Which circuit to choose to balance&unbalance in the console "Insert" section?

Hello Diy Family,

I'm looking for a good/reliable circuit to balance and unbalance signal in mixing console "insert" section.

I'm using Trident MTA console which I would like to upgrade with "balanced" patchbay.
Every channel has "Insert" socket where tip is send, ring is return. What I would like to do is

Make a small pcb near each channel. From each channel I would like to balance signal (on that pcb) , send it to the patchbay, then from the patchbay balanced signal goes back to the pcb where I would  unbalance it and send back to the channel.

I found some circuit to balance &unbalance signal,  but maybe you did this and can give me any advice which works better / reliable.

Thank you guys!

THAT corp makes some good IC's I hear.
Check out their Design Notes, specifically 140:

http://www.thatcorp.com/Design_Notes.shtml

Olegarich said:

Hello Diy Family,

I'm looking for a good/reliable circuit to balance and unbalance signal in mixing console "insert" section.

I'm using Trident MTA console which I would like to upgrade with "balanced" patchbay.
Every channel has "Insert" socket where tip is send, ring is return. What I would like to do is

Make a small pcb near each channel. From each channel I would like to balance signal (on that pcb) , send it to the patchbay, then from the patchbay balanced signal goes back to the pcb where I would  unbalance it and send back to the channel.

I found some circuit to balance &unbalance signal,  but maybe you did this and can give me any advice which works better / reliable.

Thank you guys!

Click to expand...

The vast majority of inserts are unbalanced for obvious (cost) reasons. When well engineered the signal integrity coming and going can be quite good.

To convert a TRS insert send to balanced can be as simple as using 3 wires in place of two.  (I like simple.)

To convert a balanced return to single ended is not as simple and a proper balanced receiver might be useful.

Do you know that you have a problem, or do you just have too much money and too much time?  :

JR

It would be helpful to know exactly how the insert send was being driven. I'm assuming it would just be a simple op amp and cap.

With a little modification you can make it impedance balanced. You will not get the differential signal of a normal balanced output but, you will get the advantage of noise (whatever noise the cable picks up along its length) cancelling at the receiving gears balanced input. Schematic below

As for the reciving end. The simplest way to balance it is to use a nice 600:600 ohm transformer. If you have the cash that is. Other wise I second the suggestion for the  THAT Corp line receiver chips.

Thanks guys, “insert” send is driven by ssm2017 ( which works as mic/line amplifier) and 22uf cap with 100 ohm resistor.
I thought about using transformers but for a 40 channels that would be fortune. So probably I’ll go with THAT
Thanks

Hye guys, So I  found 3 different circuits for balancing.
1 based on TL072, from the same console. Second one again from the console circuit.
3.is based on ne5532 and the last one is based around THAT,

so I wonder which one will work better?

I wonder how much THAT sounds better then NE5532 circuit?

To repeat (myself) the send does not require any active electronics (unless you have way too much money and time on your hands. )

I suggested just using a 3 wire cable...  You could add a $0.02 resistor to impedance balance the send that would make it even better (on paper).

Of your examples only the 1646 looks decent (differential outputs are not the same as balanced), but lets not get lost off in the weeds.  IMO not necessary (KISS).

Save your solder (and geld) for a true balanced return.

JR

The first 3 circuits are all the same. The THAT 1646 is going to give you the best performance. But none of these circuits should "sound better" than the others. It boils down to common mode rejection, noise, distortion, drive capability and so on.

IMHO, if you are doing 40 channels, the THAT 1646 is overkill. If you need to drive long cables with a 600 ohm load at the end, then maybe you want the 1646. But for almost all intents and purposes the other circuit is fine. For short connections in non-hostile environments impedance balanced (as someone else pointed out) would probably be better than most consumer grade console sends:



This circuit would give you ground noise rejection which is probably the most common type of noise when hooking up outboard gear (especially digital stuff, stuff with old SMPS, high gain guitar amps and so on).

I do NOT like the first three designs since they won't "play nice" if they are fed into an unbalanced destination load, which is a definite possibility with a patchbay. and a hodge-podge of outboard gear.

As mentioned, "impedance balanced" is the least complicated and costly, then followed by the THAT driver and a transformer.

Bri

Olegarich said:

Hye guys, So I  found 3 different circuits for balancing.
1 based on TL072, from the same console. Second one again from the console circuit.
3.is based on ne5532

Click to expand...

I would never recommend any of these three. They are known to exhibit the so-called "Tascam problem", which makes them not compatible with unbalanced equipment. Note that these circuits have a gain of 6dB, which will require some adjustment.

and the last one is based around THAT,

Click to expand...

I could recommend this one, however, I would probably implement impedance-balance, that involves just a few resistors and caps. It should be noted that impedance-balance does not provide rejection when used with unbalanced equipment.
THAT 1646 does, as well as another alternative,  "remote ground sensing". Check
https://groupdiy.com/index.php?topic=44786.0
Soundcraft 6000, page 3, north-east, they call it ground-cancelling.

I wonder how much THAT sounds better then NE5532 circuit?

Click to expand...

"Sound" is not the issue here. Well-implemented, all these circuits should be transparent. Balancing is about getting rid of hum, buzz and interference. Impedance-balance and THAT1646 give similar performance in that respect.
However, the actual CMRR ( Common Mode Rejection Ratio) of a connection, which is the measurement of performance at rejecting unwanted parasitics, depends onthe combination of sender, cable and receiver.

abbey road d enfer said:

I would never recommend any of these three. They are known to exhibit the so-called "Tascam problem", which makes them not compatible with unbalanced equipment.

Click to expand...

