SS Source follower to drive extension speaker

[GERPUD]

Hi,

Anybody has an idea to drive an external speaker with a SS source follower?
Input signal of the SS would be the speaker out of the amp.
Some mosfet can handle hi voltage (up to 500V, ex irf820, lnd150), so It would be perfect if I could use 85vDC (from 120vAC). Only rectifier and filtering would be needed, no transformer.
I don't want something with higher gain because i want that the two speakers have the same sound level.

The idea is to have something that does like having a line out connected to a SS amplifier, but without transformer (120V), and without gain (gain 1). This could be small enough to be integrated in the guitar amp.

This shows irf820 cathode follower:
http://www.blueguitar.org/new/articles/ ... i_v_ss.pdf


Any thoughts?

[GERPUD]

Here what I have in mind.
would something like this would work?

[Leo_Gnardo]

Here what I have in mind.
would something like this would work?

Where would the other speaker lead attach?

[GERPUD]

Rl would be much bigger than 8ohm
Rs i dont know.

[pdf64]

I can't see that the MOSFET would put out much power; maybe parallel several of them, but even then..
Are you certain there wouldn't be any Vdc across the speaker?

It would be perfect if I could use 85vDC (from 120vAC). Only rectifier and filtering would be needed, no transformer

I'm doubtful that these Vdc levels (with respect to ground) would result if the line was rectified.

Transformer isolation from the line is pretty much essential unless the equipment is double insulated (eg no grounded metalwork, hence not suitable for guitar).
50s & 60s amps without such isolation are referred to as 'widowmakers'.

[GERPUD]

Error

[John_P_WI]

DC coupled speaker? There is a fundamental error with respect to the negative rail. Try to spot it.

[GERPUD]

I was expecting it to be wrong.
too simple. I would i have seen it before.
i dont know what otherwise i would not have post it...

[Leo_Gnardo]

It would be perfect if I could use 85vDC (from 120vAC). Only rectifier and filtering would be needed, no transformer

I'm doubtful that these Vdc levels (with respect to ground) would result if the line was rectified.

Transformer isolation from the line is pretty much essential unless the equipment is double insulated (eg no grounded metalwork, hence not suitable for guitar).
50s & 60s amps without such isolation are referred to as 'widowmakers'.

Woops, missed that somehow. Yeh, I had ideas about running power supplies straight off the AC line, around 1972. Then I came to realize any minor wiring error would result in a helluva shock at best, and some serious welding sparks at second best. There were no GFI safety outlets then.

If it was all that much a good idea, why haven't manufacturers been doing it already?

Although SS power amps have much more complex schematics, there are a helluva lot of SS amps available cheap now, new, major brand names like Crown, QSC, Peavey, etc. Second-hand even cheaper. If you really need the extra wattage, why bust your brain over it?

[John_P_WI]

Safety first always.

The fundamental error was that there was no dc blocking cap between Rs and the speaker. The way it was drawn, Rs and the speaker were in series and dc was free to flow from the negative rail to ground through the speaker. Basically, a short from the neg power supply rail to ground.

[GERPUD]

If i add a couplig cap to the speaker and add a transformer woud it work?

[GERPUD]

If i add a couplig cap to the speaker and add a transformer woud it work?

[R.G.]

There is no doubt that if you add enough stuff to it, you can make something like this work. The question is whether by the time you have enough additions made to it to make it safe and usable, is it still something you want.

Here's one way to look at it.

What you have shown is inherently a source follower. In the simplest version, the speaker is just the load resistor. This works, but not with the speakers we use today, because half the maximum DC and half the maximum current have to be applied to the speaker all the time. The speaker dies.

So you can add a cap to keep the DC off the speaker. Now the source follower doesn't work, so you add a source resistor to let the MOSFET work on its side of the cap. When the MOSFET is pulling the speaker UP, it has to supply current to both the speaker and source resistor, and when it's turning off, only the resistor pulls the speaker down through the cap.

So the resistor has to have a value no higher than the speaker impedance, and that means it eats at least the same power as the speaker. Actually it's worse, as this is a simple single ended Class A output stage, and the power delivered to the speaker is only 12.5% of the power used by both the MOSFET and the source resistor - best case.

So you have two choices: change the resistor to an active current source, or put in an output transformer. These choices get the efficiency up to a maximum of 25% to 50% maximum.

Notice that every watt not eaten by the speaker is wasted in the MOSFET and its associated parts. So if the efficiency is 12.5% (1/8) you spend 7 watts of DC in the MOSFET and other parts to get a watt to the speaker. Ten watts to a speaker means you have to somehow get rid of 70W of waste heat, and all of a sudden the MOSFET has to have a big heat sink or it fries.

And you have to make that much DC power to let the MOSFET have it. So there has to be a DC supply, with rectifiers and filter caps. And they generate heat too.

If you rectify the 120Vac line, you get 170Vdc. If there was some way to make this safe (with an isolation transformer) then if you have no output transformer, the speaker is getting +/-85V peaks on the signal swing (otherwise you're simply wasting 7x the power sitting there with almost no power out). That means your speaker has to have a very high impedance. 85V peak into an 8 ohm speaker is 450W. If you do that, you had to make 8*450 = 3600W of DC, of which 7*450 = 3150 Watts are wasted as heat. To make that more palatable, you can't use an 8 ohm speaker, but something much higher.

This line of thinking just goes on and on until you get to something like a modern amplifier.

I grant you it's very tempting to think you can just use one transistor and get lots of power out. Sadly, it's not practical.

[Leo_Gnardo]

This line of thinking just goes on and on until you get to something like a modern amplifier.

All that typing I was trying to save you.

And with that idea, we're back to the technology of 40+ years ago. All the volume you want, why stop at only one booster amp? Only difference, now the amps have gotten smaller, lighter, cheaper and best of all disposable. I love posting photos of the "Wall of Sound."

[LOUDthud]

The simple source follower is a class A amp. That means it burns lots of power all the time and has poor efficiency as RG pointed out. A class AB output stage would be much more efficient. This will require more transistors.

Ampeg and others make bass amps where a 12AX7 generates the large voltage drive needed and a MOSFET source follower output stage drives the speaker. Kevin O'Connor describes such a design in one of his books. Just connect the speaker output of your guitar amp where the 12AX7 was connected and it should work. Do not use a power supply connected directly to the power line, you must use a transformer for isolation.