The AM Forum

Examining the Johnson Matchbox ATU

The Johnson Matchbox comprises an unbalanced input section, link coupled to the inductor, thus providing an excellent conversion from the unbalanced-to-balanced balun function; two balanced L networks that perform the impedance matching function; a main tuning capacitor in parallel with the inductor, forming an LC tank circuit that is tuned to resonance at the operating frequency; a second capacitor having four sections, all on the same shaft. The two outer sections of the second capacitor are the output capacitors of the balanced L networks that provide the balanced output terminals. The two inner sections are simply two capacitors in series, connected in parallel across the output terminals, and thus are also connected in parallel with the input of the balanced feed line.

The Matchbox is an outstanding antenna tuner (ATU), and has an excellent decades-old reputation as a low-loss tuner.  

However, because the designer(s) didn’t fully understand the function of the inner two sections of the four-section capacitor, they misled the users concerning the function of those two sections. The instruction manual for the Matchbox tells us that the four-section capacitor is a voltage divider, and thus performs as an impedance divider, which is why they included it in the design. Big mistake, because it in no way performs as an impedance divider. The inner two sections are unnecessary, perform no useful function, and are in fact superfluous, and can be disconnected with no harmful results.

We’ll now examine why the inner sections perform no useful function, but actually detract from the impedance-matching function. We’ll also see why they would perform a useful function if they were on a shaft separate from the two outer sections, allowing adjustment independently from the output capacitors of the L networks.

Because the two inner sections of the capacitor are connected in parallel across the input terminals of the feed line, the output capacitors of the L networks never see the true input impedance of the feed line—they see only the line-input impedance as modified by the capacitance appearing in parallel with line-input impedance. The function of this circuitry is disturbing because the modified impedance the L network capacitors see is constantly changing during the tune-up procedure, because the inner sections of the capacitor are turning simultaneously with the turning of the L network capacitors. Thus, for every change of the L network capacitors during adjustment to obtain the match, the impedance we’re matching to is also changing. It’s similar to a monkey chasing his tail. A match is ultimately achieved, but with no help from the two inner sections of the capacitor.

We’ll now explain why the two inner sections of the capacitor could assist in the matching procedure if they were adjustable independently from the outer sections, the output capacitors of the L networks. We know that occasionally the feed-line length that reaches from the antenna to the tuner presents an input impedance that is outside the range of the tuner. This situation usually leads the operator to change the length of the line so that it does present an impedance within the range of the tuner. However, there is an alternative to changing the physical length of the line to modify the input impedance—changing the electrical length of the line by adding either series or parallel inductors (or capacitors) at the line input terminals. Whether to use capacitors or inductors depends on whether the line is too short or too long. The two inner sections of the four-section capacitor are already in parallel with input of the feed line, so if these sections of the capacitor were adjustable independently of the L network capacitors, some additional impedance range of the ATU would be accomplished by adjusting the inner capacitors to change the electrical length of the line.

I hope the comments above help in understanding the function of the Johnson Matchbox ATU.

Walt, W2DU

Wonderful!
I was waiting for that Walt, have been using the Matchboxes in the shack for years.
Thanks for taking that time to write that up.

Do you have any thoughts of the performance difference, if there is any, between the 275 and 1KW models?
Aside from the obvious power handling abilities.
Our friend "Kurt Sterba" claims the 1KW model is so much better on receive that you will hear things with that box
that you'll never here with its smaller 275 watt brother.
Wonder if the reciprocal is true, that it is a more efficient unit than the 275 watt box?
I've been using the 275 watt unit with my station, running 250 watts of AM with nary a problem, even runs legal limit on SSB. Just don't have the bench space to fit its 1KW brother.


Suggest that this be filed in the handbook section, great reference material.

-Bruce

Hi Bruce,

Glad you approve of my analysis!

Concerning the difference between the KW and 275W tuners, I don't know, because I've never had or used the KW version. I've had only the 275W version for years. However, Kurt is wrong concerning the KW version being better than the smaller version on receive--that's absurd, because there's no basis for that statement. The only reason Kurt could have had for making that statement is that there might have been a problem with his KW unit.

