DC Coupled, Balanced 1/4 output?

[radardoug]

How is it THE problem?
The basic analysis of CMR is the Wheatstone bridge. Indeed, if the two bottom impedances (the common-mode impedances) were infinite, the top impedances would not matter much, but it's never the case, there are always stray capacitances, and no elecronic circuit has infinite impedance.
The solution that consists in balancing the bridge as well as possible is not a compromise, it's a perfectly valid practical answer to a practical problem.
Thrasformers are not perfect either, their leakage capacitance cannot be perfectly balanced.


Perfect is the word = impossible practically.

That is why I used the word 'perfect'. But in the practical world, transformers have always had advantages, even when not perfect.
When you hooked up a fully transformer balanced Neve in the old days, it always worked, very well. These days, it seems everyone is always bleating about their interconnects.

 

[MisterCMRR]

radardoug, with all due respect, noise cancellation of a circuit using a grounded resistor in the low side to match the output impedance of the ground-referenced signal in the high side isn't a compromise in any way - and a little redrawing of the circuit, plus some thought, will prove it. In the real world, using a transformer or InGenius at the receive end, 90 to 100 dB of CMRR can be achieved (at mains frequencies) with all three of the sources you mention, including the one with a grounded resistor! One thing I sometimes ask folks who think the signal is somehow involved in noise rejection is "if noise rejection depends on if or where a signal is present, why do we take it away and turn up the gain when listening for the unrejected noise?" It usually causes some head-scratching and some silence.

 

[john12ax7]

I've grown to appreciate the simplicity, effectiveness, and yes low cost of the impedance balanced solution for transformerless applications. Less active stages can often be superior in terms of signal integrity. It's not a performance compromise imo.

 

[john12ax7]

On a related side tangent, if you only have space for 1 transformer. Has anyone noticed real world difference with an input transformer vs output transformer?

Essentially is 1 transformer better on the input or output. I see the technical merits of both, but haven't tried enough varied real world situations.

 

[FIX]

Listen to John Roberts. He is one of the smartest people I have ever met. His LOFT test unit was my favorite for many years. I seem to remember that he made a BBD board that I used to make flangers and doublers back in the 70s, and eventually turned them into my FIX SOFTUBE flanger/doubler plugins...

 

[Newmarket]

On a related side tangent, if you only have space for 1 transformer. Has anyone noticed real world difference with an input transformer vs output transformer?

Essentially is 1 transformer better on the input or output. I see the technical merits of both, but haven't tried enough varied real world situations.

As I see it - well I guess it must depend on what is on the non-transformer end - but in any reasonable scenario it's the input that gives greater benefit.
And, similarly, using THAT Ingenious Input devices gives v good performance when fed with an unbalanced signal when wired appropriately.
I have done this to handle unbalanced signals from a TDIF audio interface going to balanced inputs on mixing desk. The 'standard' diff amp balanced inputs didn't cut it in terms of ground noise rejection. The TDIF digital interface has a common ground with the 'earthed' PC and leads to noise issues. But putting a box of InGenious chips for four channels between them gives a large improvement. Fwiw I have another four channels going in via 'budget' OEP / Vigortronix transformers and result is similar. Obviously the may be SNR implications with the extra circuitry involved but overall it's a positive esp wrt rejection of my mains related frequencies and 'digital' related noise.

 

[Newmarket]

I spent 15 years at Peavey trying to make products that customers couldn't misuse, but that turned out to be all but impossible.

Indeed. I still remember the sad and disappointed faces when I pointed out that there was no single 'Universal' and definitive wiring scheme I could provide to give optimum datasheet performance in each and every possible user scenario.
And tbh the potential users were generally ahead of the sales dept understanding at the time.

 

[Newmarket]

As soon as you introduce the resistor on one leg impedance balanced approach...

But the point is that it's not simply a resistor on one leg. The same impedance (admittedly an approximation but a fairly good one and good enough for several mic makers) is on the 'hot' signal too. Hence the two signal lines are 'Impedance Balanced" and that is what matters in terms of how a balanced input reacts to them.
Ok - obviously if you're feeding an unbalanced input then different considerations apply and you will need transformer / electronically cross coupled / Ground Sensing outputs to get any CMMR benefit.
But I was initially thinking that we were discussing 'Balanced to Balanced' connection schemes.

 

[porkyc]

DC coupled audio paths are a recipe for disaster if that DC content gets carried through to outputs and presented to loudspeakers.

