Balanced Interfaces-choose your poison

[rickcr42]

a very nice 7-Part article on the balanced interface that is a must read for anyone contemplating either balanced lines from device to device or a mixed SE/BAL scheme :

www.audiodesignline.com/howto/196603651

HINT : click "print version" which loads each article onto a single page for way easier reading then if you want to save the article use this free pdf converter to save the file as a pdf to your desktop

www.pdfforge.org/products/pdfcreator


Where the "choose your poison comes in ? Bill Whitlock is the common denominator for both Jensen Transformers :

www.jensen-transformers.com/apps_wp.html

and THAT Corp :

www.thatcorp.com/

so if you want to go with the best but more expensive option go with the Jensens (or CineMags ) or if you are transformer phobic or can't justify the cost of good iron there is the "synthetic" iron from THAT.

Even those having zero need for any balanced connections in their systems should give the article a read just for filing in the memory banks (your brain ;D ) as part of your audo library.

[spiug31]

There is a talk by this same person at www.aes.org/sections/uk/meetings/a0705.html
The audio (mp3) is good, very clear voice with no background troubles.
It's reassuring to hear the professionals can record well

p.s. it's interesting

[spiug31]

p.p.s. I don't know which forum to post this in but I found several interesting mp3 talks through looking at the links on this page. www.aes.org/sections/uk/meetings/past.html

enjoy

[rickcr42]

yeah,this is as good a place to park them as any considering the "grotto" is pretty much an audio transformer "free" zone and no one will visit this thread anyway but all good man 'cause those in the know don't need to be convinced anyway and those who are trfos-phobic never will be

[spiug31]

[toad]

rickcr42 said:

yeah,this is as good a place to park them as any considering the "grotto" is pretty much an audio transformer "free" zone and no one will visit this thread anyway

Move along please. Nothing to see here ;D

[rickcr42]

Exactly.

You make my point better than I ever could about convincing and closed minds though I think many would be surprised when they read articles by their solid state or bust gurus at one time or another stating the high quality audio coupling transformer being the best choice unless cost is a consideration (and why we even have "simulated" balancing tranformers and "simulated" inductors and NOT because the modern alternative is better) which it always is in a commercil product but should be no detriment in a DIY project where many are willing to spend $50 per on whatever cap happens to be in current fashion which they then mate to a $4 high bandwidth op-amp with zero RF protection and a noisy $5 low bandwidth power supply and think they actually acheived something worthy of showing off.

Yup.Plug and play opamps are OBVIOUSLY better for interstage/inter-device coupling than nasty ass ancient technology iron in the signal path could EVER be and any thinking human KNOWS it (psst ! Better tell Jeff Rowland and all those very expensive high-end TVC users they are making a HUGE mistake ) plus there is all the fun of "op amp rolling" even IF most have no business being in the circuit and are no improvement but a step back but hey,I'm cool with choices unlike many others who have a problem with it

[Deleted]

Rick
Transformers are no longer practical, are too expensive, and difficult to get suitable custom types.
Below is part of a couple of VERY recent emails from a friend
The "Ravens" are very expensive tweeters.
Alex


I was trying to get some hq transformers for the Ravens but Antrim said they didn't have core material that had low loss out to >50KHz.

BTW, I have sent an e-mail to Antrim in the UK asking if they can build a couple of custom transformers to transform the Raven impedance from 2.6 to 5.5 ohms. I would very much like to use them on their own and if possible to mount them into the baffle. I read in a forum that the drive impedance needs to be very low for ultimate THD.

[rickcr42]

Rick
Transformers are no longer practical, are too expensive, and difficult to get suitable custom types.
Below is part of a couple of VERY recent emails from a friend
The "Ravens" are very expensive tweeters.
Alex

100% not true.

by what measure are they no longer "practical" or too expensive for DIY projects ?

granted you will never see one in any mid-level audio and if you do will be crap part to save on cost but for me personally $60 or so per channel is not that big of a deal for a quality part that sounds better than any active alternative i have yet heard plus when you add in the requirement for a power supply to FEED any active design the cost incurred up front becomes less overall than it looks at first blush.

that your buddy can't find a proper transformer for his Ravens has zero to do with using a line level audio tranformer for stage to stage coupling purposes or for phase spliiting (single ended to balanced operation) or for equipement matching (step up,step down,impednce reducing,impedance increasing,polarity inversion).

Yes there are a few pitfalls such as not paying attention to ACTUAL impednces and or cable capacitance in some cases and yes you are limiting your system bandwidth (around 8-100khz) to the point where getting a square wave on the scope hopeless unless you compensate the output (and then remove it for music just like many manufacturers of audio gear do to deceive you) but used correctly allow more ACTUAL front end information to pass through without picking up artifacts as many/most active circuits will and with a linearity most SS circuits WISH they had on a GOOD day.

not trying to convince you dude but to not explain my side if even in my fumbling way would be a disservice to those sitting on the fence.


Best DAC Mod-10K/10K or 600 Ohm/600 Ohm if you have the "drive")1:1 line bridging transformer
Best CDP mod-10K/600 Ohm 4:1 Step down transformer
Best sound Card Mod-10K/10K line bridging trnsformer
Best Phase Inversion control-Again the line bridging transformer with a DPDT "phase" switch on the primary side
Best "Driver" for converting a Stereo amp into a Mono-Balanced-Bridge amp-Line transformer with the single ended input going to the primary then converted to a "+" and "-" signa at the secondary

Obviously some of the above "best" are peronal opinion but others undeniable and the information readily available to any looking.i use what works best in each position up to my skill/knowledge level so am not shy about using solid state where it shines (current delivery/impedance matching),vaccum tubes where they shine (voltage gain),monolithics where their strengths far outweigh any weakneses (high current line level buffers for driving long lines or headphones),discrets,analog sources,digital sources,you name it dude I use it with the above choices made not arbitrarily or because i like the idea but because to me and for my musical sensibilities the signal path iron is LESS of a sonic butcher than the active alternative in many cases and since ALL THINGS between the source and the speakers will butcher sonics to some degree i try to minimize the damage and whatever there is have it tend to the music rather than the specs even though modern audio transformers rival modern actives in almost every category that counts unless you are a ben counter looking to out spec the competition rather than a designer of music machines

[Deleted]

Obviously some of the above "best" are peronal opinion but others undeniable and the information readily available to any looking

Rick
Again, in your opinion.
MOST of the published claims for transformer superiority comes from the remaining few vested interests, and is usually stuff dredged up from the last century ! Low output MC stepup transformers, MAY be an exception, as are decent SPDIF transformers.
Unfortunately, neither of us is ever going to convince the other, because of being on different continents, which makes an AB of the different methods, at least as applied to DACs, impossible.
As far as I am concerned, the best approach by far, is to use very high quality analogue PSUs, thus bypassing the majority of problems that you are trying to bypass..

Alex

[rickcr42]

Rick Again, in your opinion

I said SOME was "in my opinion" and SOME was and is undeniable as it is and the evidence real so to say ALL the above is opinion only would be wrong because it is not

MOST of the published claims for transformer superiority comes from the remaining few vested interests, and is usually stuff dredged up from the last century

again 100% untrue and so not worth commenting on

Low output MC stepup transformers, MAY be an exception, as are decent SPDIF transformers.

So ultra-low noise passive gain is "maybe OK" but 1:1 "zero gain" bridging is not ? Careful dude.Argument is getting weaker by the post

Unfortunately, neither of us is ever going to convince the other, because of being on different continents, which makes an AB of the different methods, at least as applied to DACs, impossible

the difference being i HAVE tried both ways and made a decision on the merits.you have already arbitrarily decided a thing is so just because with no actual personal experience to back up the claim.The end medium is about the music NOT looking at graphs and if it was we would all be using TL072CP op amps and calling them perfect

As far as I am concerned, the best approach by far, is to use very high quality analogue PSUs, thus bypassing the majority of problems that you are trying to bypass..

Alex

Since most will conced every single article added to an audio signal path degrades the signal (from the lowliest resistor to the ultimate in active ciruitry) in some way,up to and including a single wire,I can not fathom how someone can possibly say no less than THREE potential sources of sonic degradation are preferable to one all else being equal.Those three ?

