Random questions

[W8ACR]

A lazy day pondering the mysteries of radio physics has generated some random questions:

1. If a big push pull final amplifier tube, say a 250TH, suddenly craps out, what is likely to happen to the remaining tube? Disaster or no? Will the transmitter continue to emit RF at the tuned frequency?

2. What specific applications are better suited to low mu tubes as opposed to high mu tubes? It seems to me that high mu tubes are much more common in both RF and audio circuits. For example, 810's are more common in all circuits than are 8000's. 811's are more common than 812's, and 100TH's are more common than 100TL's. Is there a good reason for this?

3. I am embarrassed to ask this one, but this has always been a mystery to me. Consider a conventional final tank circuit with a plug in coil and center link to the antenna. It is perfectly acceptable to attach coaxial cable to this center link to feed either a tuner or antenna directly. The coaxial shield is, of course, attached directly to the transmitter chassis, which is usually attached directly to DC ground. In this case, why is it that the transmitter chassis is not hot with RF energy, and why is it that half of the RF energy does not disappear into the ground?

I await the wisdom of the sages.

Ron

[KE6DF]

2. What specific applications are better suited to low mu tubes as opposed to high mu tubes? It seems to me that high mu tubes are much more common in both RF and audio circuits. For example, 810's are more common in all circuits than are 8000's. 811's are more common than 812's, and 100TH's are more common than 100TL's. Is there a good reason for this?

I was waiting for answers to your questions also.

I'll take a stab at this one:

Low mu tubes take less drive in RF circuits.

Hi mu tubes make better GG linears.

Some low my tubes can generate lots of power output in AB1.

Factors like the availability of multiple sources and the desire to have minimum number of different tube types in a product may also dictate using same tubes in multiple places.

I suspect that a broadcast TX designer would lean towards 810sx810s rather than 810sx8000s just based on reducing spare tube types.

811's also good as zero bias modulators -- an obvious advantage. Plus Hi mu types like 811 work better in grounded grid. So you end up with 811s being made in higher volume and lower prices than 812s.

Dave

[AB2EZ]

Ron

Regarding question 3:

Although non-intuitive (even graduate EE students would have problems answering this question)... power flows through a coaxial cable in the space between the outside of the inner conductor and the inside of the outer conductor (shield). Both the electric field between the conductors, and the magnetic field around the outside of the inner conductor (but not around the outside of the outer conductor) contain stored energy. With a load on the cable, power (energy per unit time) flows through the cable from the source to the load.

There is current flowing through the inner conductor, and an equal and opposite current flowing on the inner part of the outer conductor. If the load (antenna) is balanced, there will be no (common mode) current flowing through the outer part of the outer conductor, or through the chassis or grounding conductors.

If the load (antenna) is not balanced, some current may flow down the outer portion of the outer conductor. Likewise, if the load is not balanced, rf currents may flow through the various "ground" conductors. In that case, there may be large amounts of rf in the shack (in addition to the rf being radiated by the antenna); and objects in the shack (like the chassis of the transmitter) may be "hot" relative to other objects in the shack. This effect will be larger at higher radio frequencies, because it is differences in position... relative to a wavelength... of objects in the shack, that results in rf potential differences between these objects... caused by "ground currents".

Stu

[K5UJ]

The coaxial shield is, of course, attached directly to the transmitter chassis, which is usually attached directly to DC ground. In this case, why is it that the transmitter chassis is not hot with RF energy, and why is it that half of the RF energy does not disappear into the ground?

I don't think that's right, which may be some of the source of your confusion.  If you are asking about a swing link for coupling power from the RF final to the feedline, the unbalanced coaxial line has the inner conductor connected to one end of the link and the outer conductor to the opposite end of the link so it all floats.

Take  a look at the schematic of the Eldico TR1 here:

http://home.fnal.gov/~atkinson/TR1/TR-1schematic-enhanced.jpg

the swing link output is in the upper right corner.   It can be either connected to balanced feed or coax as I understand it.
Meissner Signal Shifter manuals also have a number of schematics for link coupling arrangements with twisted pair as to many old handbooks from the 1930s and maybe 1940s pre-coax.

OTOH, the link on a Johnson Matchbox has one side grounded to the cabinet and I guess when adjusted so the primary link and secondary coil match to give a 50 J0 view to the coax (characteristic Z 50 ohms) all of the power in the link is transferred to the coil so none appears on the cabinet.


Rob

[k4kyv]

A lazy day pondering the mysteries of radio physics has generated some random questions:

1. If a big push pull final amplifier tube, say a 250TH, suddenly craps out, what is likely to happen to the remaining tube? Disaster or no? Will the transmitter continue to emit RF at the tuned frequency?