In more recent times, I would also call it the "Digidesign problem".  Bumped into the brain-dead output driver design in what was the top of line Digi 192 interfaces approx. 12 years ago.  I was shocked to discover those interfaces had a "don't connect the minus output leg to ground" proviso.

Bri

Brian Roth said:

In more recent times, I would also call it the "Digidesign problem".  Bumped into the brain-dead output driver design in what was the top of line Digi 192 interfaces approx. 12 years ago.  I was shocked to discover those interfaces had a "don't connect the minus output leg to ground" proviso.

Bri

Click to expand...

Ouch! Some fools never learn...

abbey road d enfer said:

Ouch! Some fools never learn...

Click to expand...

There is a short list of newbie console designer mistakes we often see repeated (by FNGs).

I didn't think that was even on the list.

JR

abbey road d enfer said:

It should be noted that impedance-balance does not provide rejection when used with unbalanced equipment.
THAT 1646 does, as well as another alternative,  "remote ground sensing". Check
https://groupdiy.com/index.php?topic=44786.0
Soundcraft 6000, page 3, north-east, they call it ground-cancelling.

Click to expand...

Is it true that the THAT1646 provides rejection when feeding an unbalanced output ?
I can see from the datasheet and AES paper that it's a version of the 'Cross Coupled Output' ('Transformer Like') output circuit and see that it is claimed (and I assume true) to have better behaviour wrt being driven into a single ended (unbalanced) output where clipping occurs compared to similar parts eg SSM2142.
But I don't see that is has any means of  rejection working into an unbalanced output ?
If I'm missing something then happy to learn.

The 'Ground Cancelling' output does indeed offer rejection by means of addition at the sending end (rather than subtraction at the receiving end).
Apart from the Soundcraft designs, DDA used them. I guess others too - it's not usually clear without looking at the schematics.
But I recall DDA implementation as I was involved in work on the automation side for them back then as subcontract design.
I remember we noted the circuit but could never really 'get' what it did even after Dave Dearden had discussed it briefly.
We weren't 'seeing'  the whole picture with one connection being an input from the receiving end.

Newmarket said:

Is it true that the THAT1646 provides rejection when feeding an unbalanced output ?
I can see from the datasheet and AES paper that it's a version of the 'Cross Coupled Output' ('Transformer Like') output circuit and see that it is claimed (and I assume true) to have better behaviour wrt being driven into a single ended (unbalanced) output where clipping occurs compared to similar parts eg SSM2142.
But I don't see that is has any means of  rejection working into an unbalanced output ?

Click to expand...

Just like the cross-coupled stage from which it is derived, THAT1464 provides "rejection" - actually cancellation - by sensing the parasitics that are fed to the "cold" leg and transposing them, with the same polarity, to the "hot" leg, thus maintaining the integrity of the differential voltage between legs.
As to the "better behaviour", that is because the mechanism that measures current unbalance between legs is more sensitive than that of the standard cross-coupled stage.

abbey road d enfer said:

Just like the cross-coupled stage from which it is derived, THAT1464 provides "rejection" - actually cancellation - by sensing the parasitics that are fed to the "cold" leg and transposing them, with the same polarity, to the "hot" leg, thus maintaining the integrity of the differential voltage between legs.
As to the "better behaviour", that is because the mechanism that measures current unbalance between legs is more sensitive than that of the standard cross-coupled stage.

Click to expand...

Yes - got it. I've gotten 'rusty'  : Thanks.
But you have to 'ground' one of the outputs at the receiving end and not at the transmitting end of the cable. Not a problem generally and same as Ground Cancel output wiring.
Back then - 90s - I used SSM2142 and for unbalanced cabling generally connected 'cold' to '0V' at the SSM2142 end due to warnings of instability / oscillation if remotely connected. But I hear that it's not really a problem in normal circumstances.
Of more concern was the variability of SSM2142 measured performance although they were just about in spec' as AD pointed out when queried.

Newmarket said:

But you have to 'ground' one of the outputs at the receiving end and not at the transmitting end of the cable.

Click to expand...

That's correct. I've always advocated wiring unbalanced connections as if they were balanced. If the equipment is replaced with a proper balanced one, only the connector needs to be changed, not the cable.

Back then - 90s - I used SSM2142 and for unbalanced cabling generally connected 'cold' to '0V' at the SSM2142 end due to warnings of instability / oscillation if remotely connected. But I hear that it's not really a problem in normal circumstances. 

Click to expand...

Indeed it depends on several factors, but SSM2142 can be a little touchy; THAT were aware of this - and several other - issues and made sure they had their product right.

abbey road d enfer said:

That's correct. I've always advocated wiring unbalanced connections as if they were balanced. If the equipment is replaced with a proper balanced one, only the connector needs to be changed, not the cable.
Indeed it depends on several factors, but SSM2142 can be a little touchy; THAT were aware of this - and several other - issues and made sure they had their product right.

Click to expand...

Yeah - I just wire everything balanced now and do any shorting in the relevant plug body. Can be a bit tight with phono connections though.

re SSM2412 - yes - I recall a tube / batch where the test kit measured the output THD+N at around -75dB instead of the usual -89.9dB

There are many unbalanced insert loops working perfectly over modest distance.

You will get most of the benefit from properly wiring (3 wire) the send and return so the audio low is not corrupted by ground/shield currents.

JR

Yeah - I just wire everything balanced now and do any shorting in the relevant plug body. Can be a bit tight with phono connections though.

Click to expand...

In my home setup all the cables are balanced xlr's.  I have a bunch of 6" adaptor leads that I use for unbalanced equipment.  That way if I move anything around the adaptor cable moves with the gear & plugs into the new locations balanced xlr wiring.  Possibly a bit expensive in plugs, but I have been given quite a lot over the years as bunce when I'm working for other people.