1-AC Mains,the juice that feeds everything that follows
2-DC Power feeding the actual circuit (and part of the signal path)
3-the circuit,the actual active signal amplifying stage

each can and will impart its set of requirements on the design making EACH while part of the whole needing individul attention and get one wrong ?

The whole deal goes down in flames.

so you may be perfectly happy with your methods and that is cool,you are entitled,but to say that is the BETTER way or the ONLY way when you have not even taken the time to compare results is just not a valid point being no more than an opinion based on nothing other than "I like this better because I do"

[hypercatabol]

I got lost a few sections in but get the principle - noise rejection and more juice! Sounds good - why doesn't every designer do it in everything ? Is it really more expensive to implement properly?

[rickcr42]

Apr 19, 2008 14:04:11 GMT @hypercatabol said:

I got lost a few sections in but get the principle - noise rejection and more juice! Sounds good - why doesn't every designer do it in everything ? Is it really more expensive to implement properly?

Yes unfortunately.Most "active" options can be realised for $10 or less for 2-Channel stereo and that includes the active device performing the balancing/unbalancing duties,any possible local regulators plus any bypass/coupling/decoupling passives unless they are Boutique/Designer/Audiophile grade parts and then it can become a wash if the parts are expensive enough-many are,especially capacitors though it is accumulative cost there not being too many $70/per capacitors though if there were would STILL be justified as better as long as they are attached to an active stage.

If the unit is a "stand alone" device the cost savings are reduced to zero because NOW you not only have the active stage but must have a power supply to power it as well and that means both an AC and a DC section with the most cost going to the power transformer so we enter into the personal choice arena rather profits and as you can tell from this thread alone is one that has not a lot of middle ground being extremely pro or con for the favored method.

Add to that the "tweak Factor" where many (to me WAY too many but just my opinion) buy a thing then before the damn thing is even given a chance to have any real evaluation period set out to "improve" on it by swapping out parts for perceived better ones and if they can't screw with it then they don't want it even if they have no clue what they are doing so mod as much by how pretty a part looks in a device with other "pretty" and color co-ordinated parts even if the wrong choice if better sonics are the goal

To them having a simple passive one or two parts in the signal path device is unthinkable because where is the fun in that ? where is the challenge ? How can we take pics and post the porn of a stupid CAN with wires coming out of it ?

Talk about boring !

[hypercatabol]

The second point is a great one - regardless of whether something's single-ended or not, why aren't people just looking for good, thoughtfully designed products (stock) instead of believing that changing caps and resistors is going to transform it to the point where all the money they'd spent modding couldn't have been reinvested in getting a more capable unit upfront! The need to tweak passive components is sometimes just like not seeing the wood for the trees, IMO. Another amusing thing to note about the obsessive hobbyist mentality is that it seldom extends to any desire to improve the power supply circuitry!

This makes me wonder; is it possible to design a serious, fully discrete analogue stage that would outperform another using ICs for the same budget (assuming a single-ended design for simplicity)? If that's the case, then again, why don't more designers do that! What would be your take on that (discrete vs opamp)?

[rickcr42]

Apr 20, 2008 12:03:29 GMT @hypercatabol said:

The second point is a great one - regardless of whether something's single-ended or not, why aren't people just looking for good, thoughtfully designed products (stock) instead of believing that changing caps and resistors is going to transform it to the point where all the money they'd spent modding couldn't have been reinvested in getting a more capable unit upfront! The need to tweak passive components is sometimes just like not seeing the wood for the trees, IMO. Another amusing thing to note about the obsessive hobbyist mentality is that it seldom extends to any desire to improve the power supply circuitry!

This makes me wonder; is it possible to design a serious, fully discrete analogue stage that would outperform another using ICs for the same budget (assuming a single-ended design for simplicity)? If that's the case, then again, why don't more designers do that! What would be your take on that (discrete vs opamp)?

Discrete designs from scratch are beyond the skill level of most DIYers and why in my opinion many gravitate towards the "Plug-and-play" monolithic op amp where for pennies on the dollar anyone can change the sonics of their gear (not always for the better just different ).

