6N7 Tube QRP Power Amplifier - One Rainy Day DIY Project

QRP 6N7 Tube Amplifier
Once upon a time I'd decided to join a Big Guns Gang and made a Super-Duper Powerful Vacuum Tube QRP amplifier for my 800mW QRPP homebrew telegraph vacuum tube transceiver "3T" (I promise to write a separate article or two about this three tube transceiver project later). It was not an easy decision to me because for that legendary time I'd almost a year used the QRPP power of less than a watt, but there was a rainy spring day, I've called CQ again and again without any takers and, at very last, somebody took over my only rock bound frequency. "That's all" - I said - "Enough!" ..and turned on my Soldering Iron.
Let me describe the scheme. Grounded-Grid design has been chosen because it is simple, easy to matching to coaxial impedance, provides the necessary level of amplification and typically requires no neutralization. After several experiments with triodes and double triodes I'd chose a 6N7S (6H7C) valve - Russian glass shaped version of well known 6N7 vacuum tube. I know that there was a 6N7G and 6N7GT American glass tubes, but as I heard these ones was not as common as metal 6N7.
Homebrew QRP Vacuum Tube (Valve) Amplifier, Scheme
Anyway, pair of tied in parallel triodes inside this vacuum tube in the grounded-grids amplifier circuit offers more than six times power amplification and accurate input impedance matching via the 1:4 autotransformer. Maximal plate power dissipation of these twins is 12 watts, so you can got more than 25 watts output power without overheating - it should be interesting if you have Yaesu FT-817, Elecraft K2, Icom IC-703 or any other QRP transceiver, if you need some, but not too much extra power and, of course, if you love the magical glow of valves.
By the way, about a glow - in this amplifier I managed to use an automatic bias circuit stabilized by two LEDs. As said one of my friends: "I love that Christmas Tree biasing!", but in my case these LEDs are red, so we can rather speak about the deer's noses, I think. This shiny circuit provides a stabilized 2.5 volts bias and it is very easy to move the triode deeper into the "C" class - you just have to add more LEDs.
Input impedance matching transformer T1 is wound on a little ferrite tore found in the IF stage in some solid state radio - not so important, just try to find the ring (one for the Dark Lord on his dark throne for example) and wind 7 turns of the double twisted wire, then connect the beginning of one winding to the end of another as shown.
Seems that that's all about the scheme, you're ready to Do It Yourself. Oh, yes, L1 is a common RF choke, it should be possible to use a homebrew one, dozen coils of enameled wire on a ferrite rod. All capacitors are mica or ceramics, C3 must be HV type, 500 volts at least. Pi-network data see below.

Homebrew QRP Vacuum Tube (Valve) Amplifier, Rear View

I've made this amplifier for 30 meters band, Pi-network coil was wound with enameled wire on the pill bottle, 12 spaced turns, ~30mm diameter. If you plan to tune an amplifier to other bands, use following table:

6N7 QRP Amplifier Pi-network Data
Freq, Mc1.823.557.05510.1214.0518.0821.0524.9110
Coil, uH34.516.58.35.8432.521.7
Plate, uuF23113065453427.524.52220.5
Load, uuF988673339236189161150144140

And yet another one note about the glow - I've installed a low voltage tiny neon bulb into the Pi-network's coil, so that I got one extra Glowing (pill) Bottle in the QRP amplifier. This neon lamp shining brightly when output circuit is tuned to resonance. It can be seen at picture below:

DiY QRP Tube Radio Station

23 comments :

Anonymous said...

I noticed your blog on reddit. Very interesting projects and great descriptions. The added touch of the Jameson bottle made the pictures even better.

-Sulnibs

Nataly H said...

Good info! Keep up these nice posts.

IronGhost said...

Hi, love the circuit! I started with transistors as a child, so learning vacuum tubes in my late 30's is killing me. The resources are limited nowadays, and I am having to scrape and interpolate/extrapolate what I can. I think I can modify your circuit to broadcast on the AM band (just experimenting). I am having a difficult time finding circuit info for various designs, including calculations.

Andy Nechaevsky UR3WA said...

Hi, I suppose that it should be much better to build one of following simple AM circuits:
http://www.antiqueradio.org/transmitter.htm
http://www.dogstar.dantimax.dk/tubestuf/amtx-1.htm
Just because this 6n7 amplifier needs an external modulated oscillator to work as an AM transmitter - it is not as easy as those single tube broadcasters.
Good luck, your majesty!
de UU1CC

The Crusty Old Grognard said...

Hi,

I managed to source a 6n7, but wondered about the cap values. Are they in uF or pF?

Andy Nechaevsky UR3WA said...

Hi Matt. All caps are in pF, good luck.
73! Andy UU1CC

Anonymous said...

300 VDC supply, how much current... mA?

A great website! I love the formatting. It is very pleasing to look at.

Andy Nechaevsky UR3WA said...

I'm sorry, but I doesn't remember exact metrics - If you have a 100mA plate supply, it should be enough.

Unknown said...

Nice circuit! It will go well with my K1. So is the matching input transformer just wound on a round toroid, if building it from scratch? Also would a blocking cap on the input (to keep DC out of the transceiver) be a bad idea?

