Contacter l'auteur / Contact the author

Recherche dans ce site / Search in this site




Technical review

The Kenwood TL-922 amplifier, 850 W output.

How to select a tube HF amplifier ? (I)

Imagine that all your ham shack is invaded with rigs and accessories, that your antenna, beam, vertical or wire is performing, proved by your hundreds QSLs received from far DX stations. But sometimes, during pileups or when propagation is closed you have difficulties to work some remote stations or to get good reports exceeding 53-55. On your side you hear them all, not always loudly, but at least 53 or with a strong intelligibility or QRK.

The need and the fashion

If you really experiment problems in working far stations or pileups although you are using efficient gears but at first sight not so performing as expected, in this case the last thing to install beside your 100 W PEP barefoot transceiver is well a good amplifier, up to your class limit, 400 W, 1 kW or more, depending on your national regulation. Good news, you will tell me ! Indeed, but at the condition to use it properly. What we are going to review.

In all other circumstances, if your RTX is not equipped with performing filters to reduce QRM, if your antenna yield a low gain or is omnidirectional or erected very low above the ground, improve first the efficiency of your equipment instead of looking for the last novelty. For the same price or even cheaper, it will be more useful to improve performances or your transceiver and antenna than using a linear amplifier. In all cases adding an amplifier to your current installation will probably not help you much in working far DX stations that arrive 53 or not at all to you. 

But I have never said either that an amplifier might replace a defective antenna. Far from me this idea. Do never use a high power to compense the poor performances of an antenna. This is indeed the best way to create QRM or get serious troubles if your installation (cabling system, SWR-meter, power supply, etc) is unable to sustain such power, thus high currents. Build or purchase first a better antenna !

Improve first your antenna system in reducing the ground effect, use a low loss coaxial (RG-213, Belden RF-9913, Aircom, etc), try to get an excellent VSWR on the line, as much features that, once optimized, will increase the efficiency of your installation. Then add some filters (Collins, etc), change maybe your transceiver for a more performing model and at last, after one year of experimentation, see if an amplifier is really necessary. 

This is a wiser advice than using an amplifier at all costs and be unable to use it properly or to hear your correspondent. The impertinent reader will say that "for lack of a big gun you need a big mouth !" Mm... Indeed. Without more know-how, answer first to the question : for what usage do I want imperatively an amplifier ? Isn't there a less expensive and better solution ? Then, be aware that a bad manipulation can quickly burn your P.A or produce dangerous side effects. Here are some examples.

If your transceiver last stage is badly matching your amplifier input specifications, you have all chance to burn your P.A. The Kenwood TL-922 amplifier for example must be driven with an input power ranging from 80 to 120 W PEP, no less no more. But if you want to preserve the lifespan of both your transceiver last stage and amplifier, drive this latter with 80 W in. Optionally reduce a bit the load to get less output power (with sometimes a higher SWR). 

Then, in tuning not correctly your amplifier you will make QRM over 10 kHz or more around your frequency, without working more stations but in disturbing a lot of amateurs. 

If you place your antenna system too close to your shack, say in the first 15m away (50'), using 1 kW PEP you are to the edge of the security distance requested when using HF systems emitting a strong electromagnetic field (e.g.. a beam offering a gain of 8 dBd, transmitting 5 kW EIRP generates a max field intensity of ~27 V/m).

The solution is to place your antenna over 20m away (67') from any human being. With a mast 10m high (33'), you won already 10 meters.

Ameritron AL-572, 1.3 kW linear amplifier.

At last, without knowing exactly how powerful can be your signal, if you position your antenna straight to the TV yagis erected by your neighbour, you will quickly receive his visit. 

These few examples of behaviours do not respect either the ham spirit nor your health and are not very useful. And all the more that an amplifier is often expensive and bulky, and requires some basic knowledges of the subject if you want to respect the other hams working on the air as well as the regulation. So think about these potential problems at twice before investing in such an equipement.

If you are convinced by the utility of an amplifier here are some advice if you decided to buy or build a tube(s) linear amplifier to work in HF.

I took the example of a kW class amplifier as it is widespread in the ham community among hams who have the chance to be allowed to use so high powers. I shall describe all its internal components and their characteristics, as well as features to check with care. Basics considerations are universal and can of course be applied to all other models, working in HF, VHF, UHF, whatever, doesn't matter, with the exception that in centimetric bands, high power entail new technical constraints on electronic components and wiring systems.

Kilowatt-class amplifiers

The Mosley TA-33 junior, a five-bander beam.

You will profit at best of your amp if you work with a directional antenna, like this five-bander beam Mosley TA-33-JR-N.

