Two questions on power amps.

[jrc4558]

Good day!
I am building Craig Anderton's TDA2003 power amp to use as a bench amp for testing equipment. I have two questions:
1) Designing a power supply for it threw me against a notion of VA rating of a power transformer. Can someone explain how to interpret/understand/use it? Tube power transformers usually have their secondaries rated for mA of current. VA is alien to me, and the googling produced more mysteries...
I am mostly concerned with the amperage of the secondary windings, trying not to underrate, yet keeping the transformer small.
2) Is there a line of TDA chip amps? I know a few through repairs I did personally, replacing these chips. Are there any higher wattage ones, or maybe there's a line of them?
THank you in advance.

[brett]

Hi

VA rating of a power transformer

This is the output voltage from the transformer multiplied by the maximum number of output amps.  (Input volts x amps is slightly higher).

ie 12V AC x 2A = 24VA

It is especially important not to exceed the VA rating in warm environments (power amps, etc) where heat dissipation is reduced by ambient temperatures over 25C (80F).
cheers 

[R.G.]

Brett is correct.

There are a few things to consider on the load side.

What matters to the transformer is the RMS current. If the transformer is powering an R-L-C load, that's easy. If it's powering a rectifier/filter, it's not. This is because the current to a rectifier filter flows only in sharp pulses at the peak of the AC wave, not for the whole AC wave.

It turns out after a few pages of math that the RMS current in the transformer secondary is larger than the DC average current taken out of the filter capacitor. A full wave bridge power supply with large filter caps is the worst offender, causing RMS currents in the transformer that are up to 1.8 times the DC current out.

Further complicating matters is the issue that the load being supplied pulls DC current proportional to the loudness of the sound it's producing. Audio signals are not constant. They have an average (-ish!) level and peaks of loudness. The average DC current pulled out of the power supply depends on the average audio level, how compressed the audio signal is, and how big and frequent the peaks are.

So picking a VA rating for a transformer for audio use is either extremely complicated or trivially easy, depending on how you look at it. If you expect that you'll do full power sine waves for long periods of time, that's kind of a worst case audio signal. It needs a transformer with a VA rating of as much as four times the power output of the amplifier; since there are losses in the amplifier, the transformer providing the audio output plus the losses (could be roughly equal to the power out) times a 1.6 to 1.8x RMS factor. That would guarantee that the transformer would never overheat.

However, if you only listen to radio or recorded music at levels that don't shake the windows, chances are that such an industrial strength transformer would always be loafing. You could probably get by with a transformer with much less VA rating, as little as half the amplifier power rating or less. There is even a line of reasoning that says this is a technical advantage by protecting the amplifier in some conditions.

So anything from half the amplifier power rating to four times its rating in VA can be reasonably supported as the "necessary" VA for a power amplifier's AC transformer. Mostly people who build it themselves go with 1X because it's easy. And that's not a bad choice for musical instrument duty.

[DDD]

Константин,
если делаешь амп для себя любимого, то бери запас по мощности:
- 1.4 ... 1.6 от максимальной выходной мощности, если не боишься легкого перегруза на средней\большой мощности
- 1.6 ... 1.8 - если предпочтителен "чистый" звук;

если делаешь "для дяди", то 1.2 ... 1.4 вполне ОК.

Температура транса на полной мощности усилителя не болжна превышать 60 град. по Цельсию после 15-30 минут работы при закрытых крышках корпуса усилителя. Меряется очень просто - это болевой предел при касании транса пальцами.
Запас по мощности транса от 1.8 и выше - это уже для "ХиФилитиков" и иже с ними.
В гитарных же делах подсадка напряжения питания дает так называемый "Sag" по питанию, очень ценимый поклонниками винтажных ламповых УМ с выпрямителями на кенотронах.

[Mark Hammer]

To answer the second part of the question, yes.  Try here: http://www.st.com/stonline/stappl/productcatalog/app?path=/comp/stcom/PcStComRPNTableView.onClickFromProductTree&primaryheader=Application%20Specific%20for%20Audio&secondaryheader=Audio%20Power%20Amplifiers&subclassheader=Class%20AB%20%3E5V&subclassid=979.0&count=97&producttype=&date=Tue%20Aug%2029%202006%2014%3A22%3A50%20GMT-0400%20%28Eastern%20Daylight%20Time%29

[jrc4558]

Константин,
если делаешь амп для себя любимого, то бери запас по мощности:
- 1.4 ... 1.6 от максимальной выходной мощности, если не боишься легкого перегруза на средней\большой мощности
- 1.6 ... 1.8 - если предпочтителен "чистый" звук;

если делаешь "для дяди", то 1.2 ... 1.4 вполне ОК.

