Tube Amp

This tube amp is something suitable for use as a small computer speaker amp. Inspired by the recent discovery of the Morgan amp.

The Morgan amp is a simple three tube single channel amp using old AC/DC 5 tube radio parts.

A 50L6 output, 35Z5 rectifier, and 12SQ7 pre-amp all cobbled together to give at least an honest 1 watt of audio to a pair of simple speakers.

Cheep, cheap, inexpensive. The Morgan amp runs directly off the 120vac line. Not a safe application, but it can be made safe by using an isolation transformer, fuse, and on/off switch.

The low cost is a direct result of the simple design.

I did not have but one 35Z5 but I did discover a few 12AV6s, 50C5s, and 35W4′s. These are the miniature equivalents to the tubes used in the Morgan amp. So I decided to use them in an updated Morgan design. The only change was to replace the tubes with the ones I had on hand.

Folk who do not have a box of parts to pick from can still build this amp. The most economical approach is to find two AC/DC radios at a thrift shop or garage sale. Working or not, these radios can provide all the parts necessary to build a two channel Morgan amp. Three of the tubes in these old 5-tube AC/DC radios are used in this amp. One of the remaining tubes is an IF amplifier and the other is a mixer.

It could be argued that this amp driven by a crystal radio can provide far better fidelity reception of AM signals than the original radio. Instead of a superheterodyne style radio we revert back to a simple detector and audio amplifier. We might loose a little in selectivity and sensitivity but I doubt anyone listening to powerful local stations would notice anything but the lack of interference, noise, and squeals with the mixer gone. (Wonder why these things are called SUPERheterodynes? Most things I see with the ‘super’ label, are anything but super.)

In my design I used the goofy printed circuit mounted tube sockets found in the AC/DC radio I had. I simply cut them out of the old printed circuit board and mounted them with screws to a scrap of unused printed circuit board I had been saving. Somewhat more work than it should have been but very effective.

The housing for the amp was made from scrap pieces of redwood that had been salvaged from a demolished redwood deck. The complete enclosure is redwood and plexiglass. The plexiglass is also of the salvaged variety.

The end result is lots of noise for little cost. Actually no cost except for the time to build the thing.

You might well wonder ‘why do this’. Okay, you can buy new amplified computer speakers for under
ten dollars but then they are ten dollar amplified computer speakers. They don’t have the characteristic sound of a tube amp.

What characteristic sound? The sound of rock and roll.

All this started when tubes were cheap and amps were popular. Folk were not too concerned about how linear the sound was. Or how well a full spectrum of sound frequencies were replicated. All they wanted was something to make noise. Meanwhile, tube amp designers were not all that concerned about building expensive stuff with super good specifications. Why bother when it cost much less to build something that just made a lot of noise. Besides this is what was selling. No future in building stuff no one would buy. It was much easier to sell a cheap tube amp than a more expensive carefully engineered amp. The rock and roll crowd picked up on the cheap amps and used them. Their cheap and distorted audio gave the music a distinct flavor and rock and roll was born.

Now days real rockers can appreciate the sounds of that distinctive distortion that you can only find in music coming from a poorly designed tube amp.

So here you have it, a poorly designed tube amp for the rock and rollers.

You don’t have to dream about it. You can actually re-live the past. At least listen to the music of the past on real tube equipment.

When you can do it for no cost, why not?


#1 admin on 02.10.09 at 12:40 pm

If you google on ‘morgan amp’ you will find a schematic of an AC/DC amplifier using a 50L6, a 12SQ7, and a 35Z5.

There are some problems with this schematic. The amp will not work very well if it is built from this schematic.

The first problem is the screen dropping resistor for the 50L6. The screen should be run at a voltage below the plate voltage but 40 volts is a little too low. Forty volts is what you get with a 22K screen dropping resistor.

When compared to accurate 5-tube AC/DC radio schematics, this resistor’s value is 1200 ohms. Note also that this resistor drops the plate voltage to the 12SQ7. The 12SQ7 does not run anywhere close to full specs with only a 40 volt supply. Note that the 12SQ7 plate voltage is further reduced through a 470k resistor. That results in 20 volts on the plate of the 12SQ7 when a 22k screen dropping resistor is used.

A value of 1200 ohms for the screen dropping resistor is pushing things a little. We ended up using a 2200 ohm 1 watt resistor and that value seems to work well with our updated amp using 35W4, 50C5, and 12AV6 tubes.

The second problem is that the output tube does not have any means of reducing voltage spikes that might occur across the primary of the output transformer. Most good designs add a .001/400v capacitor across the primary of the output transformer to prevent damage to tube and transformer due to voltage spikes.

#2 admin on 02.10.09 at 1:00 pm

The final cut of this amp gave rise to the possibility of using the two 35W4s in a full wave rectifier configuration instead of the conventional half wave circuit.

The only reason we still have the 35W4s is to use their 35volt filaments in series with the other two tube’s filaments so that we do not have to use a huge dropping resistor when running the series connected filaments off 120vac.

Were is not for that, we could use silicon diodes and get rid of the 35W4′s. Getting rid of the 35W4′s may still be a good idea to reduce power consumption but the 35W4′s look good when lit and standing next to the other tubes.

In this particular design we are using an isolation transformer with two 120 volt secondary windings. This is also a little bit of overkill. True, we need two amps, one per channel, for a stereo amplifier, but both amplifiers could be run off the same single 120 vac supply.

Since this is not the case, and we have two secondaries, and two half wave rectifiers, we can go to a full wave rectifier power supply mearly by doing a little (very little) rewiring.

The filaments of both channels would still be run off of separate 120 vac windings but now the DC part of the supply can be provided as one supply to both amps by allowing both 35W4 cathodes to be connected together and then to the B+ line and filters. Each 35W4 plate would be connected to the hot side of each 120 volt secondary winding with the other side of each winding connected to ground. The ground connections provide a return for the DC as well as the AC used to light the filaments.

What does all this accomplish? Very little. However that characteristic hum common to all 5-tube AC/DC radios is gone now.