What do you think of this unbalanced to balanced circuit?

[Svart]

http://groupdiy.twin-x.com/displayimage.php?pos=-314

This is the circuit I use to balance the unbalanced signal going out to the Tape/recorder. I'm unsure of the design and feel that it might be inferior to other topologies.. what do you think?

on another note, I am working on changing the Tape/recorder input circuit as it is simply the differential signal fed to both the INV and NI inputs of a single opamp. I'm looking at a three opamp solution.

any opnions would be helpful.

before it's said and done i bet I end up with a whole new channel strip! :green:

(EDIT the opamp models were tl082s but the parts are really opa2604..

 

[pstamler]

Doesn't look bad to me. I'd be a little leery of the heavy loads you're hanging on the outputs (for meter outs? Headphone outs?), but a 2604 can probably handle them fine.

Peace,
Paul

 

[Svart]

the one 1/4 jack is unbalanced tape out, the other is a balanced elco out. the meter output is later fed to a buffer/meter driver.

what do you think about the diff receiver circuit being just a single opamp with pos. and neg. signals being fed into the inv and ni inputs?

thanks!

 

[bcarso]

"what do you think about the diff receiver circuit being just a single opamp with pos. and neg. signals being fed into the inv and ni inputs? "

I assume you don't mean that literally, but rather the 4-equal-R one-opamp diff amp configuration, right?

 

[Svart]

Well, yes, sorry about being vague. I have read many complaints about various noise issues with this configuration, however I must admit I have not measured it directly in my circuit. IMHO it would be something easy to just layout a daughterboard with a better circuit.. if indeed this one proves to have the problems mentioned.

:thumb:

 

[bcarso]

It's not that bad for line level, unless you are striving for stupendous numbers. And it is simple. Use an op amp with good inherent common mode rejection and use the lowest value R's that you can drive easily---and make sure the output R's of your driver are matched well.

There are a host of other topologies of course, some of which are quieter.

 

[PRR]

The circuit is "drawn backward". OK, a matter of taste and style, but often we draw signal left to right in to out, and power top to bottom.

The "+" input of U1B is unconnected.

There isn't any "diff receiver" there.

> the one 1/4 jack is unbalanced tape out, the other is a balanced elco out.

Did you link the right schematic??? There is only one 1/4" jack shown. The only balanced out is on three pointy rectangles. And if you are feeding home or Tascam "-10dB" inputs, you want a loss of 12dB, not 6dB.

> complaints about various noise issues with this configuration

Not sure what configuration we mean. However, many diff-amps feature ~10K resistors in series with the source, forcing noise resistance to be 10K. Noise voltage is square-root of 1,000/150 = 8 times higher than a 150Ω mike. So you reserve such techniques for places that have a gain of more than 10 up from mike level. That covers about everything after the mike-amp itself. Or looking the other way: noise is around 2uV, so if signal peaks are 0.2V 200mV -12dBu then S/N is 100dB which is generally a non-problem.

For a nickle more, I'd use a 2-amp diff-amp. Noise may be the same, but input Z differential and common-mode is really balanced.

There are topologies without the big input resistors, but resistors allow you to scale oversized inputs, and protect the chips from gross abuse (like plugging into Speaker Out, or a broadcast limiter running +30dBm).

 

[Svart]

Sorry to confuse PRR, the questions was originally about the balanced output(transmitter..) circuit in the schematic, but later I added the question about the input circuit but didn't draw a schematic for it.

the schematic is "backwards" because it was a snippet from a more complete schematic of the channel strip, the rear jacks are to the left, the internal bussing is to the right. ah but that is just a taste judgment.

the unconnected inv input on the second opamp goes to ground. I was in a hurry as usual. :sad:

This circuit was supposed to be +4 output feeding a Tascam MX-2424, +4 inputs.


Again, I am sorry to have confused anyone with trying to squeeze in another question on the same post. I was just trying to save threadcount. The other question I had was about the Input receiver. here is the schematic for the Input section.

http://groupdiy.twin-x.com/displayimage.php?pos=-323

I was just wondering if it was worth the effort and time to replace this circuit..

thanks!

 

[PRR]

> worth the effort and time to replace this circuit.

Does it buzz? Do you plan on running cables through hostile territory?

The hiss-noise is a non-issue. The real-world CMRR of the 1-amp version is not ideal, but generally more than enough for any place you'd want to be recording music.

This Plan is flawed. The difference between +4dBu and -10dBv is 4 (12dB), not 3 as you seem to have. Using 10K feedback with gain of 4 may get you onto embarassingly low input impedance. Also R3 is not changing gain, and you usually have a divider on this input. With the divider, R6 R5 are probably pointless. I'm still confused: if the output runs outside the box, put 100 ohms between it and the jack.

Plagiarize. The world is full of such designs, few of them novel, but most (at least many) of them correct for their cost.

 

[Svart]

that input circuit is NOT mine.. :green:

it's what's actually in my channel strips on my console.

I noticed some of those mentioned problems with the circuit which compelled me to post it and ask for opinions and i see it does indeed need fixing!

thanks PRR, I'll be working on a new input circuit shortly.. I'm sure I'll be pulling pieces from here or there, getting them drunk and hoping they play well together. Might I ask if you had/have something a bit better in mind before I start?

:guinness:

 

[Svart]

i've cobbled together a 3 opamp solution that i think might be a *bit* better than the single.. I'll draw it up and post it in the morning..

:thumb:

 

[Svart]

http://groupdiy.twin-x.com/displayimage.php?pos=-327

ok there are a few different designs I either plagerized or cobbled together. let me know what you think.

:thumb:

 

[bcarso]

#1 is classic but needs input resistors for d.c. reference. #2 "superbal" so called has stability margin issues in my experience but can be made to work and has the advantage of protection from insults to a degree. #3 is o.k., a bit different phase shift for one input vs. the other but of little consequence for audio.

 

[Svart]

thanks Bcarso. #1 was my initial idea to use but found instances of use of the latter two but was unsure of how they worked since I have not built/tested them. As for #1's resistors, do you mean having them in series with the signal or having them shunt to ground as in #3? i'm assuming you meant to ground correct?

updated the file but had a snafu on the old one so here is the new link:

http://groupdiy.twin-x.com/displayimage.php?pos=-329

:thumb:

I think an OPA2604 feeding a 5534 might be in order, or I have a few OPA404s to try out.. I could use the last stage to put a trim on the output..?

 

[PRR]

 

[Svart]

elegant. thanks!

 

[bcarso]

That's a tried and true solution and has the advantage of ~zero common-mode distortion and is pretty bulletproof. The tiny amount of differential phase error shldn't be a concern at audio frequencies.

As well, the likelihood that your source is going to have such high S/N as to be degraded by 20k resistors is small.

 

[Svart]

that's pretty much what I was looking for. it looks good, and uses only a dual opamp. I'll be proto-ing some daughterboards to try with my channelstrips.. I can always play around with values later too.

thanks everyone!

 

[Svart]

i have a panel of these made up and will be stuffing a few to try out. PM me if anyone wants the files for the pcb or a pdf.

 

[Svart]

UPDATE:

these work great! In fact I could hear a difference for sure, even another person( who didn't know about the mod) could hear the difference and mentioned it unsolicited.

thanks all!