Transformer for Balanced to Unbalanced Adapter?

[johnyradio]

Hi

Thanks for your AWESOME website! I have a tech question.

I have a piece of audio gear with balanced outputs, and i need to send it to gear with UNbalanced inputs.

I want to build the balanced to unbalanced adapter from:
http://www.rane.com/note110.html
(scroll down to "The Next Best Right Way To Do It")



Here's my question. Would any of the transformers below work for this purpose?

http://r.ebay.com/CWd0TA
http://r.ebay.com/yRsln2
http://r.ebay.com/Ga5xaI
http://r.ebay.com/k9tGkA
http://r.ebay.com/L0xB4e
http://r.ebay.com/TWNtAW
http://r.ebay.com/BmtLRp
http://r.ebay.com/JAlJO0
http://r.ebay.com/SgrWzs

-Thx for helping a noob!

[Seljer]

The only one that I'd put my money on working properly is the "1x PE-65831 Pulse Audio & Signal " . The others have vague descriptions or are listed as telecommunications/RF transformers or baluns.

[merlinb]

I doubt any of them will work- their impedances are too small. You want a primary impedance of kilohms or tens of kilohms (which implies coil resistances of hundreds of ohms), more like this: http://www.ebay.com/itm/Xicon-Audio-Ring-Modulator-Signal-Transformer-TM018-10K-CT-to-10K-CT-New-/111054503165?pt=LH_DefaultDomain_0&hash=item19db5d3cfd

[johnyradio]

ok, thx

[Seljer]

I doubt any of them will work- their impedances are too small. You want a primary impedance of kilohms or tens of kilohms (which implies coil resistances of hundreds of ohms), more like this: http://www.ebay.com/itm/Xicon-Audio-Ring-Modulator-Signal-Transformer-TM018-10K-CT-to-10K-CT-New-/111054503165?pt=LH_DefaultDomain_0&hash=item19db5d3cfd

Yes, that is more like it. I assumed whoever was putting up the ads didn't know what they were doing and just measured the resistance it with an multimeter and called it 'impledance'. On the one mentioned I just noticed further down "Product: T1/CEPT/ISDN-PRI Transformers" and googling for a datasheet shows it's another telecommunications transformer so no go.

Listen to merlinb

[GibsonGM]

How about just getting a direct box?  They are pretty cheap, and very effective. 

[Seljer]

How about just getting a direct box?  They are pretty cheap, and very effective. 

But make sure its a passive one. Because then it is just a transformer inside and it'll work in reverse!

[johnyradio]

But make sure its a passive one. Because then it is just a transformer inside and it'll work in reverse!

ha, thanks for that bonus info! i was wondering about that.

i'm looking for the least-expensive, DIY method. That's why i'm not going for a direct-box. I've been told elsewhere that a special assembly cable (described in the Rane article) should be fine except for long cable-runs.

thx!

[R.G.]

Let's play technology for a minute.

One of the things experience has taught me that is not obvious from most electronic theory is that there are no circuits in isolation. There is almost always something hooked to the input and something hooked to the output. Those things on input and output not only provide or eat signal, they have impedances that affect the way the circuit I *think* I'm working on works.

This is never more true than when transformers are involved. Transformers seem simple, and in many ways they are. But they have built in frequency filtering that reacts with the source impedance AND the load impedance to make them do odd things if you don't know the details.

The oddnesses of source impedances are also very prominent with the electric guitar. The electric guitar pickup is a very unusual source, from the electronic perspective. It's a *very* high inductance coil with a modest resistance, and built-in high frequency filtering from its self-capacitance.

These concepts all come together to cause very odd interactions when you try to drive a transformer from a guitar pickup. If this is what you want to do, you have issues. IMHO, there aren't any situations where you get clean, no-frequency-loss transmission when you try to drive a transformer from a guitar pickup. Unless you happen to like the built-in filtering, it's going to sound odd.

An electric guitar pickup typically has an inductance of from one to four *Henries*. That's huge. The resistive 4K-18K can usually be neglected. The internal capacitance makes the pickup output start dropping dramatically somewhere between 4kHz and 8kHz depending on the inductance, and the big difference is between single coils and humbuckers.

This is why effects inputs and amps typically provide an input impedance of 1M or over. Less than that causes perceptible treble loss. For a number of reasons, it is impractical to make a signal transformer offer an input impedance of 1M or more over a wide band of frequencies. The winding needs get out of hand. The inductance of the pickup and the inescapable inductance of the transformer also interact to cut bass. It gets impractical to wind a transformer with both good high and low frequency response even with a moderate-impedance drive on the input. It gets even worse when you're trying to do this with a raw guitar input.