There are several well documented designs available for the hobbyist from the "you choose it your way" not quite start to finsh design by Nelson Pass,the original "for hi-fi" Jensen JT-990 that you can obtain the plans for just by sending an email to Jensen for or better yet purchase fully assembled in a standard "DOA Module" plugin (Discrete Op Amps are way common in Pro Gear though many are intentionally "colored" for mic preamp use ) From Mark Hardy for a mere $70 per (damn fine value it being aimed at Pros and not whacko audiophiles who WANT to pay a lot or they dismiss it) or the Forsell JFET versions from the DIY to fully build module,again in the universal DOA pinout module.

There are also plans online for just about every commercial DOA ever designed but again,many for the pro arena and intentionally colored.I can provide links to any/all if needed

Here is the rub though.Some/most/many are actually designed to be the active stage BETWEEN two audio signal transformers (though they can be used naked) to perform the actual balanced to single-ended/DOA/600 ohm Line Driving trafos topology so again it comes down to options and taste-as it should be-and not a take it all or leave it solution.

And finally the main thing with discretes is the power supply becomes THE critical part due to the PSRR of most discrete circuits being an order of magnitude lower than their monolithic cousins and another reason most have no choice but to go to the "grab an opamp and plug it in then add pretty passives" option.They get a taste of being a designer with the reality being they are no more than glorified parts changers (me incvluded mostly ).Where this could get REAL interesting is if the "opamp rolling" extended to the DOA modules where you in effect have a plug-and-play DISCRETE circuit.It would be more expensive but has great potential IF you know what you actually want to accomplish and take into consideration the entire circuit path (input/output loads and drive)

Many paths leading to the same destination with some a quick burn down the highway (monolithics) and others the long winding scenic route (discretes) and even more extreme the overnighter 'cause you are in no hurry (optimised passive/active )

Rick out

[rickcr42]

heh.Not just getting blind in my old age but it seems "dimmer" as well since the post was on the simple single-ended class-A option (and SE MUST be in class-A by nature) and I replied with DOA options which are obviously differential designs with feedback,Sooooooooo...............

SE topology is for my taste the most satisfying be it solid state or tubed but again there are some "essential" points that you need to be aware of if the design is to work in the app.

first off we are working our way DOWN the ladder of PSRR with balanced topology substrate level precision matched device monolithics being the KING in rejection,differential discretes next and single ended designs dead last so while the actual amplifying circuit will be a simple in nature as it gets the complexity of the power supply increases in direct relationship and so is where the majority of the cost to build will be targeted.

the second important "note" is because you have ZERO ying/yang cancellation at the output the SE design having only a single device in the signal path (unless a cascade design which still has the same requirements as the single device) you WILL have DC at the output and that DC typically the power supply voltage (i.e. a 24 VDC power supply means a 24 VDC output offset !) so direct coupling not an option.

Ways to get rid of the DC are :

1-coupling cap matched to the input impedance of the next stage
2-C/L coupling-that is a capacitor and a non-gapped coupling transformer (moist line level trafos can NOT have DC on the input)
3-Gapped coupling/matching transformer

option #1 means a largish cap for low impednces (headphones,600 ohm lines,etc),Option #2 the same for a 1:1 bridging trafos or smaller if the trafos is a step-down type for example one with a 10K primary impedance and 600 ohm or lower secondary with the 10K being the capacitor "load" though be aware this means you will ALSO have an OUTPUT VOLTAGE reduction as well (4-1 for the above) meaning you must be aware of how much gain you have available and if too low for proper drive levels you are back to the 2-stage cascade type designs (single ended voltage gain stage/single ended current drive output stage) or if single stage one that provides both voltage and current gain.

Option #3 can be either of the above depending on the primary/secondary ratio but totally eliminates the need for a signal path coupling cap though it WILL be both more expensive for the part (trafos) AND will be quite a bit larger and heavier so will mean a bigger chassis footprint for the device on the rack/shelf.Choices man.All about choices

[hypercatabol]

Whew. Gonna take me a long time to digest this - in a nutshell...we'll always want good power, and move towards discrete and balanced depending on funds.

Money, eh? So limiting! Typical.

[rickcr42]

Apr 20, 2008 22:42:23 GMT @hypercatabol said:

Whew. Gonna take me a long time to digest this - in a nutshell...we'll always want good power, and move towards discrete and balanced depending on funds.

