Amplifier Technology: Tube vs. Transistor

Choice of active device technology in a kilowatt class amplifier continues to be a relevant question in 2021. Be wary of those who will not allow that there is a valid choice to be made based on their merits and demerits. Rigid adherence to one technology or the other are rarely founded on the evidence. Never mistake a preference for a universal truth!

There is no absolute best when it comes to tube versus transistor technology for kilowatt class amplifiers. However, it is certainly true that tubes own the past and transistors own the future. Although we are currently well along the transition to solid state there is a strong case to be made, today, for vacuum tube technology. Since I am making a choice for today, I considered both.

This spring I sprung for a new amplifier. I've talked about it for a while yet did nothing. There was no immediate rush since it is primarily for contest work, and since there is a pandemic -- no multi-ops here -- I delayed doubling up. The Drake L7 was good enough for the interim. But with sporadic E season approaching and my lofty objectives for 6 meters this year, it was time to act.

After considering many alternatives I settled on the Acom A1500 kilowatt amplifier which covers 160 to 6 meters. It is not in any sense the "best" amplifier. It suffices that it fits well with my current and near future operating objectives. 

I will use this article to explain my decision process. Everyone's needs are different and my choice may not be your choice. When I eventually come to replace the L7 it will not be an A1500, and indeed it is likely to be a solid state amplifier. Having different types of amplifiers has its advantages. That said, over time it is almost certain that my shack will become all solid state.

This article is not a review of the A1500. Perhaps I will do one later. My purpose here is to review the pros and cons of alternatives relative to my various operating interests. I believe it is more important to understand that rather than simply going on a quest to find the best amplifier, which doesn't really exist.

Operating objectives

One of my most important applications of amplifiers is HF contests. For that I have several requirements:

• Reliable
• Rapid tune or no tune
• Tolerant of antenna mismatches
• Protection from operator errors 
• Low distortion: modulation and harmonics
  
• Ability to switch bands under software or hardware control
• Quiet

For daily DXing my requirements are a little different:

• Rapid on
• Rapid tune
• Tolerant of antenna mismatches
• Protection from operator errors
• 6 meters
  

Contests are scheduled and typically run for 12 to 48 hours. For general DXing time can be critical; that is, you want to use the amp as soon as you hear the DX or see the spot, or you risk competing with the descending horde. Leaving an amp running all the time is not too much of a problem for solid state devices. In contrast, tubes will experience gentle wear and burn electricity that will heat up the shack. That can be cozy in winter but not when the weather is hot. However, repeated power cycling can significantly reduce tube life.

That is one important difference for choosing an amplifier and device technology. What follows is my perspective of how my requirements map onto amplifier alternatives. While your interests will differ the overlap is likely to be large and therefore of potential interest.

Tubes

Transmitting tubes in the 1 kilowatt category are becoming an endangered species. They were never primarily marketed to hams and with the near 100% transition to solid state of commercial transmitters in this power category most are no longer manufactured. There still exist NOS (new old stock) tubes, if you can find them, and there are clone manufacturers that eke out a living producing a select set of discontinued tubes marketed to hams and others.

Caution about clones is understandable, although there are products that have proven to be reliable and a good match to the original. Two that are particularly important to me are Taylor and Setec. The first makes the 3-500Z clones used in my Drake L7, and the second makes the Eimac 4CX1000A clone used in my Acom A1500. Their prices are reasonable though not cheap. Manufacturing high quality products to a rigid spec entails significant cost, and deviations can quickly sink the business when unhappy customers and amplifier builders look elsewhere.

Glass envelope tubes are typically less expensive, although you'll almost always need 2 or 3 to make a kilowatt. They are physically large and kilowatt tubes like the 4-1000 are larger yet; it takes a lot of glass surface to dissipate the generated heat. Ceramic tubes are compact, with more critical air flow needs. Their small dimensions make more practical an amplifier that covers 160 through 6 meters. Tetrodes operated class AB typically have excellent distortion specs.