I don't know why the KW unit would be more efficient than the smaller unit, but on receive the efficiency would be irrelevant, because the signal-to-noise ratio would be the same at the output whatever the efficiency.

Walt

The KW version has been used for years (decades actually) as a low noise receiving tuner. Way too many have reported its superiority to dismiss as an anamoly. It is that viewpoint that is absurb.

Sounds like time for someone to analyze it. Ive no viewpoint as Ive not used one for receive, Beverages are in use here.

Carl
KM1H

Carl, can you explain why the KW version receives better than the 275W version. IMHO, that concept is what is absurd.

Walt

a german company called Anneke licensed the rights from EFJ and made their own version for a short time. it had a series condenser in line with the link coil where the EFJ version just grounds the far end. I have my KW version open right now to change the meter over to a expanded scale RF ammeter instead of the swr bridge which never worked. Put the SWR meter back the way it came in the slant front box. Works great.

I dont see how the small and large versions could be so different as to actually hear a signal difference on rx between the two. So many other factors would matter more that would cover up the difference  - like natural QRN on the low bands, rx noise figure and other internal rx noise on the high bands.

I vote balderdash on that claim.  since loss should be reciprocal on tx or rx, the difference, if any, should b easy to measure. But it sounds whacked to me.

We agree on the misconception that the KW version receives better than the 275W version, Derb. Kurt Sterba had a lot of good stuff in his 'Aerials' column in World Radio, but he sure was off in left field on that one. Incidentally, I edited the first published edition of Aerials for WR.

In addition to my having a 275W version of the Matchbox I built a tuner using BC-610 plug-in inductors. Consequently, I used the link-coupling method, which also provided the balun function. However, I did use a series cap in the link circuit, which canceled out the reactance in the link circuit. Without that cancellation it was impossible to reach a perfect 50 +j0 match.

I give you my best wishes for a speedy recovery from your new problem, Derb--should never have happened!

Walt

The Aerials books are at home, so I'll try read and paste Kurt's comments in this forum tonight, just to be certain that I am not speaking out of school.

Kurt's comment made an impression on me at the time, enough so I recall that I lashed up both units in a A/B test a few years back, didn't seem to discern any difference on my end, but I had not real instruments other than my ears. My little 275 watt unit fits under the shelf on my operating desk, and is about the same height at the 75A1, so I just use it.

I did press the KW unit into 160 meter duty last year, sits on the top shelf, with an outboard air variable clip leaded across the right tuning cap. Think it was based off Bowie Bill's ER article, only Bill used doorknob caps.
Sure it is not too efficient, at full strap the big indictor gets somewhat warm on old buzzard transmissions.

-Bruce


(http://www.booksforcomfort.com/images/20621.gif)

thank you Walt :D There is a new wrinkle on that on which I will make another post about.

This thread also reminds me of the problem with Lew McCoy's design that he called the "Ultimate" transmatch. As with the designer(s) of the Johnson Matchbox, McCoy
also didn't understand the use of the dual-section capacitor he placed in the otherwise standard T network. Murch and other mfgrs adopted the McCoy version, also not knowing that the second section of the Ultimate performed no useful function. The same as with the designer of the Johnson Matchbox, McCoy believed the dual-capacitor provided a voltage divider, and thus an impedance divider--WRONG!

In the 1970s I analyzed the McCoy Ultimate(?) version, only to discover the second section of the capacitor provided no useful function--is unnecessary and superfluous. I was then a consultant to Denny Had, the owner of Dentron. I told him about the uselessness of the second section, and he was delighted to learn that, because he was just about to order a new batch of the more-expensive dual capacitors for their tuners.

Shortly thereafter I visited the ARRL Lab, where I demonstrated with the Dentron Supertuner that the second section of the cap achieved no useful purpose. Doug Demaw and Laird Campbell witnessed the demonstration, and later DeMaw published the results in QST.

From then on Murch and other tuner mfgrs began using a single section input cap instead of the dual, thus reverting to the standard T network configuration.

