Transformer or capacitor coupling?

[neazoi]

What is the best coupling in terms of sound quality,
a good quality output transformer or a good quality capacitor?

[Sch3mat1c]

A terrible capacitor beats a good transformer any day.

Tim

[SemperFi]

Hehe, here we go again

I must say I rather like IT coupling, and if a cap is so much better, I have something to look forward to

How 'bout choke loaded stages? Do we have all the same parasitic problems about chokes as we do transformers? If choke loading, is it better to add a cap, or simply couple the signal via the core into another wire (transformer)?

[kstagger]

completely anecdotal:

I built a transformer coupled 6J5 preamp. I thought it was the bees knees until I built an triode-connected EF86 preamp with a white-cathode follower. I preferred the EF86 preamp - more drive, dynamic, blah blah.

And slightly related - on my second system I have the choice of amplifiers - an autoformer coupled McIntosh 250 and a refurbished Dynaco ST-80 that is capacitor coupled. I prefer the '80, even with the output going through a 4700uF capacitor!

[levelfive]

Originally Posted by neazoi

What is the best coupling in terms of sound quality,
a good quality output transformer or a good quality capacitor?

Hi,

Do You mean interstage transformer?
Yes, it is much better. But high-quality IT is very expensive

Kind Regards, Konstantin

[kevinkr]

Making good transformers is much more difficult and in most instances a lot more costly than making a good capacitor.. That said my system currently has quite a few transformers in the signal path - in all instances their use either resulted in a simplication of the circuitry required to do the job or provided other audible benefits. In a couple of instances they simply replaced a several op-amps in a dac or sacd player, in yet another they provide some additonal gain for my LOMC cartridge, and finally my line stage is transformer based and has acquitted itself against any and all competion so far which is why it gets to stay.. Power amplification is via SE DHT with output transformers. Most sources in my system go through 3 transformers on the way to the speakers.. The replacement amp I am working on has an IT...

In the case of audio chokes they are subject to many of the limitations imposed by transformers. Good ones need to be built to minimize stray capacitances between the adjacent windings and the windings and the core.

In addition like a transformer the core of a choke can saturate with excessive quiescent/signal current reducing inductance to a fraction of its intended value and generating distortion due to the non-linearity so caused. In addition like a transformer there are the magnetization losses to contend with, and linearity of the core material itself with varying levels of AC flux (BH curves) whether or not DC is present. So in short much the same care should be applied in their specification and use as with transformers.

Chokes like transformers may have a certain amount of winding inductance, so the choke is reactive in nature meaning that the loading it presents changes with frequency, usually with adverse consequences for driving stage linearity, output voltage swing capability and FR response flatness at low frequencies, and due to stray capacitance these effects are also exhibited at the upper end of the choke's response range. Really bad chokes may also exhibit HF resonances due to interwinding capacitances and the unintended parallel resonant circuit that creates. (I have never seen one that bad, but I have been bitten by the other two issues mentioned.)

Some would argue that the above precludes the use of chokes, but in instances where large PP voltage swings are required without insane supply voltages the choke actually makes a lot of sense. I've found in fact that the proper choke can be quite good, as in the case of a high quality transformer, OTOH scrimp here and the results are likely to be very audibly inferior to other equally poorly thought out schemes..

Both chokes and transformers in low level applications are very subject to picking up external magnetic fields and either need to be designed to be self-shielding or encased in a mu metal can or both..

My experience indicates that all of these issues can be sufficiently well addressed for audio applications at levels of a few hundred uV and above that both transformers and chokes are viable design choices with appropriate care in application. It is mostly a matter of cost, design goals, and the tastes of the designer in question.

I never thought I would ever advocate for chokes or transformers in audio gear except in power supplies.. How things change..

[Andrewbee]

I don't think either is better or worse (How's that for siting on the fence!)

On one hand you have an iron core with many layers of wire with all its "issues" versus a conductive foil separated by a dilectric with all its "issues"

I have used both in between stages and have enjoyed both but would say I prefer a capacitor. Reason being its smaller, extended bandwidth, cheaper and different types can alter the sound somewhat, I think of it as I do cooking, different seasonings.

[BudP]

Having built a bunch of interstage transformers and a few amps to allow comparison, I can offer the following.

Assuming our level three interstage, using M3 core and having a modest +3 db hump at about 30 kHz unloaded (which is better than the other direction) and a Multicap / Reliacap PPFX poly and tin at 400v.

They have equivalent sonic characteristics though the cap has a slightly fat sound to everything, likely due to it's higher dielectric constant polypropylene barriers, more electrons in the same space equals more time to charge and discharge. This is not a major problem, but is noticeable in direct A/B subjective listening with EnABL'd drivers.

The only other difference in sound quality comes with differing loudness. When the cap passes a fortissimo from any particular instrument ,within an orchestral sound field, that instrument seems to move towards you in the sound field illusion. So, in a general sense the illusion of a sound field is always changing it's depth of field relationships with a capacitor interstage. In an IT this does not happen, at all. The same tonal characteristics and relative amplitude takes place, but the instrument in question remains in scale.

The real difference here comes in the relative cost. The PPFX is probably $13 or so for a 0.47 mfd up to $32 for a 1 mfd cap. The level three IT is around $330 with a not much less capable level one around $120. Difference in amount of low level detail and "ease" and grace under pressure being the primary differences.

So, again, a matter of taste, times wallet evacuation. There are less expensive interstage transformers, but you cannot pit them against the multicap level of capacitor, they will not measure up.

Bud

[Sch3mat1c]

'Course, you can skip it all and go DC coupled. Sure way to fix motorboating is to not have it!



This circuit specifically isn't very well suited to audio, but the biasing scheme can be applied to any case. Global DC feedback is especially interesting for stability -- you could plug in a 7KY6 for a 6SN7 half and not care the least!

Tim

[Wavebourn]

Originally Posted by Sch3mat1c

A terrible capacitor beats a good transformer any day.

Exactly.
It is one of main reasons why I started to design hybrid amps.