Well made tubes in a well built amplifier can be remarkably trouble free for a long, long time. The best tubes in a poorly designed amp, or operator abuse, can rapidly destroy the best tubes. Ground grid triodes are more tolerant of operator errors than tetrodes. It is recommended that these amps have built-in protection to prevent damage (more on this below). Vintage tube amps rarely do.

All tubes require warm up before use. Most glass envelope tubes have filaments and cathodes that are reasonably tolerant of short warm ups, although waiting a minute or two is recommended. Most ceramic envelope tubes have strict warm up requirements, typically a 3 minute minimum. This is fine for contesting but not when you need instant power to work a rare DX station.

Due to the high voltage and large output transformation ratios, it is rare for tube amps to have broadband output networks. It just isn't economical or practical.

Transistors

In the early days, transistors RF amplifiers were low power bipolar devices that required power combiners and other circuitry to put up to 8 of these devices in parallel. They were technological marvels that were rarely up to the high demands of active hams. In the decades since then the improvement of RF power transistors has been remarkable. There are now single LDMOS devices that will deliver a kilowatt of power from 160 through 6 meters.

As with any new technology, there is limited data on device lifetime. Performance is often not on par with the very best tube amplifiers. However, there are developments such as feedback between a well integrated transmitter and amplifier that deliver exceptionally reliable performance and low distortion. There is little doubt that the future of RF power amplifiers belongs to solid state.

Due to their small physical dimensions the great bugaboo of transistor amplifiers is heat. Inventing junction structures for efficient heat transport and a high frequency cutoff, while ensuring efficiency and linearity, have been challenging. Low tolerance to poor operating conditions and practices has required strict electronic protection measures. That has had implications for operating flexibility and amplifier reliability.

Progress continues apace. Increasing numbers of kilowatt class transistor amplifiers are reliable and tolerant of mismatches for an SWR of 2, and sometimes higher. The critical challenge is good design and construction since mistakes can be costly. Protection circuits are mandatory.

Mismatch tolerance

In a fantasy station every HF antenna is exactly 50 Ω at every frequency in its range. While every ham would benefit from this it is the contester that is most attracted to the ideal. Low SWR means no operating time is lost when changing band or frequency, thus (ideally) enabling a higher score by allowing time for more contacts and to catch elusive multipliers.

Real antennas are quite different. Multi-band antennas have narrow SWR bandwidths. This is also often the case with any antenna for 80 and 160 meters. For example, at right is the measured 20 meter SWR of my Hy-Gain TH7.

Even if you could achieve the ideal, weather will throw a wrench into the works. Rain, snow, ice and sometimes wind, alter antenna impedance, and in severe cases can greatly distort the pattern or make the antenna unusable. 

We simply do the best we can. Our transmitters and amplifiers need to cope with the real antennas out in the weather and their effects, predictable or not. But if your broadband amp cannot be tuned and the SWR climbs out of control at a critical moment, you are stuck.

ATU 

Adding a built-in ATU to a kilowatt amplifier typically increases the price from 30% to 60%. This or an external manual or automatic tuner may be needed for an SWR as low as 1.5 on some bands, and almost always when the SWR exceeds 2. The heat dissipation and heat density associated with high power devices demands less tolerance for impedance mismatches compared to low power equipment.

Manual tune tube amps are usually happy up to an SWR of at least 3, depending on the complex impedance at the antenna port. Since this is difficult to dynamically measure, most protection circuits keep it simple and measure either or both of the SWR and absolute reflected power, and make decisions based on that data.

There are a few tube amps that feature a built-in ATU, such as the larger version of my A1500: the Acom A2000a. ATUs are more common on modern solid state amps.

Broadband outputs or an automatic ATU would seem to be ideal for the active ham or contester to deal with our various antenna impedances. The reality is not so simple. Even if you are able to have a low SWR on all of your antennas there will be problems. Consider the following:

• Rain, ice and snow will alter the impedance of an antenna, often more than you expect
• Low bandwidth antennas -- trapped tri-band yagis and almost any low band antenna -- have high SWR at the band edges, and often over a large fraction of the bands they cover
  
• Switching among antennas on a single band -- something contesters do a lot -- will require retuning the ATU, switching the ATU in and out, or manual adjustment
  

The point is, whether you have a broadband output or an instant switched ATU, you will have to adjust tuning. It isn't a once-and-done deal. The tuning may be manual or automatic but it must be done. Choose antennas that reduce the need for tuning adjustments.

Protection

The A1500 is not my first tetrode amplifier. In 1985 I purchased a Collins 30S1 that used the Eimac 4CX1000A. As will all of these vintage amateur radio amplifiers there was no protection circuitry. You had to watch those meters and tune carefully! I was very nervous in contests since rapid tuning adjustments when switching bands and antennas was risky.

The seller included a 4CX1500B, which is plug compatible to the 4CX1000A and is a little more tolerant of poor tuning, though not by much. From what I've seen of the specs, the Setec 4CX1000A has grid specs similar to that of the more robust 4CX1500B, but don't quote me on that.

Compact tetrodes are very sensitive to secondary emissions that can quickly degrade and ruin them. It would be foolhardy to have a modern tetrode amplifier without a full range of electronic protection. DXers and contesters must rapidly tune, change bands and frequencies, and change antennas that require adjustment of the amp. Doing so without protection can be very expensive. We all make mistakes.

Kilowatt class solid state amplifiers have their own extensive protection requirements. Here the problem is excess heat, voltage and current in the active devices and transformers due to load mismatch and faults. Without protection, failure in a kilowatt class solid state amp can happen faster than you can react to meter displays. 

In contrast, a grounded grid triode amp is far more tolerant of abuse and some protections are optional rather than must have features. The smaller set of protections on late model triode amplifiers is not necessarily evidence of poor quality.

Price

Neither tube nor transistor kilowatt amplifiers are cheap, and are similarly priced. All require a hefty power supply, expensive active devices, heat removal systems, and RF circuits and switches that can deal with high heat, voltage and current. Some amps are more economical by using several less expensive, lower power tubes or transistors. There are a few manufacturers that cut the price, and corners, by reducing protections or providing little headroom in their published capabilities. Avoid the latter unless you enjoy gambling.

Before you opt for the cheaper amp you need to question whether you are pursuing false economy. Many vintage amps are available at attractive prices. Just be sure you know what you're getting into since you will have fewer protections built in and you will likely need to service the equipment soon after acquisition. This most often involves the tubes or power supply. Worse is a failing high voltage power transformer. They have a long though limited lifetime and are expensive to replace, if you can find one.

As mentioned earlier, if you need or want a built in ATU you should expect to pay a hefty premium. An external high power ATU can be purchased at a lower price, but it may not integrate well and will therefore require manual band selection. Alternatively, you can forego the ATU by making your antennas as broadband as possible, at the risk of not being able to use the amps when an ice storm hits right during a contest or DXpedition.

Service

Getting service for amateur radio equipment is worse in Canada than in the US and Europe. All the manufacturers of transceivers, amplifiers and other pricey equipment are foreign, and domestic retailers have largely exited the service market, even for equipment under warranty. When you need service under warranty the equipment must be shipped across borders, in both directions, with all the expense, paperwork and risk that involves. Amplifiers are heavy and fragile, so shipping to a professional service centre is to be avoided if possible.

The ideal is equipment that is reliable and that is easy to repair by a knowledgable ham. These days, amp manufacturers don't particularly like hams who open their equipment, let alone replace the tubes, and will void warranties when it is done. Tube equipment, especially vintage tube amplifiers, are easier to work on simply because their size is larger and you can get in there to inspect, test and replace components more easily than in tightly packed solid state gear. Unfortunately that temptation to dive in has led to many regrets, so never underestimate the size of the problem, or your abilities.

Changing the transistors in a amplifier has risks beyond the mechanical and electrical skills required. For example, heat sink compounds are almost all toxic and require expert handling to avoid health risks. Broadband matching networks are a black art for most hams and they would not (or should not) attempt to repair or rebuild them.

