A few basic noobie questions.

[audio_moksha]

Hello everyone,

I am new to the diyaudio forum and tube amp building; Noobie in the highest sense. These past few weeks I have been reading intensely about tube amp design. I Just bought Morgan Jones' Valve Amplifiers and will soon complete a K-12 tube amp (Webtronix.com)

I just have a few basic questions and I would be very grateful if someone could answer it:

1. How can I obtain a tube amp's power ratings? Is it at the anode/plate voltage?

2. I was reading about how one could build an active current sink using a Cathode follower(instead of simply using passive components. What are the tradeoffs between the two? Do people generally use active current sinks in hi-fi designs or do they just stick with passive components?

3. Are there any general opinions as to which amp configurations sound better eg. mu follower vs cathode follower? or perhaps is it futile asking such a generalized question?

Thank you everyone!

[SHiFTY]

1) Power ratings are usually supplied for any amplifier, sold state or tube. A rule of thumb is less than 33% for single-ended amplifiers, and less than 50% for push pull. For example a 2A3 tube dissipates 15W but will give you 5W audio power, and a Push-pull 6L6GC dissipates 60W per pair but will be lucky to give you 30W, unless run in class B with crazy B+ voltage and fixed bias where you can get to 100W easy. Anything is possible if you have high enough voltage supply.

2) Start with the basics- use a resistor. Then, build more advanced active devices, decide for yourself. Depends what kind of sound you like I guess, and whether you prefer simplicity over mathematics.

3) You might as well argue Ford vs GM! Cathode followers measure well but sound sometimes less than ideal, whereas ordinary anode followers measure awful and sound awful! Lots of room for endless experimentation. Depends on the tube used and the circuit too.

[Ardee]

Morgan's books are excellent, and essential reading for anyone interested in making valve amplifiers. He does, however, tend to shun the conventional method if there is a theoretically superior way of doing things. Whether the extra complication makes things sounds any better is arguable, so I'd always try simple first!

[grhughes]

1. How can I obtain a tube amp's power ratings? Is it at the anode/plate voltage?
It would require a Wattmet calibrated at the output impedance you want to measure it at. Watts are totally dependent on load. Current is delivered to the load by the load.
2. About cathode follower active sinks:
Some of them that use active mosfets have to use a rather large heat sink as they dissapate quite a few watts. A 6dj8/7308 with a 22K, 5 watt resistor on its cathode to a -125 volt supply will dissipate over 3 watts just in the resistor. Thermal runaway in a solid state device is the big fly in the ointment of transistors.

[Ardee]

1. How can I obtain a tube amp's power ratings? Is it at the anode/plate voltage?

Any reason why this should not be done by putting a known resistance on the output, together with a 'scope and digital AC voltmeter, increase the level of a sine wave on the input until the output shows visible clipping, back off a bit and note the output votage? Power is (VxV)/R.

[Sherman]

Originally posted by Ardee
1. How can I obtain a tube amp's power ratings? Is it at the anode/plate voltage?

Any reason why this should not be done by putting a known resistance on the output, together with a 'scope and digital AC voltmeter, increase the level of a sine wave on the input until the output shows visible clipping, back off a bit and note the output votage? Power is (VxV)/R.

That's the way I do it.

[tubelab.com]

Any reason why this should not be done by putting a known resistance on the output, together with a 'scope and digital AC voltmeter, increase the level of a sine wave on the input until the output shows visible clipping, back off a bit and note the output votage? Power is (VxV)/R.

I also use this method, with 2 considerations.

Many digital meters are inacurate when reading a frequency other than 50 or 60 Hz. They are also inacurate if high distortion is present.

If you have a scope you can check the behavior of your voltmeter. Use the scope to get an estimate of the voltage, check the meter to see if it agrees with the scope. Raise the frequency, and see where the voltmeter goes wrong. Most are reasonable at 1 KHz, but really wrong (200%) at 10KHz.

Most amplifiers are specified to have a given power output at a certain distortion level. For solid state amps this is usually some number preceeded by a decimal point and a few zeros. For tube amps this is often 1%, 3% or 5%. Using a common distortion figure allows comparing power outputs between different amplifiers. Measuring power at the onset of clipping can add a measure of variabillity, but it is close.

I have a fancy HP8903A audio analyzer that makes all these measurements easy. You tell it the load impedance and it tells you the power output. They are not cheap, even surplus. I also use a HP331A distortion meter. They can be found cheap. I got mine for $25 at a hamfest.

Keep in mind that some manufacturers inflate their numbers, and some flat out lie! The amplified speakers on this computer claim to put out 600 watts! They are powered by a 12 volt 1 amp wall wart.

[audio_moksha]

Thanks everyone for all the useful information. I think I'll try all you're suggestions. I just acquired a digital scope and as soon as I get it all figured out I/m gonna try using it. Cheers!

[richwalters]

Originally posted by tubelab.com


I also use a HP331A distortion meter. They can be found cheap. I got mine for $25 at a hamfest.

HP must have made mill's of these in the early 60's....I also use one of these grey ones (darned reliable), however I'm after the auto version.....these are rarer.

richj

[tubelab.com]

I have a 331A (grey one) that lived its life in the Motorola factory testing HT220 radios. I got it from a hamfest after the factory scrapped all of the non automated stuff. It has worked well for the 20 years that I have had it.

I got a HP8903A digital audio analyzer (for $75) from the same guy. It was dead when I got it. I fixed it, but it died again twice more in the 8 years that I have used it. It was a different failure each time. It has recently died again. I have not performed autopsy yet. The newer HP (or Agilent) stuff seems to die a lot.

If I can remember right you are looking for a 333A.