Schematic for XLR/balanced line driver?

[mikeb]

Hey all...

Does anyone have a schematic for a single-ended supply balanced line driver / XLR driver?

TIA

Mike

[lovekraft0]

The old-fashined way is to use your favorite gain stage and slap a 1:1 transformer with a center-tapped secondary on the output.

http://www.jensentransformers.com/as/as022.pdf

[mikeb]

Thanks ... I should have mentioned, no transformers. It needs to fit in as small a space as possible. There's a circuit over at Rod Elliots sound pages, but that one is a for a bipolar p/s.

Mike

[Johan]

google "datasheet SSM2142"...

Johan

[Marek]

Try this:

http://www.henryeng.com/Support_pdf/Portamatch.pdf

Greetings,
MM

[mikeb]

Thanks .. heh-heh, I'm getting fussy:
- the SSM2142 is a crazy AUS$19 per IC, making it too expensive even for a one-off
- the portamatch circuit requires a bipolar power supply

Hmmm .... guess I'm going to need to convert one of the bipolar circuits to single-ended.

Mike

[Mike Burgundy]

It's real easy to do with run-of-the-mill IC's such as 072's or 62's.
Aahhmm... (rummage rummage)
hold on, I'll come up with something.
Basically you use one opamp for buffering and + out, and the other for the inverted out. The trick is matching both outputs but this is relatively easy I think.

[Mike Burgundy]

Okay, can't really find anything at this moment, but if memory serves, a regular gain driverstage (with gain pot in the feedback loop, or any configuration you like. Buffer, too.) can be used. Take the output of that opamp directly to a 10k to the inverting input of another opamp. +input to mass (or vcc depending on how you make your bipolar supply) and a 10k in the feedback loop of that opamp.
Make sure output series resistors match, and bob's yer uncle.

On bipolar supplies: that's not hard. Have a look at all the loads and loads of pedals using dual opamps (4558, TL072, etc) and see what they use to fool the chip into believing it has a bipolar +/- 4.5V supply.
The trick is you use a voltage devider to creat half of the supply voltage (Vcc), and float in-and outputs (well, they sort themselves out-the opamp does that) *at that level* isolating the rest of the circuit with caps. The opamp now thinks that Vcc is ground.

If you need a drawing, holler (I'm lazy, and flu-ridden )

[lovekraft0]

I'm using this to power a TL062 from a single 9 volt.

It hasn't given me any problems, but if there's anything troubling about it, I'm sure someone can chime in with any necessary mods for your application.

[Marek]

Basically you use one opamp for buffering and + out, and the other for the inverted out. The trick is matching both outputs but this is relatively easy I think.

Isn't this what people call "pseudo balanced" because it's not really balanced?

This article I found recently:

http://www.film-tech.com/warehouse/wareview.php?id=917&category=2

It may, of course, be that I got it all wrong :-)

Marek

[Mike Burgundy]

It's easier to match both signals that way, but yes, there's a stage more in the inverted signal, so it's not truly balanced. Electronically, it's safe to assume it is for all intents and purposes.
If you want what that paper titles as "true balance" , you'll need to feed both inverting and non-invrting opamp from the same source. Bit harder to match both outputs.
If you *really* want to pick nits, it could be argued the only true way to go is transformer-balancing, but that's expensive.

Have a look at the Rane site, discontinued products. Lot of schems in there for just this stuff.
http://www.rane.com/oldman.html

[Paul Perry (Frostwave)]

If anyone ever has a need for a not-quite state of the art balanced driver, and didn't want to spend more than a dollar, maybe a single transistor phase splitter (you know, 10K in the emitter, 10K in the collector) would do.

[bwanasonic]

Not DIY, but there are a few relatively inexpensive direct boxes that you might want to look at. Rolls makes a DI in the $65 US range, and the Aphex line (punch factory, compressor) can also function as a DI. They go for @ $120 US, and are nice effects as well. Since you mention size constraints, I'm not sure either of these would work for you, but...

Kerry M

[JimRayden]

Or check out the first link of this thread, there is a non-transformer version too.

I know, I'm good.

-----------
Jimbo

[mikeb]

Thanks for all replies ... I guess I really should have been more specific. I knew pretty much how to do it, and how to use a single-ended supply, just wanted to see a 'ready to go and verified' schematic somewhere; as it turns out Mr Tremblay from tonepad pointed out a useful snipped from Jack Orman's CD which, combined with the buffer stage, did the trick. BTW, I think part of the balanced thing is not necessarily to provide inverted signals on both 'lines' (if you have a look a lot of so-called balanced line drivers only actually provide signal on one 'line') and also the output impedance is important. However, for my DIY application, Jack's snipped is just what I needed.

Mike

[B Tremblay]

...Mr Tremblay from tonepad...

That would be runoffgroove.com  :wink:

[toneman]

the PAIA SIAB(StackInABox) has a differential out on the front panel.
it uses both parts of a 5532(lownoise) opamp.
the 5532 can drive headphones, so it *can* drive 600ohm inputs.
PAIA uses it in their HeadphoneBufferAmp kit as well.
As far a your BiPolar "delimma" , also look @ the SIAB for
a DC/DC inverter schemo.   Look @ all the pwrsupply section.
Running monopolar, U will have to match Rs and Cs very closely,
to get a "truely balanced" out.
Easier 2 use a 600/600 xfmr.
W/Bipolar your outputs are centered at ZeroV.
Otherwise U'll have to use coupling caps 2 get rid of DC.
Maybe U can breadboard w/5% Rs and do a test listen(?)
staybalanced
tone

[Paul Perry (Frostwave)]

BTW, I think part of the balanced thing is not necessarily to provide inverted signals on both 'lines' (if you have a look a lot of so-called balanced line drivers only actually provide signal on one 'line') and also the output impedance is important. Mike

It is true that you can drive a balanced cable with unbalanced signals. But, you lose the immunity to noise being picked up by the cable. Which probably isn't a problem if you have a reasonably hot signal, considering also the low impedance driving the cable. If you are just trying to get a guitar signal into a balanced input, the noise question hardly arises.

AS for the question of bipolar supplies, it helps to note that the chip itself can't tell whether the supply is bipolar or not. All it knows, is that there is (or should be!) more volts on the + power input than the - power input, and that the input should be biased to half way between them. How that is done, isn't something the chip worries about.

[mikeb]

It is true that you can drive a balanced cable with unbalanced signals. But, you lose the immunity to noise being picked up by the cable. Which probably isn't a problem if you have a reasonably hot signal, considering also the low impedance driving the cable.

Hmmmm ... well, that isn't what I read here:
http://www.sound.westhost.com/project87.htm#cheat

Mike

[mikeb]

...Mr Tremblay from tonepad...

That would be runoffgroove.com  :wink:

Oops ....

Mike