Power supplies are easier. Many hams can repair these when the need is parts replacement. Care is required when working on high voltage supplies (tube amps) and high current supplies (transistor amps). When in doubt, seek a professional.

There is a false meme that says the more expensive a product the less you need to worry about pricey repairs. The underlying assumption is that you are buying peace of mind for the higher price. This is as untrue of amplifiers as it is for appliances, vehicles, entertainment equipment and other products. The truth is that lifetime service cost tends to track the purchase price. Parts are more expensive and service rates are charged at a premium for premium products. Of course, cheap, poorly built equipment will fail sooner and require more frequent repair or replacement, but these repairs are typically less costly, and you may be able to do it yourself.

Station automation

An amplifier that matches the transceiver looks very pretty on the operating desk. That is true whether it is an SB220 paired with an SB102 or K4 paired with a KPA1500. In my opinion, appearance is at best a secondary reason to consider a matched pair. Although the interconnection between any transceiver and amp is straight forward, the benefits of well integrated products can pay big dividends in station automation and therefore operator effectiveness.

Some of these benefits include:

• Automatic band changes, following the transceiver, or both following a software app
• Low distortion, by managing gain by negative feedback (ALC) and, now emerging, compensation for amplifier generated distortion (e.g. PureSignal)
• Software control of the ATU, antenna port, metering, etc.

Perhaps surprisingly, there are a number of modern tube amplifiers that do these things well. After all, the electronics and software control surrounding the core of the amp can be the same for both tubes and transistors. However, most of the design and product energy is directed at the latest generation of solid state amplifiers. This is not surprising since, whatever you think of tubes, they are not the future.

Obsolescence

Would you purchase an amplifier that uses obsolete tubes or transistors? What will you do if and when they need replacement? Take advantage of current suppliers of NOS components to stock up immediately. If you wait until they're needed they may well have become unobtainium. The transistors in older solid state amps are becoming hard to find, new or NOS. Not only tubes become obsolete.

It is not just these headline parts to be considered in an old amplifier. Consider meters, transformers, relays and all the rest. I have had problems locating obsolete digital ICs to repair old equipment. Looked at individually, these are reliable components. Put hundreds of reliable components into an amplifier and there will be inevitable failures among them. Know what you're getting into when you buy an old amplifier.

How my amps measure up

The 40 year old Drake L7 is a well-built product. It is pretty easy to adjust on the fly due to its grounded grid design and 3-500ZG tubes. But it's old and parts fail. So far these have included the high voltage filter, T/R relay and variable capacitors. The lack of 6 meters is not a concern for those only interested in the HF bands.

The A1500 is a well-built modern tube amplifier that makes the most of the Setec 4CX1000A tube. It also covers 6 meters, which I had to have. During contests it requires more attention than the L7 because the tetrode is more sensitive to impedance changes when I change frequency and switch among antennas. It takes only a moment to tweak the load control to compensate, and you must do it to avoid tripping the protection. Protection circuits trip more easily at the high end of the amplifier's power range.

Despite the extra attention it requires, it is nice to have an amp that will comfortably and cleanly drive my diversity of antennas. As I evolve away from tri-band and other high Q antennas it may be possible to replace or complement it with a broadband transistor amp. When that happens the L7 will likely be kept on the shelf as a spare.

My problem now with the two amplifiers is that the operating desk is crowded! What you see in the picture is messier than usual due to the parts and test equipment for the various projects I am working on. Stuff gets shoved aside or I do a proper cleaning before major contests. 

One nice feature of fully automated amplifiers is that they can be taken off the operating desk. Should  you do that, keep them readily accessible for service or replacement during a contest. Manually tuned amps must be on the operating desk. They are typically laid out horizontally due to their weight (the A1500 is almost 30 kg). The L7 is light enough to stack on the transceiver because the hefty power supply is a separate unit.

Until the band pass filters are ready I am not using both amplifiers. All should be ready for the fall contest season. I will not be revisiting the amplifier choices in my station until the antennas and station automation are significantly improved.