plate transformer selection

Discussion in 'Amateur Radio Amplifiers' started by KB1SEL, Jan 18, 2011.

Thread Status:
Not open for further replies.
ad: l-rl
ad: abrind-2
ad: Left-2
ad: l-BCInc
ad: L-MFJ
ad: Subscribe
ad: Left-3
  1. KB1SEL

    KB1SEL Ham Member QRZ Page

    now i understand.I know someone who bought an amp now i know why the supply is rated for what it is rated for.You have to factor in all the things that use aditional wattage and also the effiency factor of the tank circuit.

    Thank you everybody.
     
  2. WA9SVD

    WA9SVD Ham Member QRZ Page

    And THAT is "downhill with a tail wind.":rolleyes:

    Using a linear amp for AM is notoriously inefficient.
    The best way to achieve efficiency in AM mode is to use high level modulation of the final amp stage.
    THAT way, you can run the final in class "C", (the most efficient mode in analogue use) and apply modulation to the plate circuit of the final amp.
     
  3. WB2WIK

    WB2WIK Platinum Subscriber Platinum Subscriber QRZ Page

    Linear amplifiers (Class AB2, B) typically run 50-60% efficiency depending on a lot of factors.

    Class C amplifiers (good for FM, CW or plate modulated AM) can be >70%.

    Very linear amplifiers (Class A) won't even achieve 50%.

    But all this isn't terribly important in the grand scheme. You want a power supply that doesn't drop more than about 10% of its no-load voltage under full load, and one that doesn't overheat any components. If you start out with an unknown component like a transformer that you didn't design to meet a specification, the best way to determine this is by building the circuit and making measurements.

    Thermometers that are just adhesive strips that can be attached to stuff is one way to find out about thermal rise. I use them all the time, because they're reasonably accurate and very cheap.

    Regulation can be measured just using a voltmeter.

    AM is very hard on linear amplifiers because they must be tuned for the maximum (PEP) output level, which renders them very inefficient at carrier power. They work fine for AM, but they're stressed, so extra cooling and other measures to reduce stress are always a good idea.

    My linear amps here (about six of them) don't run more than about 25% or 30% efficient on a carrier, when properly peaked for operation at PEP levels. They actually dissipate as much power "idling" (carrier) as they do on voice peaks. That's counterintuitive, but that's the way it is.
     
  4. AF6LJ

    AF6LJ Ham Member QRZ Page

    :) This is quite true.
    I like the idea of high level modulation for AM.
    As Steve pointed out AM is hell on a linear amplifier.
     
  5. WA7PRC

    WA7PRC Ham Member QRZ Page

    I agree that using an "afterburner" amplifier is most inefficient. Though, for the casual AM-er, it works.

    Most of the inefficiency in a classic high-level plate/collector/drain modulated AM transmitter is in the modulator (usually run push-pull Class B). The most efficient way to produce AM is with a Class D or E modulator.

    The HV supply is twice the normal DC voltage and the modulator is either in series with the cathode/emitter/source to ground, or in series with the plate/collector/drain to the HV. The modulator is driven with a PWM/PDM signal. It is switched on/off at a rate that's at least twice the maximum modulating frequency. With no modulation, the duty cycle of the switching rate is 50%. This puts the average plate/collector/drain to cathode/emitter/drain voltage at the level that the tube or transistor would normally see. At 99% modulation on negative peaks, the plate/collector/drain to cathode/emitter/source voltage is nearly zero. On 99% modulation on positive peaks, the plate/collector/drain to cathode/emitter/source voltage is equal to the HV supply voltage. The PWM switching rate is trapped out in the output network. Because the modulator device is either on or off (a conduction angle higher than Class C), efficiency is quite high and you can get by with a switching device with a small dissipation rating.

    Broadcasters have been doing it that way for decades. They are/were allowed to have a certain percentage of carrier shift (change in avererage power) with modulation. While the negative peaks have to be limited to 99%, they are/were allowed go well over 100% on positive peaks.