Money, eh? So limiting! Typical.

yeah,sucks.But better to spend where it counts-performance-than to have most go to a pretty chassis with marginal/iffy circuitry.

Seems to me you could charge a damn lot of loot for a simple opamp based circuit as long as you have the pretty innards and heavy metal shell to justify the price which tells me most have forgotten why we even bother-the music...........

[hypercatabol]

Isn't that why most "high-end" commercial gear come in heavy, shiny boxes to match their price tags / owners' penile substitutes? More weight, more shiny!

[rickcr42]

the second important "note" is because you have ZERO ying/yang cancellation at the output the SE design having only a single device in the signal path (unless a cascade design which still has the same requirements as the single device) you WILL have DC at the output and that DC typically the power supply voltage (i.e. a 24 VDC power supply means a 24 VDC output offset !) so direct coupling not an option.

Oopsy

just got a PM from SandyK inquiring as to if I meant 24 or 12 volts DC which was cool 'cause he COULD have made me look like an idiot by asking openly (instead I am taking care of that detail ).Obviously the number would be 1/2 Vcc which is the Class-A bias point so a mental slip.Still damn high though ;D

[Deleted]

Hey, Rick.
I only deliberately give XTRProf a hard time, because he keeps pushing Zen over (hopefully) sound technical methods.
Leo and I could do with some help from the Top Gun, though!
(Actually, the Singapore Stallion is a pretty good sport.)

Alex

[rickcr42]

Apr 21, 2008 5:08:05 GMT @hypercatabol said:

Isn't that why most "high-end" commercial gear come in heavy, shiny boxes to match their price tags / owners' penile substitutes? More weight, more shiny!

Absolutely but to be honest it is OUR fault as consumers and NOT the manufacturers who cater to our needs.If I had a dollar for every post or revue that said "sounded nice but couldn't they have made it better looking ?It lloks more like a DIY project than a commercial amp" then a year later you see "V2" of the very same product at double the price flying off the shelves with the only change being the cosmetics !

Look at the products by AVA or Bryston or Hagerman gear for example.SERIOUSLY fine circuitry but since it is both affordable and utilitarian pretty much dismissed as serious high end equipment except by those looking for performance over pretty.

Off the top of my head while still decompressing from a three week no day off even with a weeks vacion (vacations are more tiring than work !) I can't immediately think of examples other than the original Slee Solo headphone amp.The original was about as basic looking as it gets being a box with jacks.Reviews were good yet the "chatter" in the forums pretty much zero until the amp was "repackaged" to be more inline with what folks look for in audio gear and then what ?

Started showing up all over the planet at various audio dealers it now being considered a "serious" effort.If most think hard enough they will come up with examples of dismissed gear until made both better looking and more expensive yet pretty much the same internally and THAT is the sad part.

we claim we are all about getting the best performance from our systems but the reality is the majority would take lesser sound as look as the ensemble looked good so they can show off the pics (how many forum gear reviews go to "give us the pron ! We want the pics !" with nary a "could you give more detail on the sonics" ?).

Manufacturers are in it for the loot (as they should be) so are more than happy to wrap $50 electronics in a $500 chassis and even DIYers willing to spend $100 for laser etched panels to slap on a $20 chassis box hiding $50 in parts so it is the "audio as eye candy" consumer setting the market.

My own crap is mostly utilitarian "form follows function" and so would be considered very boring or even amaturish if posted.

I use mostly stock "off the shelf" boxes/chassis then paint.My idea of fancy would be adding a wood end cap or a brass control panel sub-panel strip.My efforts go to the internals and spending the loot on getting (for my taste) good sound and long term reliability.

For example my Line controller,a multiple input/multiple output passive volume and balance control (not passive preamp which is not an accurate wording ) has in the vicinity of $300 internally-heavy duty silver contact SEALED four deck rotary and 6T4P toggle switches,$10 per PEC pots,$70 per CineMage line input transformers,heavy gold plate isolated RCA jacks,silver contact TRS and XLR jacks-in a $20 hammond chassis with the only cosmetic touch being black wrinkle paint and a satin black painted top and bottom plate (may be heading to a full wrap walnut cabinet at some point ).