Why using so much power ? This is a pertinent question. But don't be offended by the ham telling you that you have a big mouth, because there are some good objective reasons in using such a power. As said some Old Timers, you might answer to this people, "you know kid, the live is too short for QRP, go QRO !". But enough of this joke.

Going up from 100 W to 1000 W PEP, the power ratio is 10 and looks high if not huge. But knowing that you add only 3 dB each time you double your power, 1 kW output produces only a 10 dB power gain on the S-meter, what is not really important. However your QRK will be much improved and your correspondent will have please to hear you in better conditions, mainly if you work with a directional  antenna, in which you can concentrate all your RF signal.

In all quality tubes amplifiers (i.e. Ameritron AL-1200X, QRO HF-2000, Kenwood TL-922, etc) removing the cover we are often impressed with the design, the few components but how big, the care and finish of the work brought up to the band-switch and cosmetics. All high-end kW amplifiers display big, strong and rugged accessories, large coils, big capacitors, heavy toroidal transformers, big choke, for short all is big and seems to offer an excellent quality. Even the vernier, switches and potentiometers are big but all them move softly without interplay, like professional or military material, specially the Kenwood TL-922. 

Looking closer inside the amplifier, between the large components there is a large control board full or small colored active or passive components. It provides low voltage bias, screen trip and control circuitry. In addition several smaller circuit boards complete this installation, like the high voltage rectifier board, the screen supply board, the RF I/O board, the ALC circuit board and the metering board. Let's see in detail each of these components.

Voltage, current and impedance in amplifiers

We can see a linear amplifier as a "black box"; it doesn't matter what goes on inside. Whatever the band used, most,if not all linear amplifiers have 50 W input and output impedances. Used with a 100W emitter, a kW linear should have a power gain of 10 dB.

If it is a 12V solid-state linear, the voltage levels will be low; at the device terminals the output voltage is no more than about 20V PEP, but the RF current will be high, up to 80A peak for a 500W amplifier. The impedance will be rather low, less than 1W.

If it is a valve linear, working at about 2 kV anode supply, the RF voltage at the anode could be as high as 4 kV PEP.

A PM QB 3.5-750 tetrode offers a plate dissipation rating of 250 watts.

At so high voltage the RF current will be less than 100 mA RMS, the impedance being around 16 kW.

In each case, the input impedance will be practically of a similar value. The input and output (tuned) circuits will take care of matching these values to 50 W, because your RTX and your antenna need to "see" that impedance to offer as less as loss as possible and the lowest SWR.

Generally, such a device will (at lower frequencies at least) have only a single valve or transistor or tube as an amplifier. Depending on the design, it may have one or more additional active components providing bias, switching or even a receive pre-amplifier. In the case of a tube amplifier there is no pre-amplification; transmitting power tubes ensure themselves the power gain.

Power tubes

 In kW linear amplifiers RF power tubes are still made of a glass envelope, a design disused by manufacturers since the middle of years 90's (1997 at Eimac). Most metal/ceramic tubes are also discontinuing (i.e. 8873, 8874/3CX400A7, 8875, 3CX1200A7, D7, Z7, ...) as well as most tubes made in Russia, Chzek or China. The last tubes still in production until 2020 or later are 3CX1500A7/8877 triodes and alike made by Eimac/PCI and the Svetlana production. Many other tubes which production has been interrupted are however still available as spare (see links on the last page).

An Eimac 8877 triode.

But until recently, most Russian tubes were considered as not reliable, and the rumor extended deep among the ham community who rarely accepted to buy such tubes. A Tesla 8877 triode for example was not consider as reliable as the same model manufactured by Eimac, like the one displayed at left. Idem with the 8802 aka 3-500Z. 

In some cases the filament of these tubes blow out after a few dozen hours of work or they simply refused to switch on due to mechanical problems. But on the other side I can confirm you that the same tube manufactured by Eimac shown also mechanical problems. 

When we know how high is their price and the so-called quality control, such problems are inacceptable. Of course such tubes are covered with a warranty, but when this problem occurs 2 years after your purchase, it is no more question to ask for a standard exchange...

With these positive and negative experiences, nowadays there are two ways to select a tube amplifier, for short the expensive and the cheap one; in other words selecting high or low quality tubes. Let's explain this idea. 

For decades quality tubes were manufactured by Eimac, RCA or Westinghouse among others U.S. companies. But for some years similar tubes but twice, three or... even ten times cheaper were manufactured in Russia and eastern states.

 Most of these "Rusky" tubes (8877/3CX1500A7, 8802/3-500Z, etc) are always available and reliable, as performing as their U.S. counterpart. So if you have the choice and some money left, you can purchase an expensive 8877 tube and alike made by Eimac, which production is ensured for some decades. This is an excellent warranty in case of failure.

