Amping up space charge tubes

[SolderFodder]

I've built this amp:


I am currently waiting for a regulated power supply to come in, but it does work.  The main problem with the amp is that it's really quiet (it's down in the milliwatt range).  I was looking at the schematic and wondering the best way to get it up to around 1W, without a LM386 instead of the tube power section.

I've considered the following:
1) Running multiple 12k5's in push pull configuration.
2) Running the 12k5 into an op amp to boost the voltage level.  With 12v powersupply this might not have the headroom (and admittingly this is beyond my knowlege of such electronic subjects).

I'm open to any ideas though.

[PRR]

> it's really quiet (it's down in the milliwatt range).

Put a number on that. If everything is per data sheet, you get a rousing 35mW.

> best way to get it up to around 1W

OK.....

> multiple 12k5

In pure parallel, you need 1000/35= 28.6 power bottles. In push-pull you may or may not do better: you can't do the tricks we use with "real" tubes, so you are surely looking at two dozen bottles.

This may be inconvenient.

You can raise the supply voltage. Using brute math, 64V would bring you to 1W; however the tube is only rated 30V and has very odd curves, so this is not going to reach a whole Watt.


> without a LM386 instead of the tube power section.
> Running the 12k5 into an op amp to boost the voltage level.

What is the difference between a LM386 and an opamp? They are more alike than different: transistor chips under large NFB.

The main difference is that LM386 will make a whole Watt; the average opamp won't.

There IS (or was) a standard way to bump this tube to the Watt level. A big Ge PNP on a big heat-sink working SE Class A into a choke. That's what this tube is for: big slow transistors arrived before devices fast enough for radio frequencies. The fat Ge was pooping-out by 5KHz, but that's ample for AM car radio.

Also: try a bigger better speaker. 35mW into a Full 8x10" Stack will wake the baby.

[SolderFodder]

Thanks PRR.  Kind of what I suspected.  Still learning tubes and transformers and how they interact.

Currently I am feeding it into a 12" junker.  About to order a speaker for it and was wanting something to A) make it louder and B) be a little silly with it.  Like make the build a replica of a Marshall cabinet only with four 6" Jensen mods.  Is bigger always louder with speakers?  I was assuming smaller speakers were going to be more effecient with low watts.  Do I need to focus on quantity and size?

[PRR]

> waiting for a regulated power supply

BTW: if you are anywhere near 12V, there's no reason for regulation. Tubes aint fussy. And this series was designed for rough use: car radio. It has to play OK as the battery drains to 11V, and survive long cruising at 14V.

> Is bigger always louder with speakers?

It Depends.

The softest speaker I know is also one of the largest, 18 inches.

However another 18" is one of the loudest air-slappers I ever met.

The difference is heavy or light cone, long or short coil. Also magnet, though over a wide range the magnet size has more to do with power handling (melting) than efficiency.

> I was assuming smaller speakers were going to be more effecient with low watts.

Air is very "thin". In a speaker, area is good, mass is bad. Obviously large cones want to weigh more, so this is a compromise. But a small cone CAN'T grab air well.

Less than 6" nominal size generally won't have efficient grunt. Over 12" nominal size beams the highs. That's why most serious (not "practice/beginner") amps run 8"-10"-12" speakers. Even a 12" won't grab a whole bass-wave, so the four/eight-stacks rule for MAXIMUM LOUDNESS. They will make the most of a 200W Marshall; they also make the most of a 0.035W flea-amp.

Still, you won't get near "1 Watt loudness" with 0.035W into any practical speaker.

I'm just saying that a speaker built to make the most out of modest-power tube amps would be the best bet even at flea-power.

Four MODs might be fun. If the Eight is the same price as the Six (and assuming that size is not a problem), I'd go for the bigger model. However one MOD12-110 might be a wee bit louder and more holistic than four-8". (The -110 has high power handling, but also for marketing reasons it has the highest efficiency of the series. Yes, it does seem silly to buy 3,000 times more power rating than you need.... at least you'll never blow the speaker!) The MOD10-70 looks like serious efficiency for the price.

There is well-developed tube technology for about a Watt. Every kitchen radio had a ~~1 Watt tube output stage working with 100V supply. 35L6 or 35C5 does the deed. Today, I would favor 12FX5: gobs of gain, sweet bend, and very cheap ($4). But that does mean keeping your fingers off the circuit while power is on, very different than debugging 9V stomps by finger-poke.

[SolderFodder]

Good stuff, PRR.  Thanks.