600mW single-ended power amplifier

© Jim de Kort

A few weeks ago I was doing some measurements on a few different tubes to see if they reacted as the formulas had predicted. The final test was seeing if the calculated output power was correct as well. I don't have any fancy measuring equipment so I can only measure power output with a sine wave signal. I hooked up an 8 ohm resistor to the secondary of the OPT and started measuring. Hey, the output was really head on within a few tenths of a Watt.

This made me think of measuring the same with the amps I am using to see what I was listening to in Watt terms. Would I really need that 5W amp? I started out by listening to music for a while so that I could set the volume to a "nominal" setting. This would allow me to measure the Watts that were being produced at normal listening levels (not max output). I put a 1kHz sine on the input to see what the level was. Hmmm... my scope said 0.04W... This can't be right!! I turned up the volume so that the 1kHz sine was just bearable to my ears and looked at my scope again; 0.19W. MAN!!!!  Taking into account that this is only a sine wave (static sound) and not music I can say that this really proves that you don't need that much power when you use high sensitivity speakers.

The amp in question here is a 45 SE amp which is able to produce about 1.8W out rms output before showing visible clipping. This amp has headroom like a midget in a MAC truck. I am really impressed with what a measly 0.25W can produce in terms of sound pressure. This immediately set me thinking... Why not use a smaller tube? The smaller output power means a lower voltage which in turn equals lower rms input on the grid which in turn means that I would need less gain.

The most likely candidate in my mind was the 71A, the little brother of the 45. The output was about half that of the 45 but otherwise it was pretty much alike. Rp and mu were the same which means I could use the same OPT (= save money). Max output was rated at about 700mW, right what I need. Wow... a real "power amp" with less than 1W of output power, hah, most solid-state preamps would put this amp to ridicule :)

The driver stage(s) would normally need two stages to get enough gain and still use decent tubes. The 26 comes to mind, but I am already going to use it in my preamp and the filament is quite beefy and can be a hassle (read the 26 projects). I want a simple yet excellent amp that acts as integrated amplifier. Using two stages in front of the output tube only complicates the circuit and adds more components. So I'll need a high-mu tube to do the job. After doing some math I came up with the KC3. This is a european small-signal DHT with a mu of 30 and Rp of around 10K. This would do the job perfectly.

I use RC loading for the KC3 since this amp is only an experiment and I do not want to go all out without knowing if the planned 650mW will actually be enough. The load is about 3x Rp which is nice for distortion figures, but it drops a lot of voltage. The KC3 is running at 3mA and 125V between anode and cathode, B+ was measured at around 230V. The filaments are heated by a 2V battery to reduce noise and hum pick up. Lately I have found that battery supplies are THE way to go. They are small, simple and quiet. After my problems with the 26 I use batteries where ever I can these days.

The 71A is running at about 180V on the plate and biased at -45V over the 2K cathode resistor.. Current is a bit higher than I wanted, but no problem at all at 22.5mA. Load is a 5K OPT that was specified for 90Hz and up.

I finished the 71A modification to my existing AV5 'Århus' amps after receiving the new power transformers. Modification was done in a jiffy after re-drilling the mounting holes and soldering the wires to the appropriate pins. The general design of the amp has already proven itself during the Triode Audition in Denmark a few months back. The KC3 does a marvelous job of driving the output tube even though it's high-ish impedance and low current capability. I have yet to play in the tubes, so a taste test will follow in a few hundred hours ;)

The amps have been playing for three days now with a pair of used 271A globes. I am very pleased to say that these amps have met and surpassed my every expectation. Sound is still stabilizing since the parts are almost played in now, but as to sound levels: Whoah!!!! If someone would have told me about a 600mW amp about four years ago I would have laughed at them. Now I am just speechless... MAN! The pressure that this tiny little amp can produce is just amazing. Granted it is not the amp to blast my Prodigy cd's on and expect the plaster to fall from the roof. But it is LOUD still the same. If I wouldn't know better I would look for a secretly hidden amp that was really playing!!!! WOW!!!!