A GU-81 pentode, 26 cm tall, almost 1 kg, it develops 600 W for... $15 without its socket. Like me you can use it as decoration item !

 The positive side is that you can buy several of these reliable tubes together with your amplifier at the time of purchase. But if you want to spare some money you can consider purchasing their cheap counterpart made in Russia or even at Eimac. You do also an excellent choice because this is a plentiful tube and you can blew out several of them without be brough down by this loss.

The drawback of purchasing an expensive Eimac tube is that in case of failure you have to break your moneybox and invest between $400-1100 to replace it... With a tube manufactured in the East in case of failure you will only pay between $170-350 plus charge (about $40 for an express oversea shipment) ! With the Svetlana 3CX800 tube for example the street price can even drop as low as $50 instead of...$525 in buying a similar triode at Eimac ! Of course I cannot assure you the quality of this bargain, but is this necessary ?

To read : How to preserve your tube lifespan ?

Make a 3-kW linear with 2x GU-81M (schematic in PDF), EX8A

If I cannot take the decision at your place, objectively you have to select a tube that first suit your needs, which specifications match the ones of your amplifier (in terms of power, current, temperature, fot short its comptability). 

Then you have to select an amplifier which transmitting tubes will still be available in 10 years or later or, more secure, be able to invest today in 2 or 4 spare tubes. If these tubes can still be purchased today at a reasonable price, in one decade or two, they could reach a high price, and be much more expensive than today. Of course you will not benefit of the one year or two years of warranty offered with these tubes in stocking them in your drawer but at least you have the "warranty" to be on the air today or later at full power.

At last select also your tube according the amplification class you desire. No positive grid current, high plate-dissipation, no regulation need ? Then select a Class AB1 amp. Need a high efficiency associated with a high amplification factor, then select a Class AB2 amplifier.

   Gain and efficiency

A Svetlana TM 3CX400A7/8874 power triode made of ceramic/metal. It is mainly used in RF amplifier, in Class A, B or C.

One say that a class AB amplifier can only turn about 50% of its input power into radiated power. The rest of the input power that cannot be radiated is lost as heat. This explain why an amplifier rated at 1 kW PEP DC input develops only 0.5 kW PEP output. In the same way a device offering 70% of efficiency loses only 30% of its input power as wasted heat.

In this matter the term "gain" for an amplifier can be confusing. Here is an example compared to "efficiency" : 

An amplifier driven with 100 Watts input develops 500 Watts ouput to the antenna as radiated power. Therefore its gain is 500/100 = 5.

But its "efficiency" is different. Its efficiency is 1000 Watts input for 500 Watts out as radiated power. In others words that represents 50% efficiency for its power device, tube(s) or transistors.

But for the amplifier seen as a whole, the total power into the entire amplifier is 1000 Watts input to its final, from the main in which I plugged it into *plus* the additional 100 Watts input to the drive (base) circuitry coming from the transceiver. So the efficiency for the amplifier as a whole is :

500W / (1000W + 100W) = 45.4% overall.

But for the station as a whole, the total power into the transceiver is 200 Watts for 100 Watts output. So, overall, the station's true power efficiency is only : 

500W / (1000W + 200W) = 41.7%

RF circuit

Vacuum tubes operate at high voltages (and moderate currents); typically, a power amplifier stage requires an RF anode load resistance of approximately 2 kW to get optimum efficiency. Therefore the most visible and important part of an amplifier is the famous "RF tank circuit" named after the large tank coil that stands in the middle of this module, near the plate choke. After the transmitting tube, the RF tank and in a lesser extent capacitors determine the overall efficiency of the amplifier. 

Closeup on the Wingfoot 813 amplifier made by Dr. Greg Latta alias AA8V. At left of images, the plate RF choke (the yellow "chimney") and in the middle the large cylindrical tank coil here wounded around a ceramic cylinder. At the foreground and to the right is the small antenna RF choke.

Tank coil

The tank coil sometimes wounded around a large ceramic cylinder is made of a copper tubing (or wire), 0.5mm tick or larger. To work on the 160 meters band, manufacturers prefer using toroid cores to keep the coil in a reasonable size, but this solution offers sometimes problems if its diameter is too small or badly insulated.

A performing toroid core should display a diameter over 50 mm (2") made of 3x T-200-2 cores, red mix type. The coil should be a 12 ga (inner 2.07 mm) Teflon insulated wire or at least made of 1.25 mm polyurothane grade 1 (PUR1) insulated winding wire. It should be wrapped in glass tape, stacked on 3 deep. At last the core should be isolated from the other components and placed away from any metal to prevent arcing.

Next chapter


Page 1 - 2 - 3 - 4 -

Back to:


Copyright & FAQ