Lower Output Capaitance

[galyons]

Just a reality check!?  Is there any reason that I cannot reduce the Smash output caps from 2.2Uf to 0.47uF?  The Paramounts' input impedance is 249K, so at 0.47uF, I calculate 1.4Hz Fo crossover frequency.  The 2.2uF Fo is 0.3Hz.  My experience has been that lower capacitance caps tend to sound a bit better.  My speakers don't have much output below 50Hz, so 0.3Hz is slight overkill. ;-)

Any issues?

Cheers,
Geary

[Paul Birkeland]

As posted elsewhere regarding the Eros by Paul Joppa:

"I have to thank long-time Bottlehead and occasional poster VoltSecond, who first brought this to my attention. The output interconnect is exposed to electromagnetic fields, many of which can impose hum through capacitive coupling. Thus hum is reduced when the impedance of a conductor is low, so a rational criterion is that the impedance of the output capacitor should be no greater than the output impedance of the source at 120Hz, the most common (non-magnetic) hum frequency. The output (source) impedance of Seduction/Reduction and Eros is around 4000 ohms, so doing the numbers we get 0.33uF. That's why Seduction uses a 0.47uF cap. The Seduction cap is a 250v part, so we need a different part anyhow, and the 1uF/630v part was in inventory for the then-current version of the SEX amp - that's how it got into the Eros."

In the Smash, this moves the minimum capacitance much closer to 2.0uF.  I personally wouldn't go any lower.  (Plus, if you ever decide to add a line level sub, many of them have a 10K input impedance anyway).

-PB

[johnsonad]

Great post PB (and PJ), thank you for increasing my understanding!

[galyons]

snip...
In the Smash, this moves the minimum capacitance much closer to 2.0uF.  I personally wouldn't go any lower.  (Plus, if you ever decide to add a line level sub, many of them have a 10K input impedance anyway).

-PB

Thanks PB,
I think I understand the theory put forth by VS.  Just to clarify, a sub is not likely to ever be considered with my current horn speakers.  Would the manifestation of the potential increased hum be interconnect dependent?

During the last week, or so, from my original post, I took the empirical route. I have several 0.47uF quality caps on hand, so I switched out the, then currently installed, 2uF Russia KGB PIO caps for 0.5uF KGB's, (aluminum foil,paper, oil). Now after a week of listening/burn-in, my ears say that the improvement is more than subtle.  The major difference, I noted, was increased speed and impact.  The leading transients are more dynamic.  Bass is more clearly defined with more impact.  There is also an improvement in micro detail and timbre.  There is no perceivable increase in hum.  My Smash-to-Paramount interconnects are AlphaCore Goerz flat silver ribbons, no shielding.  Hence my interconnect query.

I am going to try the  other caps I have, AmpOhm copper foil, Jensen copper foil and Dueland Cast copper foil.  I will listen for a week to each one. 

I will report back, but that will be more a cap review than the original "lower capacitance" question.

Cheers,
Geary

[Paul Birkeland]

Would the manifestation of the potential increased hum be interconnect dependent?

I certainly would imagine that this is the case.  When you swap a cap, I would listen to it for a week, then put the pair of caps back in that you removed prior, then re-listen.

[galyons]

This week+ I put in a pair of 0.47uF Jensen Copper foil/paper/oil caps. They have about 100 hours in place.  They are only slightly better than the Russian caps, a very small bit warmer with, perhaps, a bit better definition.  Most of the difference was in the upper mid range with a small improvement in timbre. 

Per PB's suggestion I put both the 2.0uF and 0.5uf Russian KBG back for a listen.  The smaller Russian cap seem to have slightly better dynamics than the Jensen's.  I think that the Jensen caps are just a little more musical.  Hard to define, but I could listen to either one, Jensen or KBG without feeling I was "missing out"!

But the bottom line, so far, is still, to me, the smaller the caps sound better!! Everything was just a bit clearer, more defined.  To my ears, there was no perceivable loss of bass or hum!  Jensen's are back in for a while more.

I have no sense for how well the Smash is selling, but it is a damn fine line stage.  At its low price point, it is  truly a high end value! (Which is really an oxymoron!)  The early microphonics have subsided, only slight "tube sing" when I switch volume and/or selection.  But, as I have have posted earlier,  my stepped attenuator and selector switch are much more mechanical than the stock devices.  There is just something magical with the DHT architecture and PB has captured it very well!

Cheers,
Geary

[galyons]

This week+ I put in a pair of 0.47uF Jensen Copper foil/paper/oil caps. They have about 100 hours in place.  They are only slightly better than the Russian caps, a very small bit warmer with, perhaps, a bit better definition.  Most of the difference was in the upper mid range with a small improvement in timbre. 

I replaced the Jensens with Dueland Cast Copper foils!!  The Duelands are more than a bit pricey, but they are a quantum upgrade from the Jensen's.  I now have the Dueland's in the Eros, Smash and Audio Note DAC.  The Eros is the latest and still breaking in, but the improvement has been dramatic. 

The upgrade has had consistent impact in all three units.  The most obvious improvement has been a  dramatic  extension in the bass and treble.  The imaging and soundstage seem more stable, on a blacker background. The Eros had the Ampohm Copper foils, (production ended a few years back), a wonderful value, if you can find them.  With the Dueand's, there is an increase in purity of timbre, but subtle.  The timbre improvement over the Russian KGB aluminum foil caps was more obvious.
The Russian caps are still a high value bargain providing much of the benefits of the pricey caps at less than 5% of the price.

For now, my caps experiment/search is over.  My ears are happy!!!

Cheers,
Geary