"As IS" I would have to sell identical units for $650-$700 just to cover labor (wiring three 4-deck rotaries,five toggles and two transformers with #24 solid wire is no joke) and make a small profit.The SAME EXACT DEVICE in an Audiophile Approved wrapper would go over the $K mark and would gain zero in performance

Sitting alongside is my Naked buffer which is a simple 5x6x3 "utility box" from Radio Shack (purchased a bunch when they discontinued them ),the ones with the aluminum chassis and vented steel tops which for me are PERFECT for anything that produces heat such as my buffers and low voltage power supplies with no more "pretty" than painting the aluminum satin black and that if seen as a "product" would be laughed at even though for my taste blows the balls off of most so called "line drivess" or "headphone drivers".Cost to build sans external PS (but internally regulated) is in the $100 range total and because of would be say $250 or so to sell but again,adding a pretty wrapper and with NO internal changes we are looking at $500-$600 just to have a shot !

MY efforts are internal and because of means I concentrate only on the circuits and THEN the box and in this case even the BOX efforts are mostly not visible to the naked eye being all about chassis damping .adding additional "mass" so the heavy cables don't have the tiny box walking all over the rack from cable weight and finally front to back panel stiffeners/dampers stopping any "flex" when plugging in.

If I ever tried to make a buch off my designs I would have to also "repack" the electronics then back down on parts quality or the device would be sooooo expensive only a few would be able to afford them even though I would maybe get reviewed well and that is just so wrong it blows my mind to think on it....

[hypercatabol]

/puts down the popcorn
/applaud
/salute

It's an interesting pathology, though, the conflation of use-value with the fetish for aesthetics as value. It's the same in visual arts - whither the authentic intentions and spirit of Bauhaus in modern advertising - all surface, no feeling.

[rickcr42]

Apr 21, 2008 23:44:11 GMT @hypercatabol said:

/puts down the popcorn
/applaud
/salute

It's an interesting pathology, though, the conflation of use-value with the fetish for aesthetics as value. It's the same in visual arts - whither the authentic intentions and spirit of Bauhaus in modern advertising - all surface, no feeling.

all true but being an AURAL medium once you close your eyes and listen all that pretty dressing is stripped away leaving the music.

I find it interesting that domestic loudspeaker based systems being to the front of listener and with the majority of the image a phantom center almost FORCES a person to focus visually on the point between the stereo pair unless the lights are out (best way to lose the room),car audio comes from directions that make no sense in the real world of performance music yet is enjoyable and headphones being more "between your ears" most likely to be listened to either eyes closed or while doing something else but again not even close to any REAL performance in space yet all are enjoyable.

Why ?

Because music is an emotional art form (when done right anyway)


You FEEL in addition to hearing and why even the lowliest boom box or pocket radio or even TV speakers can produce a sound that though lacking in any defination of Hi-Fi as we define it you may find yourself singing along to or tapping your toe in time or maybe even dancing to the kitchen for a beveredge without even knowing you are doing it

Ever notice that NO two sound systems sound the same yet you can identify the song ? That the same performance on a shitty table radio or TV speakers or car system of boom box or domestic playback system or someone elses system is recognizable and even enjoyable even though it sounds TOTALLY different on each ?

The difference is we are LISTENING not LOOKING so it matters not what the source looks like until we get to high end audio and then our brains seem to say "if it looks good it sounds good and if it looks like shit is is" even though that very same song on a POS AM radio had you singing an hour earlier.

Maybe the old timers had it right when they hid everything behind a cabinet/console door so rather than LOOK at their music they listened to it thus getting the full measure of

[Deleted]

Maybe the old timers had it right when they hid everything behind a cabinet/console door so rather than LOOK at their music they listened to it thus getting the full measure of

They had to ! The woman of the house didn't want to see that bloody big chassis with TRF radio tuner using mechanically coupled tuning capacitors, and several huge (potted ?) transformers, as well as a few big triodes for the audio section of a high quality unit! Heaps better than the Radiograms that followed later.
I remember stripping down one of those monsters when I was in my early teens. The cabinet wasn't too shabby either !

SandyK