Back before some of you were born Crown made a DC coupled power amp (DC 300). It was a notorious speaker eater, if something upstream leaked DC into the audio input path.

JR

But they were the best thing for your 15" drivers in Control Room environment. Eastlake/Westlake Studios.
PC

 

[abbey road d enfer]

On a related side tangent, if you only have space for 1 transformer. Has anyone noticed real world difference with an input transformer vs output transformer?

Essentially is 1 transformer better on the input or output. I see the technical merits of both, but haven't tried enough varied real world situations.

For years, I have produced products where output xfmrs were optional, but the inputs were electronically balanced as standard. It was in a time where the dedicated chips did not exist. Differential inputs were always much easier to put in production than EBOS.
In particular unbalancing them would always result in significant loss of performance (increased THD and rail pollution).
Nothing of that sort happens if you connect an electronically-balanced input to an unbalanced source.
Just make sure that you use 2-wire+shield, whatever the nature of the connection, balanced or not.
Although very different in nature, the same conclusion applies to xfmrs.
Driving a xfmr requires attention to undesirable currents due to stray capacitance, that can pollute the rest of the circuit.
The Jensen load isolator obviates that issue.

 

[JohnRoberts]

Listen to John Roberts. He is one of the smartest people I have ever met. His LOFT test unit was my favorite for many years. I seem to remember that he made a BBD board that I used to make flangers and doublers back in the 70s, and eventually turned them into my FIX SOFTUBE flanger/doubler plugins...

Thanks but save that for my eulogy.

I'm not dead yet, but closer than I was.

JR

PS: Paul Wolff has his own track record of market successes.

 

[MisterCMRR]

It's a fine point and not really worth arguing about but, by definition, the signal on a balanced pair is differential and has no intrinsic relationship to ground (that's a property of the common-mode or CM component of the signal). But, in the real world, it's not desirable to have an infinite CM impedance at both ends - for one thing, such "floating" requires extreme shielding to avoid picking up high common voltages from relatively weak electric fields - it's why open telephone lines had a center-tapped transformer at at least one end. So having a low CM impedance at one end of a balanced line is a good thing. And generally, because the line driver must deal with supplying high currents (to drive cable capacitance at high frequencies), it's simplest to make the driver low impedance for both differential/signal mode and CM. Mics with balanced lines are usually an exception. A dynamic mic has very high CM impedance - much higher than the typical mic preamp input. So the interface can work well either way, it's strictly a practical consideration. But the signal in a balanced system doesn't require ground-referenced signals, it just happens when the line driver is a pair of voltage sources, whether symmetrical or not. Since the op-amp driving the high line is ground referenced in most inexpensive designs, in order to maintain impedance balance, an equal impedance must be used in the low line. Grounded voltage sources (op-amps) are the simplest and cheapest line driver designs. Of course, a balanced driver that "floats" for CM will theoretically be able to overcome the ground-voltage-difference (GVD) issues that plague unbalanced interfaces (provided the cable's low side is grounded only at the destination - an adapter at the driver output defeats the purpose). An undriven low side op-amp or matching grounded resistor obviously can't do that. It can only prevent high (and usually severly distorted) audio current from flowing into a remote ground point (potentially creating crosstalk) or permanently damaging itself. But, as a balanced output, it's as good as any other at driving a balanced input. A transformer, whether real or simulated with something like THAT's 1646, is the preferred solution. Cost, of course, will dictate the choice for many.

 

[MisterCMRR]

If you can only afford or have room for one transformer, you'll always get much more noise rejection from a good Faraday-shielded (the Faraday between windings, not the can the transformer's in!) input transformer! Output transformers with Faraday shields are rare, more expensive, and have lower bandwidth that their non-shielded counterparts. So most good output transformers are bi-filar wound to get very low leakage inductance (which is essential if the output cable ever becomes lengthy). But the ground-noise rejection of output transformers in general is much worse that an input transformer. In practical applications, the main advantage of an output transformer is the ability to drive unbalanced inputs with no problems. Good output transformers (those with good bass performance in particular) tend to be much larger and heavier than input transformers but generally less expensive than input types. Output transformers are relatively easy to build, but the Faraday shields and multi-layer construction of input transformers makes them much more labor intensive to build.

 

[JohnRoberts]

Looking at this with a sharp pencil (what engineers do) output transformers have to pass more power cleanly so need to be larger/heavier (more copper/iron, i.e. more expensive) than input transformers all things equal.

