Upgrading coupling caps - which ones first?

The Mundorf caps are not in the same league as the Jupiter or Duelund in any important sonic attribute. However, they do need at least 150 hours of burn in. Mundorf are good mid tier caps, but lack the natural musicality of the top tier Jupiter and Duelund after burn in. 

So far I've been playing around with the .47uF output caps on the amp and right now I am trying to decide which ones are going in there permanently.  After that I will target the DAC.  So far I have tried Jupiter copper foils, Clarity MR, Mundorf Silver Oil, and of course the stock Nichicon polyester film caps. 

Surprisingly the stock Nichicons have a lot going for them.  They have a clear, wide open sound with the best bass of the bunch, and extended highs that seem tipped up and can be a bit thin sounding and on the shrill side at times.

The Mundorfs have an extremely smooth sound that is very musical, and are a tiny bit darker with a smaller soundstage in direct comparison with the Nichicon.

When I first installed the Jupiter, the imaging was what I noticed first.  It is more pinpoint but at the same time the sound was a bit closed in.  After a few days I reinstalled them for a second evaluation and they seemed to have opened up.  Their presentation is similar to Mundorf but not as smooth, and bass is not up to the levels of the Nichicon.

I am a big fan of Clarity TC caps in the power supply so I was eager to hear the MRs in coupling duty.  The first impression of these caps are that they are clear and have a big sound to them.  They don't quite have the smoothness of the Mundorf, they are instead more neutral and true to the source it seems.

These caps were all burned in for a week with a tuner on the speaker output of an old amp connected to a dummy load.  I'm not sure if they still need to be run in some more but I could swear that the Jupiters sound different now then when I had first installed them.

So what I've discovered is that all these caps do some things great but not everything.  Based on what I've heard so far, the ideal cap would have the bass slam and transparent open sound of the stock caps, combined with the smooth liquid highs of the Mundorfs.  Does such a cap exist?  We will see.  I can only imagine the can of worms I am opening up once I start with the DAC.  Why can't I leave well enough alone!?

Please let us know in what way you would like to change the sound. Those dac output caps will have a big impact on your overall sound. 

I support the concept of  beginning at the upstream  position  (source) and then progressively moving  down the audio signal pathway. Replacing the  output  coupling capacitors in my DAC yielded more impact than changing the  input  coupling  caps in my amplifier. Now I'll admit that changing the  single tweeter cap in my speaker's crossover was a very noticeable  upgrade without question. I used Duelund CAST in the DAC and speaker and Jupiter copper foil in my amplifier. I can recommend  either brand enthusiastically.
Charles,

The best place to start is your source as the whole system will benefit! ! Start with your dac's output caps. Please let me know what brand they are and what sound change or difference you want. I am very versed on what various caps sound like in dac's and amps being a chronic cap roller:)

Dont  try to change too much, but rather start here and give it time to settle into your system and mind. Some caps sound fuller, some warmer, some darker, some more focused on the upper mids and highs etc... 

Sounds good re the change to 1 uF.  As you may already have calculated, in conjunction with the amp's 70K input impedance that will result in a 3 db bandwidth extending down to 2.3 Hz, and only about 6.5 degrees of phase shift at 20 Hz.

Based on that and on the values you supplied, and without knowing anything further about the design of the amp, what seems likely to be the most optimal prioritization would be doing that pair of caps first, then the 0.47 uF caps, then the 0.1 uF caps.

Good luck.  Regards,
-- Al
  

The caps in the amp are .1 and .47 uF at 600V, and the output caps in the DAC are 3.3uF 200V. 

it is probably reasonable to expect that higher values of capacitance and voltage rating will tend to correlate with less ideal behavior of a capacitor.

My thoughts also.  Amp input impedance is about 70k which means I could get by with a much smaller ~1 uF cap on the DAC without any frequency rolloff.

It may be helpful if you could let us know the capacitance values and the voltage ratings of the three pairs of capacitors, and also the input impedance of the amp. If the amp has lowish input impedance (although it probably doesn’t, given that it is a tube amp) it would tend to increase the criticality of the coupling cap at the output of the DAC. And I’m thinking that as a rough rule of thumb, and everything else being equal, it is probably reasonable to expect that higher values of capacitance and voltage rating will tend to correlate with less ideal behavior of a capacitor.

Regards,
-- Al

I concur the furtherest up the signal chain.

That in itself isn't much help though.  Perhaps, think of it like this: the capacitor is a bell and music signal is a hammer.  The further up the chain the smaller the hammer and the less the bell will ring imparting its artifacts to the signal.

So, as you move from small to large signal the capacitor needs to have more and more dampening effect.