which transformer for unbal. instrument out (from guitar Fx) to bal. line in?

[iandy4]

Hi Guys,

I recently got an excellent Otari reel to reel and after several recording tests I want to run my roland space echo (stereo outs) straight to the inputs of the R2R.  The output of my space echo pedal is unbalanced of course and the input of the R2R is expecting line level, balanced.  

If I want to get that hooked up, which transformer should I use?  I understand how the conversion happens from balanced to unbalanced and visa versa when using transformers and I have 10k:10k transformers which are center tapped but I don't really understand enough about transformers to know if 10k:10k are appropriate values.  I assume 10k is the load that must be driven by, in this case, the space echo output (or a buffer if I must add one to drive the 10k primary).

I assume the best way to do this would be to have the stereo outputs from my echo pedal go straight into a buffer (for each channel) then to the primary outer lugs of the transformers (one for each channel).  The secondary transformer lugs then go to the appropriate xlr connectors which run to the inputs of the R2R.  Is this correct to assume?

I am interested in getting the best sound possible so links to information and/or schematics about this kind of interfacing from bal. and unbal. is definitely welcome.  Rather than whip something up quickly I'd like to understand what I'm building and learn how to make a professional interface.

P.S. I have a few Xicon TM018's and two Edcor 10k:10k center tapped tranformers (all 10k:10k w/ ct) (the Edcor tranny's are super quality.. or at least cost a lot! So I'd love to find a way to use them).

[iandy4]

I should mention the otari r2r has switches in the back for setting the level: off, 0db and -20db for each channel.  If I set it to -20db and then use just the 10k:10k transformer without a buffer for both channels, would that be a quality solution?

[haroldjenkins]

As far as the transformer is concerned, wouldn't an lower output impedance be in order to go to a line in? If so, would a transformer like this http://edcorusa.com/products/160-xsm10k-600.aspx work? Someone can correct me if I'm wrong. The 10k input seems about right for the guitar's output.
I'm no expert on transformers, but it seems like Edcor's prices are pretty reasonable. Especially when compared to others like Hammond. And they're American-made (wow).

[R.G.]

It's quite difficult to use a passive transformer to match a guitar to a balanced 600 ohm input without losing lots of treble.

Most passive guitar pickups need to see an impedance that's bigger than about 1M or the treble gets lost selectively because of the inductive nature of the pickups. A 10K transformer input is problematic.

If you go active and buffer the guitar, you can use almost any good audio transformer with a CT secondary to do the matching. The impedance seen by the secondary of the transformer is not the controlling issue.

Edcor does make good transformers.

[iandy4]

Alright! Then I'll go with an 072 based buffer right at the output of the space echo, then to the edcore transformers and then I have the long run as balanced xlr's to the R2R.

Does this look ok to everyone?  Anything I should add or change?  Are C1 and C3 redundant as coupling caps because they are directly followed by the transformers?



Hey R.G.,  I started reading semi conductor circuit approximations which I think was your suggestion.  Having to reread this over and over but it's great.

[iandy4]

Just bread-boarded it without C1 and C3 and it sounds great.  Seems like I got a lot of clean headroom even without a grounded metal enclosure around it so I imagine it'll work pretty well when I build it in to one.  If anybody sees an improvement to the circuit, advice and suggestions are very welcome!

[R.G.]

Only one issue I can think of (especially since you've already made it and it works fine!  ) and that is with C1 and C3.

Those values work fine, but if you use polarized caps in a NON polar environment, the signal voltage reverse polarizes them half the time. This can lead to faster degradation over time than they'd otherwise do, depending on how big the signals are. Every volt-microsecond of reversed voltage, even from signal, eats a little bit of the aluminum oxide off the insulating layer.

So whip in Non Polar (NP or bipolar, BP) types for those two.

And C1 / C3 are not really optional. True, it works without them. But those opamps do have some DC offset, and if it runs right into the transformers, it puts in a DC offset with various low-level ill or at least not-good effects. Using a series cap with transformers is good if you can use a big enough cap and not have the side effects of using the cap overwhelm the good it does you by breaking the DC path.

This stuff is obviously second- and third-order stuff. You got the basics. Good work.

[Earthscum]

Question... D.I. should be lower impedance out, correct? 10k:10k would reflect a heavier load to the amp if I were driving to a mixing board? So, if I wanted to reduce load to the amp, a 10k:600R would be a better choice, correct? I'm just curious... I need to put DI out on my pedalboard, and I only find bits and pieces about it here and there. The rest of the time it seems that some mystical soul of a Beringer sales rep possesses the boards whenever the question is raised about DIY DI's. 

[R.G.]

Question... D.I. should be lower impedance out, correct? 10k:10k would reflect a heavier load to the amp if I were driving to a mixing board? So, if I wanted to reduce load to the amp, a 10k:600R would be a better choice, correct? I'm just curious...

If you drive a 600 ohm input on a mixing board you get one of two kinds of inputs. One is really  about 600 ohms resistive, another says 600 ohms, but is really higher impedance. If you're driving the real 600 ohms input, the 600 ohms is reflected at the impedance ratio. So the buffer sees 600 ohms for a 10K:10K transformer and 10K for a 10K:600. So yes, there is a heavier load to the buffer for a 10K:10K.

That may or may not matter.

Almost all opamps are fine driving loads down to about 2K with no substantial degradation from the 10K or 100K loading case. Some opamps, notably the NE5532, LM833, and others are specified for driving loads down to 600 ohms for their full output swing (which may be +/-15V). Other opamps are only specified for driving 2K to full output swing, but may do fine into 600 ohms for smaller outputs, like under a few volts peak. The TL072 is in this group, as are many others. The opamp may not care much whether it gets a 10K load or a 600 ohm load in terms of loading.

Notice that this is the load presented to the opamp. The opamp's driving impedance is already very low as long as you don't drive it into clipping. It's output below clipping may be as low as a few ohms, and that's the impedance that drives the transformer primary and is seen on the lines leading to the board input. Using a stepdown transformer will reduce this even further, but the whole point of 600 ohms is that it's already low enough. So using a transformer to reduce driving impedance from 10 ohms to 0.6 ohms doesn't buy you much in line driving ability.

The 10K:600 transformer steps the signal voltage down though. A 10K:600 transformer causes a 16.7 to one voltage drop on the 600 ohm output compared to the 10K input. If you're already at a low signal level, like from guitar, your 100mV nominal signal will come out at 6 millivolts in the mixing board.

As to which is better, again you have to say what "better" is. I don't think a 10K:600 is needed for low impedance purposes. I'd be happy with 10K:10K; Using a 2K:2K would probably be optimal in some senses, and I'd use a 1K:1K, or 600:600. But I wouldn't want to lose signal voltage for noise reasons.

[Earthscum]

Damn... that cleared up a bunch of stuff for me. I had the info all around, just hard to piece together sometimes. In one pedal I was going to run the NE5532 for DI (that one will have Phantom power availability, similar to the design at Elliot Sound), but it runs on 9V. You'd run out of decent signal pretty fast with the 10k:600R.

The one I was most concerned with is my pedalboard's DI... I'm running +/-15V, which (according to datasheets) is too high for the 5532, but that's fine. I was going to run TL072's all the way through it, but wasn't quite clear on how much loading is actually presented. I could've just copied the DI out from some amp schematic, but I don't get to learn anything that way. 

Thanks much.

[R.G.]

I highly recommend the LM833. It was designed for audio, takes large power supplies, and will drive a 600 ohm load to nearly full power. And it plugs into the same socket as the TL072 or other dual opamp.

[a soBer Newt]

Check out the DRV134 by TI may suit your needs