audix distribution amp as a headphone amp?

So, today I was ready to canibalize a pair of audix distribution amplifiers but I give it a second thought and decided to post the schem here.

I was thinking about using it as a headphone amplifier. It has a 2.2k resistor shunting the output to the ground, but I think it can drive pretty low impedances. Do you think it would work nice? What´s the lowest impedance this thing will drive?

Off coarse I´ll be taking those input iron out to use as my 1081 clone input. It´s a T1442, or marinaire 31267.

Here it is:

http://img.photobucket.com/albums/v172/rafafreddy/3804_50.jpg

The output transistors pair is in fact a 2n3055/2n2955 on the PCBs...

> What´s the lowest impedance this thing will drive?

The 2K2 is just the least load they want on it. Mostly to drain leakage on the output cap. Look at that cap to get one clue to minimum load impedance. 4,700uFd! Assuming 20Hz response, that cap could support loads under 2 ohms! It isn't unusual to over-spec output caps, but 4,700uFd isn't cheap. If this is a distribution box, it may be that large to supress back-talk from other outputs accidentally fed into this boxes output. Still.....

What's the supply voltage? What is the heatsinking on the output transistors?

If it has 24V supply, it should drive 8 ohms up to 8 watts, before the 1A output fuse blows, unless the output transistors melt first. The raw circuit can probably drive 1 ohm loads, ignoring details like overheating and supply sag.

The real problem is that it is dangerous to itself and to low-Z headphones unless you add output resistors. With 24V supply it can put over 2 Watts in 32 ohms. On a short, it is a race between transistor meltdown and fuse blowing. In a fair fight, a 3055 will usually kill a 1A fuse, but Murphy's Law says otherwise.

If it is 20V-30V supply, and any kind of heatinking on the 2955/3055s, put 33 ohms in series with headphone jacks and don't expect to dive more than about a dozen low-Z phones, or a half-dozen shorted headphones. If it is 12V supply (seems unlikely to be that low), try 15 or 22 ohms to each jack, and don't use a lot of shorted headphones.

What is this thing really made for? How OLD is it? That's real 1970s design. EARLY 1970s. No emitter resistors.... how do you set the bias pot? Even if you meter the supply current, this thing will be awful twitchy to bias and liable to runaway. If it is working, don't mess.

"old fashioned" indeed. The biasing arrangement is a disaster. Picture what happens if the pot on the base of VT4 goes intermittent, which it will, undoubtedly, sooner or later.

At the very least, I would rework the biasing and add small emitter resistors to the transistors that do not have them.

Unless there's an output network that's not shown on this schematic, it's very, very unusual for a distribution amp to have no isolation resistors on the outputs. A well-designed DA will still function even if a portion of its outputs are shorted; this one, as shown, will not.

Sorry for the information missing.

The output transistors are mounted on big TO3 heatsinks. Also, VT4 (bc182a) is attacked to the 2n3055 heatsink throught the three red wires you can see in the module scan.

I´ve checked the mixer documentation and it seems that this thing drives at least 10 and up to 20 transformer coupled inputs in parallel. I don´t know what is the primary of the transformers. Those are T1501 and I guess it´s a regular 600ohms primary. It looks like the transformers are connected directly to the output, without any resistor network.

It´s from 76. Not that old, no? Almost 30 years. Older than me, for sure.

It runs on single +24v.

VR2 (500ohms, VR1 on the board) is set at 65% in both units. Should I just substitute it with resistors for the sake of safety??

Also, It´s writen in the schem that the values of R5b (18k) and Cx (470pF) depends on output load. There´s a table on the left of the schem, with places for writing R5b and Cx values for different number of transformers it´s driving, but all the values are blank. This makes the thing even more problematic. Note that those parts are not fitted in my pair of amps, and those came from a working AUDIX console, so I guess this is not needed? What are they doing, anyway? Frequency compensation network?

Well, it´s getting harder and harder. Do you think I really would be able to use it as a multiple headphone amp? It would be great if I could, because I have a pair of these amps and if it can drive a dozen headphones it would be just perfect for my needs...

Here is a pic of one of the modules:

http://hps.infolink.com.br/rafafredd/3804_module.jpg

And the schem again, because photobucket sucks...

http://hps.infolink.com.br/rafafredd/3804_50.jpg

Oh, and I haven´t tested this yet. Should I just put a beefy 24v supply in, a sinewave, and see what comes out? I´m afraid now.

But, if I can´t really use it, it will not be the worst thing in the world to just canibalize those modules...

Thanks for your time!

I would not even worry about the bias pot until you had some emitter resistors. I guess they rely on the fraction-ohm parasitic resistance inside the transistor; if you ever have to change a transistor, you'll have to trim the bias.

It also has gain of only 2, and you often need more than that to boost monitor feeds to headphone levels.

And while I am not a big re-cap fan, those caps are old enough to make me fret. It was 1970s when low-price caps really got better. And while yours may be from 1976, this is surely an older design and there may have been older caps in the parts bins.

What I'd end up doing is using LM1875, a 5-pin TO-220 power amp which will do all that this one can, no muss no fuss. Simple, easy to use, low distortion, enough output for anything short of Sennheiser 600s, will drive under 2 ohms if it must, and won't die no matter how many outputs are shorted.

You could build an LM1875 onto this board, even using a heatsink. Power and output cap too, if you are sure they are in good shape.

Salvage the 2N3055H. This is the correct part (and vintage) for Neve outputs. There is a difference between the -H and the common 3055. In most designs, even Neve, it hardly matters. But save them and sell them to someone who really-really wants 2N3055H.

Thanks PRR. So, this one is going DOWN!!!!

I´ll use the RCA 2N3055H on my neve boards that are ready to be soldered. Thanks!

Maybe I´ll also use this board to put a better headphone amp together.

[quote author="rafafredd"]Thanks PRR. So, this one is going DOWN!!!!

I´ll use the RCA 2N3055H on my neve boards that are ready to be soldered. Thanks!

Maybe I´ll also use this board to put a better headphone amp together.[/quote]
I ran into a few RCA 2N3055H as well - based on what PRR said I understood it'll hardly matter but did you happen to do some listening ?
RCA 2N3055H vs say a Motorola 2N3055 ? I do have Moto 2N3055 as well so could perhaps just use these but for grins I could use that RCA -H.

Bye,

Peter

I took this module apart, but I didn´t tried the transistors in a NEVE circuit yet. May do it some of those days...

Cool, if you do so then just let us know.

Of course I realize this all might be trying to improve on an already clear choice (just use a Moto or On-semi) with the risk of missing the boat altogether in the process of overdoing things.

Would make up for a nice story to brag about in the pub ! :wink:
"Then I took this TO-3 thing out and had the guts to use that other brand !"
Sad case of too less :sam: :guinness: