, as I see it, wouldn't the coil and C1 form the balanced L network, and the combination of C2 inner and outer caps merely alter the feed point impedance of the OWL?  Actually, instead of simply removing the inner caps, it would appear to me that re-wiring the inner caps so they are disconnected from ground and connected in parallel with the outer ones, and the rotor plates re-positioned so that both sections would mesh unmesh together, would give the tuner a greater range, since you would now have an increased range of variable capacitance in series with the feed line.  When C2 is rotated so that the outer section has maximum capacitance while the inner section has minimum, the inner caps are in effect nearly disconnected from the circuit. Placing those sections in parallel with the outer sections would allow a higher series capacitance, something you would want if the feed point of the OWL were at a voltage loop with the tuner looking into a high-Z with minimal  reactance.

Also, it looks to me like using the stock tuner to feed a low impedance, such as the case of feeding the OWL at a current loop would be inefficient, since you would have a large circulating current between the coil and C1, while simultaneously trying to force-feed a high rf current to the OWL through a small capacitance at the outer section of C2 (since, to form the capacitive voltage divider that gives lower voltage/higher current to the OWL, the differential cap  would have to be adjusted towards minimum capacitance at the outer section/maximum capacitance at the inner section).  Add to that, any reactance,  capacitive or inductive, appearing at the feed point would drastically alter the match, probably unpredictably.

Frankly, I don't see how the thing works at all when attempting to feed anywhere near a current loop.

K5UJ:
<<<
Frankly, I don't see how the thing works at all when attempting to feed anywhere near a current loop.>>>

Right.  I think I slightly understand you Don; (most of it over my head) but I know the Matchbox wasn't intended to match low Z loads where there's a lot of current like very small << 1/2 w. dipoles.   this is something Matchbox bashing hammies whine about--the crowd who want an ATU to handle every possible Z on every frequency both balanced and unbalanced. 

Let's see...I have to think about what you posted Walt, thanks.   Also, I did not know the inner sections were 180 degrees opposite the outer sections on C2.  Yes, putting inner and outer in parallel is an interesting idea.   I'm trying to imagine the reasoning for the stock MB scheme.   For the first time I began to think that the cap arrows opposing each other in direction on the schematic might mean something (duh).   but if they do I'd think the outers would point in the same direction and ditto for the inners but opposite.  And C1's would also point in the same direction.

If you are right about C2 Don, what if we think of the inner sections of C2 as being parallel with C1 (maybe you are already doing that, what do I know) so if we run out of C1, can C2's inner sections augment it.  by increasing the inner sections of C2 the series capacitance of C2's outer sections would proportionately decrease but would the overall effect of C1 increase?  What does that mean/do?   to illustrate by going to the logical conclusion, imagine just one side for simplicity and C1 is maximized.  between its hot side and the feedline you have a C2 section in series.  If it is minimum it is almost out of the circuit and the other C2 section (the inner one) is maximized and parallel with C1 to ground.  C1's capacitance is augmented?   As we decrease C1 by lowering the value of the C2 inner section the series side of C2 goes up.  I'm not an engineer so I don't know why this would be desirable (but mine works so I'm not complaining hi hi).   I don't know if I contributed anything or not, but this sure beats cleaning out the basement  ;D

Rob

Ralph W3GL:


   Don,

   I'm here to tell you, it don't match low Z, high I loads!  

   In that case you must add a section of OWL to bring the feed Z to a
   reasonable value or (like I do) use a piece of coil stock as a balanced
   auto xformer... This effectively increases the matching range of the unit.

   Another weak point on the 275 Watt box is the fiber coupler between
   the sections of the differential caps.  If one hits the box with over a
   couple watts of power in any unmatched settings of the tune/match
   controls it WILL flash over, causing a carbon path making the box useless
   till repairs are made...  Ask me how I know!  

   The big boy has an isolantite coupler in there so it isn't prone to this problem.

k4kyv:
Looking at the Johnson schematic, I see the link is grounded at one end, and what looks like two taps and an unused portion of the coil, but it doesn't show a switch to select the taps or to connect anything to the top end of the coil.  Also, there is no variable cap in series with the link.

How do you change the taps on the link coil?  With no resonating capacitor, it would seem to me that you would have to change the number of turns on the link every time you changed bands if it is fed directly with coax, and the coax is to work into a matched load.  Also, it shows  the receiver connected to a different tap on the link from the transmitter.  Is that always the case?

I have found that to feed coax directly into an untuned link and achieve a 1:1 SWR, the coupling has to be extremely tight, and the number of turns is critical.

On my link coupled tuners, I use a series resonating capacitor, and it takes just the right combination of main tuning cap and link resonating cap adjustments to achieve a perfect match to a 50Ω coax line.  If the adjustment is made at the middle of a band, I can usually cover the entire band by simply rotating the main tuning cap without further adjustment of the link resonating cap.  The SWR may creep up to something like 1.3:1 at the extreme band edges, but it's not worth the extra step in the tuning procedure to keep adjusting it back down to a flat 1:1 all across the band, so I just leave it as is.

K5UJ:
The tap on the link that gives more turns is the one for a receiver.  The TR relay switches between the taps.  The idea was that originally an old rx would have 300 ohm antenna input terminal strip on the back instead of a UHF jack for 50 ohms.  I think you'd connect the rx twin lead to the back of the Matchbox at a terminal barrier strip at the bottom of the rear of the cabinet.  This is the stock wiring.   A lot of guys have removed the TR relay and rx tap and all that extra stuff and just run the tx line from the UHF jack on the back of the MB right to the link tap and use that for tx and rx.

I do that and run mine with that one tx tap on the link, no cap in series to ground or any other link taps mine seems to work okay, but I had to add about 10 feet of ladder line to get it to match the low band dipole.  to get the high band dipole to work with it I had to make it a full 1/2 wave on 20 m.    Maybe that means its range is limited but it is hard for me to tell because my situation is complicated by antennas interacting with each other.  If I can't get a match I find that futzing with another antenna like grounding the verticals makes the difference. 

I don't really know how much adjusting I have to do at N kc away from where it is tuned because when I was QRV I pretty much parked on one frequency for the evening and did not move so it was not something I bothered to notice.

rob

Navigation

[0] Message Index

[#] Next page

[*] Previous page