Walt

PS--I posted this somewhat earlier today, but apparently it got lost somewhere.

Other then driving the cost of the tuner up, does this extra two sections have any ill effects?  I cant see any.  I own two KW matchboxes. I use one everyday.  I own 3 of the 275 watt versions.  I dont use them much anymore. I have heard for years the KW has better RX.  I cant imagine that being True.  But I have heard or read somewhere, that the large metal box makes a difference in the performance.  Something about the Q.  I am not an expert on this subject, I am just trying to remember what I read or heard over the air. 

Maybe I will Drag a 275 in, lay it ontop of the KW, And Test this out on WWV with the IF scope. Maybe I can determin if there is any truth to the statement.  I have a feeling that one person wrote this somewhere and hams picked up on it as fact. 

There was also discussion about having the RX bypass the KW matchbox.  EFJ gave provisions for that if you used the built in TR switch.  Some people had much better results NOT using the tuner on the RX.  This of course turned out to be some RXers being at 300 ohms input, Military at 98 ohms and some being at 50 ohms.  Depending on the atenna and the RXer, You might have been better off not going through the KW. 

A simple test is to tune the tuner flat for transmit.  Monitor a signal and retune for highest RX.  I was very suprised to see some of my RXers go up a good deal on Recieve when tuning OFF the Transmitter settings.  Some have an antenna TRIM or TUNE knob and I could bring this back with that setting. Some do not have that feature.  I ended up trying to make a matching balun for 50 to 98 ohms but gave up and ran out of time.

C

This is vy interesting.  Maybe EFJ used the four section cap to make folks think they couldn't homebrew their own tuner? hi hi.

That four section cap is hard to find isn't it.   Thanks Walt for this information.  I know the KW MB has faults but it is still by far the best manufactured bal. tuner I have ever used.  Didn't know about the inductor heating up though.  At least I have not seen any vswr creep during long transmissions like I saw with my old bal. tuner, the Bliss Matchmaster. 

You all who have tried the var. cap in series with the link, pse tell what v. and value you used.

It'd be interesting to see a side by side A/B with the 275 w. and KW; guess all you need besides the two tuners is a big knife dpdt switch and a coax switch.   

On mine, there's no rx relay, that all got taken out by some past owner.  I'm gg to go see if there is some higher noise level settting on rx other than what I have it set on for 1:1 tx vswr.

Well I went downstairs to the shack to power everything down for the night and found out in my case on 3558 KHz the low vswr setting was also the high rx noise (i.e. sensitivity) setting.

Rob

Good info and discussion folks. I'm going to poop the case on my Murch 200b to see if I have a two or four section cap in it. I T/R the tuner out of the receive circuit.

I think the Q of the match box stripped off some RX noise. Today most receivers have a broadband input unlike the days of tuned preselectors.

The Q of the Matchbox is unlikely to be high enough to have a bandwidth more narrow than the IF bandwidth of any receiver. So, unless the front-end of the RX is overloading, there will be no SNR improvement from the Matchbox.

Inductor Heating
Rob, the inductor heating which I observed was only on my KW Matchbox padded for 160 meters, seems to work fine
and run cool on 80 up. Suspect the heating is my problem from the clip leaded air variable I tagged on to make it play on 160, and it really only happens on continuos carrier old buzzard transmissions.

RX Matching
On the 30K station, I do run the 75A1 through the Matchbox for receive, does seem to help the signal level, suspect its more of a matter of impedance matching with the front end of the Collins receiver. I've read that you can change the tap point on the inductor for the rx input, have not changed mine, seems to work fine.

Some day, I have a some of "some days" scheduled, I hope to have that station, the 30K tx and 75A1 rx, all set to run on balanced line. Right now its lashed up with coax from the rx and tx to the Matchbox. I need a suitably old buzzard looking 120VAC relay to mount on the wall for T/R switched and get rid of the Dow-Key.