Transformers used to be the only way to make mic preamps that didn't suck (with decent S/N) until the mid 70s or so when silicon got quiet enough for active preamp designs. It took years after that for IC preamps to get that quiet.

I started dropping iron from designs about then (mid/late 70s), but I was always cheap.... helped me fit in at Peavey.

JR

PS: In the fixed install industry they still disproportionately favor transformers, not for sound quality but because transformers suffer fools more graciously (wiring mistakes, et al).

 

[radardoug]

Looking at this with a sharp pencil (what engineers do) output transformers have to pass more power cleanly so need to be larger/heavier (more copper/iron, i.e. more expensive) than input transformers all things equal.

Transformers used to be the only way to make mic preamps that didn't suck (with decent S/N) until the mid 70s or so when silicon got quiet enough for active preamp designs. It took years after that for IC preamps to get that quiet.

I started dropping iron from designs about then (mid/late 70s), but I was always cheap.... helped me fit in at Peavey.

JR

PS: In the fixed install industry they still disproportionately favor transformers, not for sound quality but because transformers suffer fools more graciously (wiring mistakes, et al).

Johns PS says it all. In the old days professional engineers built the installations. Nowadays you have wannabees sitting in their bedroom studios with lots of unbalanced connections, and wondering why they cant get rid of hums and buzzes!

 

[MisterCMRR]

How true! But it's those wannabees that saved Jensen back when, after Deane Jensen's suicide in 1989, Jensen transformers was on the verge of bankruptcy. Deane had never wanted to sell to end users (only OEM console makers) but I decided that, if we put them in little boxes with connectors, car audio fans and audiophiles might buy them. Then we decided to exhibit at the home theater tradeshow CEDIA in 1994 and they were crazy for them - sales were brisk enough to turn the company around - and we designed full line of plug-and-play isolators and other interfaces called "ISO-MAX." That same year, I started teaching seminars to explain ground loops: what they are, how they generate hum and buzz, how to troubleshoot the system, and where to install isolators to cure them. I still do those seminars at AES, CEDIA, and anywhere they'll cover my travel expenses.

 

[Tubetec]

The old classic is the guy who wants a twin amp guitar setup and removes the ground on one amp and relies on the signal cable between the amps , all going well you get away with it , worst comes to worst you become the path of least resistance for a fault current , a simple jack to jack cable with screen connected only at one end and clearly marked with coloured tape could save your life and get rid honk and hum .

Just happened to look up the history of Jensen transformers , wow very much a story of pheonix rising from the ashes Bill.

 

[FIX]

Thanks but save that for my eulogy.

I'm not dead yet, but closer than I was.

JR

PS: Paul Wolff has his own track record of market successes.

Thanks! One huge thing I learned from you is reducing the number of parts I use. I currently have my own SMD machine and oven, and I have one DC filter value, one input value, one output value, etc. If I need a 301K resister, I use a 200K and a 100K... except those audio ladder volume controls. They use every odd value that exists...

 

[JohnRoberts]

Back in my kit business days I tried to standardize on a few common capacitor values and used different value resistors to make them work in different applications (TH resistors were cheap). For SMD designs the efficient purchase quantity for resistors is a whole reel holding thousands of parts, so it can make sense to use series/parallel to avoid bringing in additional reels of components for a new value or two, but there is an expense in using more real estate and more pops (machine operations) than needed. For general purpose resistor applications I like quad resistors. I recall trying to use x8 resistors in a friend's design years ago and they were too fine pitch for easy serviceability. For critical audio paths larger single SMD resistors work better.

JR

 

[abbey road d enfer]

Standardizing resistors is not a big concern, since they cost pennies. 5 years ago I have bought an assortment of 100 each 120 different values metal film resistors. I had to reorder only a few values since. Whatever value I need to reorder, I buy 1000.
Now electrolytic capacitors, I use only 3 values: 100uF/35 for both coupling and decoupling, 470uF/35 for PSU's, and 100uF/63 for phantom power. When I want a smaller value, I use 1u, 2.2u, 4.7u and 10uF MLCC, that I buy by hundreds. Much better than electrolytics.
I stock about a 1000 of 0.1uF/50 for rail decoupling.
I stock E6 series of film capacitors, again in 100's. When used in selective circuits (filters, EQ...) I always use pairs. I've never met a case where a combo of 2 E6 values would not come close enough to the target value.
I don't do SMD's since I make only prototypes and one-offs.