Tnx,

Robert
KG4KGL

Andy Nechaevsky UR3WA said...

KG4KGL de UU1CC

Hi Robert.

1. Sure, it is usual small toroid, I've used a low frequency toroid from IF tract, as I recall.

2. Blocking cap is not necessary, but if you feel better without DC connection - you're welcome :)

Anonymous said...

Hi,
Is this circuit suitable for operating a 6BX7 twin triode ?

73,

Damien

Andy Nechaevsky UR3WA said...

I think so, Damien, just try it ))

Anonymous said...

2 questions
if i want to add a second 6n7 would i parallel the 4 plates and would i parallel the cathode also.

second question you are running this amp in class c if you you wanted to bring it up to class b would you remove the leds?

I'm new to tubes and trying to learn as much as I can.

thanks in advance
73's

Troy

Andy Nechaevsky UR3WA said...

Hello Troy

If you would like to add a second tube, you have to parallel everything, but take in account that the input as well as output impedance will be divided by two, so that Pi-network should be readjusted (more plate capacitance, less inductance). As for input circuit I'm not sure - let's experiment!

Second answer: yes, one by one. By the way, you'll finish with something like the AL-811 amplifier light - four triodes in parallel. It would be quite interesting, as for me :-)

Jim Doyle said...

I'm considering this a starting point for a retro-cool linear to pair with a Peaberry SDR. I'd like abit more gain, as the drive from the Peaberry is about 1W.

I see no reason not to try the 211 or 811A at 430-450V as that is a common plate voltage available from off-the-shelf guitar amp transformers. However, the cathode impedance is going to very high.

Why not use a MOSFET in the Cathode to a grounded grid tube? Such a thing is called a Cascode. It would have low Zin and give access to more gain. Thoughts? I realize the MOSFET gate capacitance is an issue to watch for, but for 20m operation, I think I can find something (IRF501?)

Trevor Ialen said...

I've got the 6N7, but I've only got 160v. Will it still work at this low voltage ?

Anonymous said...

transceiver "3T" (I promise to write a separate article or two about this three tube transceiver project later).

Здравствуйте Сергей. Помнится что ВЫ уже публиковали материал об этом трансивере. Сейчас не могу найти. Вы убрали публикацию или я что то путаю?

Andy Nechaevsky UR3WA said...

Я Андрей Ж-)

Нет, тут не публиковал - есть кратенькая страничка со схемами http://www.qsl.net/ur3ijc/3t.html, но я же на нем долго в эфире работал, схема немножко менялась, да и хотелось подробно расписать всё, с монтажными фотками. Как-нибудь сделаю, после войны.

Anonymous said...

Я извиняюсь конечно же Андрей, ошибся.
То есть схема вполне рабочая и повторять можно? Но кварцы, хотелось бы без них, могли бы Вы схемку черкануть, и вообще возможно ли передатчиком настроится на регенеративный приемник? Кварцев нет и хочется как то иметь возможность настраиваться по диапазону хотя бы в небольших пределах

Anonymous said...

Ok here it is 2016 and I finally saw the 6N7 amp. Well I gave it a try using 2 6N7 tubes at 430 volts DC plate. It draws 25 mills with 3 leds in the cathode. Seems to work just fine. Now I have to find the right toroid mix and size to get the input SWR down. Been a fun project.

Unknown said...

Hi, following this construction I made approximately the same with the same tube at spring of 2015. But powering from 12V DC using cheap DC - DC converter from China (the plate voltage 450V). Later used russian GU-50 tube with the filament voltage 12V. You can see the pictures (sorry, the text ir in Lithuanian language) of my construction here:
http://www.qrz.lt/ly2bok/Konstrukcijos/HT%20Romantiskasis/Romantiskasis.htm
And also here: http://www.qrz.lt/lhfa/index.php/lt/video-foto-4/166-link-kauno-mari-pakalniskiai
The power on 80 and 40 m is near 50W. On 20m about 40W.
73 from Lithuania. Vidmantas (Vidma) LY2HT.

Unknown said...

Hi, following this construction I made approximately the same with the same tube at spring of 2015. But powering from 12V DC using cheap DC - DC converter from China (the plate voltage 450V). Later used russian GU-50 tube with the filament voltage 12V. You can see the pictures (sorry, the text ir in Lithuanian language) of my construction here:
http://www.qrz.lt/ly2bok/Konstrukcijos/HT%20Romantiskasis/Romantiskasis.htm
And also here: http://www.qrz.lt/lhfa/index.php/lt/video-foto-4/166-link-kauno-mari-pakalniskiai
The power on 80 and 40 m is near 50W. On 20m about 40W.
73 from Lithuania. Vidmantas (Vidma) LY2HT.

Unknown said...

Hello Andy, thank you for this project, quite amazing!
I'm about to build your circuit, but I'm planning on using two EL84 tubes in parallel connected as triodes (?), accodring to the datasheet g2 connected to a. and 250V plate voltage.
But I'm not sure about the impedance in that case... could you help me understand the calculations for the impedances? Also I'm winding L1 on a 50-43 ferrite toroid, 12 turns #22 wire, is that right?.
Thanks, 73!
Dan.
HK4DEI