Sterba Quote
About the 275 vs KW Matchbox comparison. I didn't have a chance to find the Sterba reference last night, Nancy and I went out for dinner, never had kimchee before, surprisingly good. Should be a barn night tonight as I've got to finish up the Layette PB-46 RF deck. I'll post the exact wording just so I am certain that I did not misquote ole Kurt.

Good discussion and feedback!  :D

HUZ,
Both Jay and I have seen ther AD overload LED flash when there is high noise conditions. The HPSDR just has a low pass filter on the input set just above 6 meters. So I bet there are conditions where you could overload a first mixer with noise

No, as I dont have one, dont intend to get one, and have never used either. Just saying it is absurd without doing an evaluation is the real absurdity particularly since several well known DXers have stated differently.

Carl
KM1H

Which well-known DXers were these and what was the basis of their statements?

If your A/D is overloading from typical HF band noise (not T-storm strikes), you need to get a new A/D!

Steve, How do you like your KW matchbox and 275 matchbox?

C

I've had both in years past. All coax fed antennas at the moment.

You dont even own them?  Oh. 

C

They will flash under high noise conditions like storms in the area. when conditions are ok there is no problem. The A/D saturates around -10 dBM so that is a very strong signal.

What's your point?

I have owned them and also built my own link coupled and T-type antenna tuners. I never saw any improvement in SNR with any of the tuners when compared to a reference antenna with no tuner.

I've also used tunable preselectors with high peformance low noise preamplifiers ahead of receiving systems. Here the preselectors would often provide an SNR improvement by preventing overload by removing strong signals at frequencies other than the tuned frequency. These were 5-Section, Chebyshev, high-Q filters designed for this purpose. But even here, if the preamp was not overloaded, the preselector did not provide any SNR improvement. In this case the preselector was removed to avoid its insertion loss.

I thought you could run a test for us. But if you dont own them, I guess you cant.

C

<<<RX Matching
On the 30K station, I do run the 75A1 through the Matchbox for receive, does seem to help the signal level, suspect its more of a matter of impedance matching with the front end of the Collins receiver. I've read that you can change the tap point on the inductor for the rx input, have not changed mine, seems to work fine.>>>

I think the stock matchbox had two taps on the link, one for tx and one in another point a few turns away for rx which was of course switched in by the stock TR relay.   Many (including mine) were modded at some point with the relay and rx tap removed.    I guess that rx tap was for older rx like the A1 that had 300 or 600 ohm antenna terminals on the back.  There was a KW MB on eBay a few weeks ago that still had the TR relay so a few survive I guess.

Rob 

Yes. Exactly.  Mine, Is straight through.  If you use the TR, on RX, It will match to the older 300 ohm recievers. This is why some are missing performance on RX using 50 ohm recievers.

C

I have the original 275 matchbox owned at one time by W6PIG.

It's been modified so that the TR switch is still in service, but the receiver tap was moved for 50 ohms.

It has the instructions on how to move it BACK to 300 ohms written in pencil.

I can pop it open and count turns if anyone wants one designed to have 50 ohms on the receiver tap.


--Shane

That might be good info to have posted.  You can use an MFJ to figure it out. Recently, I have been experimenting with matching the Rxer to the antenna. It makes a big difference for weak signal. But I fully realize why most people dont bother. I normaly talk AM when the band is in. If I guy is showing 40 over to me, It wont make any difference in the copy.

Good to see you posting shane.

C

Written in various old paper DX and contest journals back when 5BDXCC was announced. Back then 80M DX was a rarity and requirements little known. Then a 160M DXCC came along.

WOW I thought I was the only one who fell into that trap!!

The Bliss tuner did pass muster for the lowest loss, but when components started to fail as that POS coil did, the manufacturer-owner said there were no replacement parts. And talk about a real POS to get into the DAM thing....geesh.
It's in pieces and I built the K1JJ SuperTuner.
Nice info on this thread BTW.
Fred

We had one of those Bliss tuners. The big one.  I think my elbow still hurts sometimes from turning that damn Crank on the side.  It was sold off.

C

my KW match box produced exactly the same feeder current as the much larger HB fugly