One month later :)  I hooked up a pair of 5K Amplimo (US: Plitron) OPT's today to see if there is an improvement. These suckers are huge! My 71A could fit inside this tranny with a few of it's buddies and still have elbow room. Specs are impressive, but these hardly ever tell anything about sound. WOWOW! Even brand new out of the box, they are a huge improvement over the Bartolucci OPT I was using before. Detail in the mid/high and high went through the roof, voices were smoother, bass was tighter. MAN!!! There goes my second kidney and another two quarts of blood for the bill, but it was worth it. You'll never guess, but they don't fit my chassis.. ARGHHHHH!!!!!

Today I modified my DAC output which was giving about 0.7Vrms at 0dB signals. This required me to get extra gain either in the preamp or in the power amp since sensitivity was too low to drive even the 71A's with that kind of signal. The DAC was running on two sections of an E182CC paralleled. This setup gave a gain of around 17, this had to go up! I rewired the output stage to hold the 5842, a very nice tube with a mu almost twice that of the E182CC. After the replacement I now have a gain of 31 times, enough at last.

This modification made it possible to remove the high-mu KC3 in the power amp. This is one hell of a good tube, but the high-mu is very audible after a while. The music sounds a little bit confined and stressed than compared to a low mu tube. Maybe this is subjective, but I wanted to try. The new driver tube is the 31, a dutch tube-buddy (hoi Rob) is using this tube with great success. The filament is also 2V and B+ is not much different (besides current). The Rp is almost 1/3rd that of the KC3, but gain is about 1/8th :)))  I am still warming up the amps as I type... hoping for some positive result.

I got my hands on a pair of amorphous OPT's to try out on my 71A's. YES!!!! I have been looking for a pair for more than a year. I couldn't wait to try them out. I had been running my 71's on a pair of Amplimo OPT's these last few weeks and kind of liked them a lot. The huge slabs of iron now made way for a pair of tiny OPT's. For 700mW of output power, you don't need a lot of iron :)  It was kind of plug-and-play, four wires per channel and it was running. The first few hours were a bit of a disappointment, but you cant judge a new component right out of the box. I let it play for  about three days and then finally sat down for a serious analysis of these precious little jewels. WOW!!!! Details in the highs like I've never heard, like my Lowthers have an integrated tweeter now. Very fast and clean response, details all over the place that I could never hear. Voices were very clear and imaging was holographic. These things are not coming off of my amps again :)))  I'm a happy camper!

A lot of changes since the first version :)  I hooked up the amps to a battery supply to make it all battery-supplied like my new 26 preamp. The 31 and 71A were intended for battery use, or at least they are very suited for it. See more info on the Puppy Power 2 page.

Back to the rectified B+ while I'm finishing the chassis to house all those batteries.

I switched back to the original OPT's today, Bartolucci custom made 5K's... Argh... withdrawal syndrome!!! Wow, the difference is even bigger now that I have gone back after several weeks of listening to the Tribute OPTs. Man, I didn't know the improvement was this great. I will have to hurry and finish the new chassis really quickly to try to avoid the cold turkey that's coming.


This is what they look like. I was given a
surprise with the wooden pots :)))  I LIKE IT!


I wouldn't be a technician if I didn't open it up :))) This
proto was not potted yet since it would be a shame to
waste these expensive cores on a hunch. This
amorphous stuff is really weird, very flaky metal layers.


This shows the size compared to a 71A.
More down to earth than the 15cm Amplimo's.


Close-up of the cores and windings.


If you own a Lowther or a speaker with similar efficiency that is running wide-band or bi-amped as high-pass system, then I definitely recommend experimenting with small power tubes like the 71A, 45, 12A etc etc etc... Those hundreds of milli-watts are all that you need for every-day use.

For further info look at my AV5 'Århus' page. This amp is basically identical with exception of the output tube. 

Signal section schematic of the KC3 version. About 1V input sensitivity.


B+ supply for both tubes.

Top view that reveals the layout of the mono amp.


Front view of the amplifier. The clip leads are for charging the filament battery.


Close-up of the two signal tubes, Telefunken KC3 and Silvertone 271A


Close-up of the AZ1 mesh rectifier, beautiful.