Температура транса на полной мощности усилителя не болжна превышать 60 град. по Цельсию после 15-30 минут работы при закрытых крышках корпуса усилителя. Меряется очень просто - это болевой предел при касании транса пальцами.
Запас по мощности транса от 1.8 и выше - это уже для "ХиФилитиков" и иже с ними.
В гитарных же делах подсадка напряжения питания дает так называемый "Sag" по питанию, очень ценимый поклонниками винтажных ламповых УМ с выпрямителями на кенотронах.

Спасибо! Я запитываю TDA2003 и предусилитель на трех транзисторах, так что выпрямитель диодный. НА ламповых тоже зачастую ставлю диодные мосты, т.к. "sag" не очень люблю. Те цифры что ты указываешь, это пропорция от максимального тока на выходе усилителя? То есть, например,  TDA2003 (www.st.com/stonline/books/pdf/docs/1449.pdf ) в параметрах указывает что на 4 омный динамик может выдать 6 ватт. Тогда амперы будут: квадратный корень из ватт поделнных на омы, верно? То есть корень из 1,5, что будет приблизительно 1,25 ампер. То есть транс должен быть на ~2,25 ампер. Верно?
Я не совсем знаком с расчетом блоков питания, как видишь, так что не взыщи. Проверь расчет, если не тяжело. Заранее спасибо.

[jrc4558]

thanks Mark for the info!
thanks brett, RG, and DDD for scientific input!
I've put the unit toghether. The power transformer turned out to be a 21V, 27VA, which yields ~1.25A of the secondary current. I dropped the voltage down to ~20V by means of a 100ohm resistor in series with the power supply. Before that, the raw dc was at ~27 volts.
I should mention that this is a rebuild of an existing no-name 14 watt discrete amplifier (so it says on the labels), made in canada somewhere in seventies, judging by the components used (three digit 2Nxxx transistors, etc...)
The speaker is a 12", but I didn't measure the impedance yet.
The face plate metiones depth, speed, tone, and volume controls, so I assumed that there was a tremolo there. Not there is a runoffgroove EA tremolo in there, followed by a simple treble cut tone control, and then the TDA2003. Tremolo power supply is decoupled and a voltage regulator was used.
The sound is pretty distorted an fairly quiet. Perhaps I should check the impedance of the speaker and bypass it with a large capacitor i nseries with a 4ohm resistor to reduce the Rl and increase the wattage of the output.
Will post more later, as I progress.
Thanks again!

[DDD]

Константин,
если не ошибаюсь, то КПД выходного каскада двухтактного транзючного УМ в режиме максимальной неискаженной мощности равен 78%. Добавь сюда потери мощности на диодах выпрямителя и на питание предв. каскадов. Получишь примерно потребную мощность, а потом примени тот коэфф. , который тебе по душе.
Да не скупись на запас по мощи, вопрос же у тебя не стоит о серийном выпуске изделий, где экономия так важна. Потрать лучше пяток каксов на транс большЕй мощности - не пожалеешь.

[Ardric]

I've put the unit toghether. The power transformer turned out to be a 21V, 27VA, which yields ~1.25A of the secondary current. I dropped the voltage down to ~20V by means of a 100ohm resistor in series with the power supply. Before that, the raw dc was at ~27 volts.

I don't think that 100 ohm resistor will work.  The TDA2003 is a class-B amp, and so it only idles at 50ma.  It draws much more current under load.  It can't get that current through the resistor.

I've used the TDA2040 in a couple of amps.  It's a 20W version of the same chip.  It would be happy at 27VDC.

[brett]

Hi
an easy way to drop some voltage is to put some 1N400X diodes in series (instead of the 100R).  Each one gives a voltage drop of 0.65V that is virtually independent of current. 

If you definately  want to use a resistor, a 22R (2 or 5W) would be a good choice.
cheers

[jrc4558]

I've put the unit toghether. The power transformer turned out to be a 21V, 27VA, which yields ~1.25A of the secondary current. I dropped the voltage down to ~20V by means of a 100ohm resistor in series with the power supply. Before that, the raw dc was at ~27 volts.

I don't think that 100 ohm resistor will work.  The TDA2003 is a class-B amp, and so it only idles at 50ma.  It draws much more current under load.  It can't get that current through the resistor.

I've used the TDA2040 in a couple of amps.  It's a 20W version of the same chip.  It would be happy at 27VDC.

I will try as you suggest. IF the chip burns - to hell with it, I have another 5 or so. They cost 1.50 CAD, which is cheaper than some transistors...
It turned out to be a very easy build by the way. I made it all point to point, and it still doesn't take much space! Chassis look empty now, without the pcb.