It is possible that you won't see much difference and plate current and rf output.  Of course, the other tube will be severely overloaded, since it's taking on the full load of both tubes, so I would immediately shut it down.  Or you may see some reduction in plate current and rf output, but not nearly down to half.

2. What specific applications are better suited to low mu tubes as opposed to high mu tubes? It seems to me that high mu tubes are much more common in both RF and audio circuits. For example, 810's are more common in all circuits than are 8000's. 811's are more common than 812's, and 100TH's are more common than 100TL's. Is there a good reason for this?

For class B and class C applications, the lower mu tube is usually preferred for the class C final, and the higher mu one for the class B modulator.  The lower mu triode tends to take less drive for full class C operation.  The higher mu triode requires less bias and less peak-to-peak driving voltage for a class B modulator than would a low mu triode. For example, the 8000 takes only about 75% the driving power as a plate modulated class C amplifier as does the 810.  RCA used to run ads recommending a pair of 203As modulating a pair of 211s. Also popular was a pair of 812(A)s in the final modulated by a pair of 811(A)s  But the tube charts give ratings for each one of most of these pairs of tube types for both class B and class C service, and usually their efficiency and maximum power output are about the same.

3. I am embarrassed to ask this one, but this has always been a mystery to me. Consider a conventional final tank circuit with a plug in coil and center link to the antenna. It is perfectly acceptable to attach coaxial cable to this center link to feed either a tuner or antenna directly. The coaxial shield is, of course, attached directly to the transmitter chassis, which is usually attached directly to DC ground. In this case, why is it that the transmitter chassis is not hot with RF energy, and why is it that half of the RF energy does not disappear into the ground?

You don't have to ground one side of the link.  In fact, you will be less likely to have common mode currents on the feedline at certain lengths if it is simply left floating.  Grounding it provides a ground plane for the outer braid to work against as a Marconi antenna.  If it is floating, there is less likelihood of the final tank arcing over to the link if the final tank is series fed so that the main coil has full modulated +HV on it.  But if everything is working properly, grounding one side shouldn't make a lot of difference, since if there is no common mode current (sometimes called "antenna current") on the feed line, there is nothing to be carried to ground.

Think of coax as a 3-conductor cable: the inner conductor, the inner surface of the braid, and the outer surface of the braid.  The current on the inner conductor should be exactly equal and opposite to that of the inner braid, while the current on the outer surface of the braid should be zero. Any outer surface current is undesirable, the result of common mode current, and this is what causes the feedline to radiate.

I used to run my link-coupled tuner about 3 ft. away from the the link-coupled transmitter.  The transmitter had a swinging link and the tuner, a fixed link.  I coupled the two links together with heavy duty zip cord, with the two conductors pulled apart and separated.  Neither side was grounded. I tuned by dipping the final, loosely coupling the variable link, tuning the tuner for a peak in plate current, then adjusted the link for rated plate current.  I would then re-check dip and repeat the operation if it was very far off. I also left the main coil floating from ground, since that could also provide a circuit for common mode currents.

But at some feedline/antenna leg lengths, when the total is an even number of quarter waves, common mode currents may develop as it tries to act as a Hertz antenna.  In that case, it may be desirable to ground  the mid-tap of the main coil.

The best thing to do is to try it both ways, link floating and link grounded, and see which way works better.

[WA1GFZ]

I would think coax connected to a swinging link without the shield grounded would allow RF currents to flow on both the inner and outer surface of the coax. Also you will have an RF voltage on the outer shield. The shield will radiate.
Grounding the shield at the link will force zero volts on the outer surface of the shield. All voltage will be between the center conductor and inner surface of the shield because the link is driving the center conductor against the grounded shield.

[W7TFO]

Properly done twisted pair will not radiate, nor will it be susceptible to interference if it is used for receiving or audio transfer.

The key to that is using it in a truly balanced configuration.  If there is any unequal preference to ground in either source or load, it will upset and not work properly.

Some balanced systems us a centertapped arrangement, with the CT grounded.

If one side is forced to ground, it is now a single-ended transfer circuit.

73DG

[KM1H]

1. If a big push pull final amplifier tube, say a 250TH, suddenly craps out, what is likely to happen to the remaining tube? Disaster or no? Will the transmitter continue to emit RF at the tuned frequency?

Depending upon the type of "Crap out" the remaining tube will be from severely to highly overloaded but still be producing RF since its half of the tuned circuits are still functioning. Reduce drive to a safe level and run Ip as you would a single tube.

2. What specific applications are better suited to low mu tubes as opposed to high mu tubes? It seems to me that high mu tubes are much more common in both RF and audio circuits. For example, 810's are more common in all circuits than are 8000's. 811's are more common than 812's, and 100TH's are more common than 100TL's. Is there a good reason for this?