    The downside to PWM is, you don't get to hear the modulation transformer sing and it doesn't keep you as warm on Winter nights! ;):D
     
  6. W8JI

    W8JI Ham Member QRZ Page

    You seem to be under the misunderstanding that transformers supply a certain power and that is all. That is not true. There are a few myths in this thread.

    A good modern high power transformer is in the 90% efficiency range under extreme loads, and is over 98% efficient at modest loads. A typical power mains transformer for a house used by utility companies is 98% or higher efficiency.

    The AL1200 plate transformer produces about 175 watts of heat at 3000 watts load. The core is a 1.8 kVA CCS core, but the transformer will easily supply 10 kVA for short periods. There actually is no power limit other than heat and voltage drop.

    Many people think a transformer is good for a certain power, then it saturates, but that is not true. Running a linear SSB/CW amp on AM, the heat limit will virtually always be in the PA tube and not the transformer. The power saturation limit will never be in the transformer, it will always be in the tube.

    As for efficiency, the theoretical efficiency of a class B amplifier stage is in the upper 70% range. A class C stage approaches 100%. In a good AB2 tube amp it is possible to get around 70% plate efficiency. A few percent is lost in the tank, but that is tank heating and not anode heating. Typically expect around 60-65 percent in a class AB2 amp.

    I have actually seen commercial class C transmitters that had over 90% plate efficiency. We had a 5 kW RCA transmitter with a single low mu triode that did well over 90% overall PA efficiency at WSPD.

    A linear amplifier is a form of efficiency modulation. This is because in a normal amplitude modulated stage plate voltage is doubled at modulation peaks. This doubles plate current in a proper class C stage, and power is thus four times. In the linear amp plate voltage is constant and plate current doubles, so something has to make the amp go 4x power on peaks. That extra something that allows the peak power is the efficiency change. Ideally plate current doubles and so does efficiency on peaks. It is exactly the same as a grid modulated stage.

    If you don't reduce power enough on carrier, the amplifer will be non-linear. Most CB amplifiers are intentionally non-linear so they compress the audio like a speech processor and "sound loud".

    By the way, nearly all those Hams who think they have plate modulated rigs do not. Most rigs, unless they are triodes, are a combination plate and screen modulation.

    See this link:

    http://www.w8ji.com/amplitude_modulation.htm

    73 Tom
     
  7. WA7PRC

    WA7PRC Ham Member QRZ Page

    Hmm. Was it this one? It appears they obtained higher efficiency by intentionally injecting some 3rd harmonic energy into the PA to raise the conduction angle higher than Class C:

    [​IMG] [​IMG]

    I recall hearing about this technique back in the 70s. Apparently, this xmtr technology is vintage 1960. It appears the latest technique is Class E. An interesting method is used to cope with harmonics:
    I should've mentioned in my previous post that asymetrical modulation wouldn't be worth the trouble for a ham xmtr. This is because the FCC limits us to 1500W PEP output, and the AM broadcaster's power specification is the carrier power (not PEP).
     
  8. W8JI

    W8JI Ham Member QRZ Page

    Hi Bryan,

    It was not a conduction angle thing, but the transition time between on and off.

    The RCA transmitter used a very low mu triode with high bias, and banged the grid hard. Third harmonic resonantors in the grid and anode allowed the grid and anode to have a steeper slope on the waveform so the tube spent less time in transition between full on and full off. Minimum dissipation occurs at two places...full on and full off. Letting the tube transition between those two states faster without increasing on time improves efficiency.

    73 Tom
     
  9. WA7PRC

    WA7PRC Ham Member QRZ Page

    I was curious if THAT was the xmtr you recall. My (legal) broadcast engineering experience was in FM so I only know of this technique from reading about it.

    Something to think about: when hams try to increase an amplifier's efficiency, we're talking about maybe a few hundred watts difference for an amplifier that isn't on 24/7. But, in a legal limit (up to 50KW) AM broadcast amplifier, the difference in dissipation could be 15KW, and the xmtr is sometimes on 24/7. This power savings can translate to a lot of money not spent. So, this xmtr was a huge step forward.

    IOW, whether you're trying to reduce dissipation or win a drag race, less transition time is better. :D ::rimshot::
     
Thread Status:
Not open for further replies.

Share This Page