The net of all this is that it is very, very, very difficult to do a good job of guitar -> transformer -> something else unless you happen to like the filtering and frequency losses that happen automatically. And transformers that even come close are very difficult to wind, need special materials, and are therefore quite expensive.

It is far, far cheaper and more direct, especially for a DIY setup  to use opamps to do the balanced-to-unbalanced transformation. You can get opamps for a couple of bucks which will run rings around the best transformers you can buy for all considerations except two. These are (1) the need to power the opamps and (2) you don't get galvanic isolation.

Powering the opamps is a trivial annoyance IMHO. You're almost certain to have 9Vdc available to do this. If you don't, 9V power is very cheap to get, an insignificant fraction of the cost of a really good transformer. You can get the isolation by using a transformer.

What? R.G., you just talked me into ditching the transformer. Now you're telling me to use one!

Let me explain. Good response from a transformer driven by a raw guitar signal is HARD. Good response from a low-ratio transformer driven by an amplifier can be simple, easy, and cheap. The amplifier can be set up to minimize or eliminate the issues with the transformer and give you the isolation if you really have to have isolation.

So, in response to your original question, I have a few back to you for you to contemplate:
(1) Do you intend to drive this balanced to unbalanced setup with your guitar straight in?
(2) Do you have to have ground isolation?

[merlinb]

Since the OP only mentioned "audio gear with balanced outputs", and guitars almost never have balanced outputs, I think it's safe to assume he's not talking about a guitar...

am I wrong?

[Derringer]

FWIW, I am REALLY happy with this $25 passive direct box from ART
http://www.sweetwater.com/store/detail/Zdirect?device=c&network=g&matchtype=&gclid=CjgKEAjwzIucBRDzjIz9qMOB3TASJABBIwL1FAG0nmeXTDA80GQwAgjqg2GWp92sSOBvujj0zr94sfD_BwE

I use an transformer laden XLR-to-1/4" adapter from my microphone cable into unbalanced FX boxes and then take the output of the last FX box into the ART direct box to make everything balanced again for the PA mixer.

[johnyradio]

Heh. Merlinb is correct. No guitars involved.

But that's a fair assumption R.G.-- this is DIY Stompboxes, after all! I really appreciate the tutorial. I love in-depth technical answers more than simplistic "you can't do that" answers. Thanks for the detail!

You've shown me that my particular source is a critical factor. It's a rather unusual source: two piezo pickups arranged in a "balanced" configuration.

from http://christian.liljedahl.dk/guides/noise-free-piezo-microphone



i'm looking forward to the response

[merlinb]

It's a rather unusual source: two piezo pickups arranged in a "balanced" configuration.

Ouch. Can you spell "high impedance"? Think you're gonna need TL072s rather than a transformer...

[R.G.]

Yep. The only thing worse for transformer coupling than guitar pickups is piezo pickups. You're going to have to have very - very! - high impedances after the piezos.

I recommend what amounts to a FET input instrumentation amp. National Semiconductor did what they called a "Low Noise Transformerless Balanced Mic Preamp", which used an LM387 low noise mic preamp chip feeding an LF356 differential amplifier. It's a mostly-standard instrumentation preamp setup, but it was intended for low impedance mics. You'd need to redesign it for two very high impedance followers feeding a diffamp in instrumentation amplifier format. But this would get you the balanced to unbalanced transformation without a transformer.

And a transformer by itself simply won't work. Transformers get really hard to design somewhere around a 10K impedance level. I would hate to even try one at the 10M or so you'd need for this.

As a practical matter, you're stuck with an active implementation.

[johnyradio]

you're saying the piezos should be looking at a megohm range impedence?

thx everyone, these answers are awesome and informative.

[R.G.]

Piezo pickups are very - very! - high internal impedance. They're easiest to think of as capacitive voltage sources. 1M is a minimum, and some piezos need 5M - 10M to get decent bass response.

Again I'm making an assumption; this one being that you want full range audio out of this. If you can make do with 40kHz sonar pulses, the problems get smaller.   

Does your piezo pickup manufacturer offer any suggestions?

Note that high resistances have their own problems, one of those being thermal noise. If you're going active (and I recommend that) you can use bootstrapping techniques to give effective impedances in the 10s to 100s of megohms without using actual resistances in those ranges, and save a lot of thermal noise.

[R.G.]

Just saying - the advice would be more practical if you told us what you're doing.

[johnyradio]

Just saying - the advice would be more practical if you told us what you're doing.

would be happy to. What more info would help?

[R.G.]

would be happy to. What more info would help?

Overall description and schemo as it exists now, even rough.

[johnyradio]

Overall description and schemo as it exists now, even rough.

best i can do at this point is my post above, http://www.diystompboxes.com/smfforum/index.php?topic=107342.msg974802#msg974802

Feel free to ask specific questions.

thx!