The Triton IV needed a Checkup

Figure 1: The old S meter on the Triton 4 is registering signals above zero again.

The Triton IV is a radio I’ve had close to my inner ham’s happy thoughts for a long while. It’s antiquated by today’s standards, having been manufactured first in 1976. It’s almost 100 percent analog, unless you count a few old gate chips on the control and calibrator boards. Its 70s era-styled case is infamous for being the flimsiest ham radio chassis enclosure ever built. For this reason, it has always been advisable to place it on a stout table, keeping the brass pounding and attendant vibration damped down to a suitably less energetic level. But, the receiver is pretty sharp, has no hiss or phase noise, and the sensitivity has always managed to be plenty enough for my rag chewing QSOs.

It’s that sensitivity that plummeted one day, while I was working twenty meters. Suddenly, I could hear the other station barely. More alarmingly, the noise had also disappeared. It was as if the antenna disconnected, and I was picking up the little bit of signal that could penetrate the circuits directly through the case. To check this, I sent out a signal and watched the SWR, letting the other ham know I was going to have to QRT.  The transmitter was OK, and the SWR was perfect.  That mean’t that … Oh Oh.

Note: Quite a while ago, I had anticipated an eventual “Oh No!” moment. Accordingly, I located a half dozen or so NOS 40823 MOSFETs for the junk drawer.  So, the dread that might normally accompany the last paragraph was diminished somewhat. HiHi.

It immediately dawned on me that the RF amp’s MOSFET input (front end) transistor had given up the ghost. So, I took the Triton apart to do a replacement, feeling that was something I most certainly would have to do. Checking the voltage on the calibrator wire coming out of the preselector partition seemed to show that gate #2 on the MOSFET was operating correctly. The drain, source and first gate terminals of the MOSFET still needed to be checked. This was no easy chore.

I am one to compliment Ten Tec on almost every aspect of their builds, but the preselector stack on the Triton is definitely a PITA. The resonator knob is attached to an elevating core assembly that moves the ferrite cores of L1 and L2 up and down on the RF amp board. This is a bit nifty, as an idea. It works well. But, the RF board is nested below the elevator assembly, and is screwed into the chassis from the other side of the board. So, you have to pull the board out from the bottom AFAIK. I suppose you might work a screwdriver in between the partition walls somehow, to loosen the screws, and then take it out from above. Before any of that can happen, you have to desolder about a dozen wires and component leads. This was a long, messy job just to get a couple measurements on the MOSFET.

Run across a dental pic or a dental mirror at a flea market? Buy it!

So, I put my tiny dental mirror to work, hoping that with the mirror, I could successfully land a probe onto the PC board foil for the other gate, the drain, and source, without blowing anything up.  BTW, if you ever run across a dental mirror at a flea market or antique store, definitely buy it! It was worth the $5 already, and I got a nice pic along with it.

Well, getting the flashlight to line up with the mirror while working the probes was seriously challenging for an old guy with only two hands.  I don’t think I landed the probe properly, but in the process of pushing and pulling things, I miraculously solved the problem.  Suddenly, the S meter popped up to 20 db over nine.  W1AW/4 was calling “CQ contest”.  What a beautiful signal that was to hear!  Of course, this meant that I had a cold solder joint.  After forty seven years of life it for some reason decided to lift its head.  It just goes to show ya! Sometimes it’s easier than it looks.

In retrospect, it might be that the resonator knob and attendant elevator assembly put a little mechanical push or pull on the RF board as it “elevated” or “de-elevated” – and this eventually pushed the cold solder joint open.  Or not.  So, I made sure that the cores were centered in the shafts, so that further mechanical stress could not happen, and heated some of the terminals.  Don’t know if I heated the right one(s), but at least I know where to look if the problem ever returns.


A Shortwire on 80 meters.


Figure 1: The trifilar wound “unun” in a weatherproof minibox.

I had an assortment of antennas, but on 160 and 80 meters I had nothing of consequence.  I have an 80m horizontal loop, but stopped using it due to its proximity to my building structure and attendant problems. I had been looking at VK6YSF’s longwire web page, and the little 23 meter (75.5 foot) end-fed “long wire” antenna that he had formulated was quite interesting to me. He had taken the time to do a complete analysis of the impedance of a variety of not-quite random wires, assessing the impedance on each ham band. He settled on 23 meters as a good fit for almost all the bands (except 17 meters). I already have a nice loop for 17m, so didn’t feel hobbled by the antenna’s high impedance on that band.

He uses a 9:1 unun to feed the end-fed longwire (it’s really only a longwire at 20m and above). Probably one should slap a 1:1 current balun in behind the unun to keep coaxial radiation at bay, in my non-expert opinion.  Balun Designs sells a nice hybrid unit that is both a 4:1 unun and a 1:1 current balun in one box.  I ordered one for a 4:1 loop connection, but I don’t think they have a 9:1 combo version, which would simplify my longwire antenna arrangement.   I procured a nice 9:1 unun with a 2 inch core (1.5 KW capable) – and tried a few of the bands.  I never bother with the little cores because the big ones will work with QRP as well as QRO. For end fed wires, it’s always only a QRP (max of a couple watts) power level that I use.  I could easily load power from the Omni A into the antenna using, in addition to the outside transformer, a tuner in the shack. The impedance transforming unun is outside in a minibox, and a run of RG213 coax feeds the tuner inside the shack. 

Some hams think that they can slap a 9:1 voltage balun onto the wire, and all will be good.  But, such a device does nothing to isolate the antenna to prevent mantle radiation or the feeding of RF into your shack.  A good quality 1:1 current balun seems to be in order for that task.  I am a big fan of using transformer baluns instead of chokes.  Some baluns are built with short coax sections that are covered with concentric ferrite beads, but IMO these may not isolate the antenna in all cases.  The “Balun Designs” page gives info on transformer balun versus choke considerations ( “Striking a Balance.pdf”).  You can search for the pdf on their site.  It seems to me that they are expert in this arena.  They successfully pushed me in the direction of their own product, LOL.

Note I’m not giving advice for how to build your antenna here, but am just writing about what I did with mine.  Incorrect (or for that matter even correct) usage of end fed wires, baluns, ununs, and counterpoises can create high voltage nodes on your wires.  BEWARE of end fed wires and other such things, and don’t say I told you to do it any particular way!  I’m not a qualified balun/unun/counterpoise design expert. For my own preferences, I would never run much power to an end-fed radiator such as what is described here.   QRP power levels of one or two or five watts are in the domain of what I run.  

There! Disclaimer entered.

I found the performance on 40, 30, 20, and 15 meters to be mediocre, which probably is the expectation beforehand.  On twenty meters it may be a bit better than my dipole or my vertical in its favored direction.  The graphs on the VK6YSF page show wildly varying radiation patterns, depending on the band in use. On 17m, my tuner could not do the job.  I didn’t try 10 or 12 meters. Just for grins I decided to load the antenna on 80 meters.  On the low end of the 80m band, the wire is about a quarter wavelength.  On the other end of the band it is approximately a 1/3 wavelength radiator.  I had not harbored great expectations for the use of it on 80m, but after a few days I’ve decided it ain’t all that bad! 

Granted, I currently have nothing else on the band, so cannot make any valid comparisons, but in the first few days many CQs seemed to garner a 579 or 589 or 599 report, so long as the station was within maybe five or six hundred miles.  Conditions may have been above normal, but maybe it is (at least for now) my 80 meter antenna, and I’ll let the vertical and the loops handle the other bands.  When it’s all ya got, of course it’s viable, if it works at all,  depending upon your patience level. It seems to do pretty well on the 20m band.  On that band, it’s a true longwire (equal to or greater than one wavelength). Due to the up and down complications of band conditions, this is all very much a horse shoes swag of course.

The VK6YSF page is at

Figure 9 on his page is the important one – and shows the tune-able impedance on 80 meters and the other bands.  This antenna should eventually be upgraded to something more traditional, like a high dipole on 80m, but for now it’s easy to run out a length of 75 feet of wire, to get a signal of sorts onto that band. Due to my lot, the full length 80m antenna will be much more labor intensive.

I went so far as to try this wire on 160 meters, and made a QSO there with difficulty, three states away (Hi there W8BJO!).  On 160m it is only a 1/8 wave horizontal wire, so expectations are considerably lowered.  The efficiency may be a bit better due to the counterpoise I use in preference over an earth ground. I DO have an earth ground on the coax feed for the equipment!   

Addendum 02/25/23:

I played around with the counterpoise length quite a bit.  Its length seems to make a substantial difference, depending upon on the frequency of operation.  Looking at the info at – you can get an idea about counterpoise lengths.  For a half wave antenna, end fed, he shows that the impedance starts to climb quickly as the counterpoise reaches about 80 percent of the antenna length, ultimately going to about 5,000 ohms.

Figure 3 on the aa5tb page shows the impedance chart vs counterpoise length.  I still haven’t settled on a good length for the counterpoise, because it is a multiband antenna and the counterpoise is subject to a percentage of the vagaries associated with the antenna wire itself.  So, I’ll continue to experiment, and maybe eventually I’ll settle on the perfect counterpoise length for a 23m longwire LOL.

Again, none of this is design advice or a recommendation!  73, – Ron