As others have mentioned, however there is no hard rule. Often the choice was due to price and availability and high or medium mu tubes such as the 811/811A, 810, 100/250TH, etc were more readily available as WW2 surplus and are still fairly plentiful. OTOH the 304TL had the same attraction to hams.

Hi mu makes much better GG linears since it requires much less drive due to zero or very little grid bias. While a 304TL will barely make 6dB gain in GG and require several hundred volts of regulated bias a 3-500Z with a 200 mu is good for close to 12dB. Most all the popular ham use ceramic triodes from the 8874 to YC-156/3CPX5000A7 are high mu with some having gains close to a tetrode.  OTOH the 304TL makes a lousy Class B modulator but is one of the very few triodes that can run PP AB1 driven by a 6SN7 to 700+ W and will put out gobs of reasonably clean audio in Class AB2 at low drive.

3. I am embarrassed to ask this one, but this has always been a mystery to me. Consider a conventional final tank circuit with a plug in coil and center link to the antenna. It is perfectly acceptable to attach coaxial cable to this center link to feed either a tuner or antenna directly. The coaxial shield is, of course, attached directly to the transmitter chassis, which is usually attached directly to DC ground. In this case, why is it that the transmitter chassis is not hot with RF energy, and why is it that half of the RF energy does not disappear into the ground?

Suffice it to say that floating the coax shield at the amp is a sure fire way to get RF burns on your lips in most installations as the RF will be picked up by the house wiring and most metal surfaces. Im one of those that believes its near impossible to have a balanced line inside a house and almost as hard outside. Grounding the shield to the chassis is going to require that the chassis have a good RF ground. There are really 3 RF currents on a coax, 2 inside (center conductor and inside of the shield) and the 3rd on the outside of the shield. There is also some small leakage between shield currents with single shields such as RG-213. It gets worse with the economy cable offerings and is lowest with the LMR-400, and RG-214 double shielded variety. Others use 50 and 75 Ohm hardline as its become very affordable on the surplus market. Something with 3/8" Superflex in the shack and 1/2" or 7/8" Heliax outside with ferrites will be as clean as you can get with a perfectly balanced unbalanced feed.
You want to minimize the outer shield currents at all times and modern ferrites are very good at this whether feeding a tuner or coax directly to the antenna where more ferrites should be used. The ancients didnt know about them and those who still operate that way likely refuse to use them so suffer from RF in the shack and noise pickup from all the electronics in the house and from neighbors from severely unbalanced open wire line and coax.

Carl

[k4kyv]

1. If a big push pull final amplifier tube, say a 250TH, suddenly craps out, what is likely to happen to the remaining tube? Disaster or no? Will the transmitter continue to emit RF at the tuned frequency?

Depending upon the type of "Crap out" the remaining tube will be from severely to highly overloaded but still be producing RF since its half of the tuned circuits are still functioning. Reduce drive to a safe level and run Ip as you would a single tube.

I would reduce the grid drive to a little over half the normal value (or run as much drive as the tube will safely handle), and drop the loading back so that the plate  current reads half the normal value.  If each tube in the final stage uses its own grid leak resistor, then the drive  could be dropped to half value.  If they share a common grid resistor, dropping the grid drive back to half also drops the grid bias proportionately, so the tube could be operating with correct grid drive current, but insufficient grid bias for good linear plate modulated class-C service. Of course, this is ignoring any fixed bias supplied to the tube.

I have never had rf in the shack using a floating link.  The only time I ever had that problem was with my first serious transmitter, a pair of 807s modulated by another pair, which used a pi-network output, which I fed directly to dipoles which were fed with 72-ohm twin lead.  When I increased power to about 600 watts, the twin lead  crapped out, couldn't withstand the voltage across the minuscule gap between the wires.  I replaced the spider web of separate dipoles with a single 80m dipole, fed with open wire line and a tuner for multi-band use, and that cured the rf problem, plus the house looked a lot better without all those dipoles strung everywhere.

When I moved my  station to the present outdoor shack, I had a problem on 40m.  No "hot" mics, but enough rf was getting into the audio to cause distortion and feedback on 40m.   I finally cured that problem by winding a simple current balun with a piece of coax.  It's about 8 turns wound about 6" in diameter as an air core coil.  I also tacked about a half-dozen quarter-wave radials under the shack, to the underneaths side of the floor joists, and ran the common point up through the floor right to the transmitter.  I can put the transmitter on the air and  crawl under the shack, and see that the ends of those radials are hot enough with rf to light a neon lamp.  The 8' or 10' run to the outside ground rod was too long to establish a good low impedance ground at the transmitter on 40m.