Hi everyone,

I'm the third owner of the ZSTAGE that was originally delivered to  meraklya (Sean) in August 2012, went back for warranty service and came back to him two weeks later, and was later sold to Orangecrush (Will) in March 2013, who sold it to me three weeks ago.  Amazingly, it's still in "Mint" condition - truly, it looks brand new - but I'm not selling it - not yet, anyway!



This poor thing has had a hard time finding a home, it seems, but I'm already feeling attached to it.  Still, I've made some unexpected observations that are somewhat disconcerting.

First, let me proclaim that I am completely satisfied with the transaction I had with Orangecrush (Will) - no complaints, whatsoever.  He sweetened the deal with a couple of PSVANE 12AU7s in addition to the stock JJ, I paid a lot less than what the bundle would have cost new, and I didn't have to wait three months.  



Prior to this purchase, I've been a solid state guy.  My goal in buying the ZSTAGE was (is) to make the following chain, which is fairly neutral and slightly laid back (fourth row seating), and lacking in both dynamics and bass energy, a lot warmer in the mids, more forward (front row), punchier, and hopefully, a little stronger in the bass:

WAV > Foobar 2000 w/WASAPI event > USB > CEntrance DACmini CX (line out) >  Burson Soloist > Beyerdynamics T1

The ZSTAGE has been inserted ahead of the Burson Soloist, but on first receiving it, I was unimpressed.  I had no reason to believe it wasn't working as designed and I'm still not convinced that it actually has a problem.  It seems to be functioning perfectly, except for one annoying observation:

With the JJ 12AU7, for example, The ZSTAGE is supposed to permit a decrease or increase in dynamics, with a simultaneous decrease or increase in weight.  Steve Deckert writes (in the ZSTAGE manual) that a decrease in dynamics and weight can be had by adjusting the ZSTAGE gain control to something less than unity gain (less than the DAC's output - as with the ZBOX), while volume-matching with the amp's volume control to maintain a desired SPL at the headphones or speakers.  An increase in dynamics and weight can be had by adjusting the ZSTAGE gain control to something greater than unity gain (greater than the DAC's output), while volume-matching.


Quoting the ZSTAGE Owner's Manual:

Approximately 1/2 way up is “unity” meaning it is neither adding or subtracting gain.

Now, it's easy to understand that the point at which unity gain would occur could vary with the tube I'm using, especially if I were to use a 60 mu 12AT7 instead of a 20 mu 12AU7 - at least I think that's the case, as I've not yet tested any 12AT7 tubes.

Using an SPL meter (an iPhone app by JL Audio) and a white noise WAV file, I've determined that, with the stock JJ 12AU7, unity gain occurs just below 3:00 o'clock on the gain control of my ZSTAGE.  Thus, nearly 75% of the range of adjustment available to me will only reduce gain to the amp (theoretically reducing dynamics and weight).  Only the top 25% of the ZSTAGE's available range of gain control actually increases the gain to the amp,  (potentially increasing dynamics and weight), AND when ZSTAGE gain is turned all the way up (even a few degrees shy of maximum), I can hear distortion - unacceptable distortion.

Without explaining why I was asking, Burson support has responded to a query telling me that they've tested the Burson Soloist with inputs as high as 12Vrms, without any evidence of problems.  CEntrance support has assured me that the line level output of the DACmini CX is 2.0Vrms, so I'm thinking that I'm somehow overdriving the JJ tube when ZSTAGE gain is set to maximum - rather than overdriving the Burson Soloist, but I really don't know for sure where the distortion is happening.

Keep in mind, however, that the ZSTAGE manual specifies it having an adjustable output of only 0 to 5 Vrms - far short of the Burson's claimed tolerance for 12Vrms.

One thing's for sure:  I'm only able to use about 20% of the total adjustment range of the ZSTAGE gain knob for increasing the gain above unity gain, and thus, I have a very narrow range within which I have any hope of increasing dynamics and weight.  

The result was that I really wasn't hearing much, if any, improvement in dynamics and weight, before I start getting distortion.  

Rolling in either of the PSVANE 12AU7s yields the same results.  

It then occurred to me that perhaps the ZSTAGE was just not happy with the DACmini CX' DAC section putting out 2.0Vrms (why not?), so I whipped out my trusty Sanza Clip+, which I once read has been measured with an output voltage (at full volume) of only 0.55Vrms.  

Replacing the 2.0Vrms DACmini CX with the Sanza Clip+ did the trick!  When I feed the ZSTAGE with the low, 0.55 Vrms output of the Sanza Clip+, "riding the gain" of the ZSTAGE actually makes a dramatic difference in dynamics and weight - a big difference from what I have been experiencing with the DACmini CX feeding the ZSTAGE with 2Vrms!  Previously, my SPL meter evidenced unity gain occurring at just shy of 3 o'clock - using the DACMini CX and any of my three 12AU7 tubes.

But when using the 0.55Vrms Sanza Clip+, unity gain occurs with the ZSTAGE knob at only about 8 o'clock, which gives me more than 3/4 of the ZSTAGE gain adjustment range with which to increase dynamics.   I have to set the amp's volume much higher than when using the 2Vrms DACmini CX, but the weaker 0.55Vrms output of the Clip+ permits a much larger adjustment range on the up side of unity gain, vs. on the down side of unity gain - and I haven't detected any noise problems that could be associated with driving the amp with less gain.

Best of all, when I spin the ZSTAGE knob fully clockwise (and turn down the Soloist's volume control to maintain a desired SPL), I'm no longer getting any of the distortion I was hearing with the knob near or right at maximum gain, where it was especially noticeable in the bass.  

Given that ZSTAGE specs show a maximum output of 5Vrms, I'm (somewhat) convinced that the distortion I was hearing with the DACmini CX when ZSTAGE gain was at or near its maximum was coming from the ZSTAGE itself - as if the ZSTAGE does not like having an input as high as 2Vrms.  It is MUCH happier with an input of 0.55Vrms, where I hear no distortion at any gain setting.

The lower output voltage of the Sanza Clip+ has converted the ZSTAGE from being nothing more than a tube buffer (offering no gain)  into the gain stage it was designed to be - offering a great deal of sonic tailoring, just by "riding the gain." (Not to mention tube rolling.)

So...  With this observation, I ordered some (inline) 6dB Harrison Attenuators to bring down the DACmini CX output, when using it to drive the ZSTAGE.   They're working great with the DACmini CX line level output reduced by 6dB.  Unity gain with the DACmini CX is now at about 7:30 on the ZSTAGE gain adjustment, instead of at 3:00, with no distortion heard at maximum gain - making the DACmini CX output more like that of the Clip+.  

:)

Since installing the Harrison Attenuators, I've ordered two more 12AU7s - a Siemens and Radiotechnique, and they, too, just as with the stock JJ and the two PSVANE's, are a lot happier when the DACMini CX line level is attenuated.  Keep in mind that I'm getting almost exactly the same gain from the 6dB attenuated DACmini CX as I get with the Sanza Clip+, so it's not as if the DACmini CX was somehow putting out a lot more than 2.0Vrms.  No... my ZSTAGE just can't handle 2.0Vrms of input voltage.  It's much happier with 0.50Vrms.

So, now I'm waiting on my first 12AT7 - a Mullard.  At 60 mu vs. the 12AU7s' 20 mu, I'm wondering if I will hear distortion and where exactly unity gain will be with the 12AT7 rolled in.

Comments?

Thanks!

Mike

Thanks for the quick reply!

I hear you regarding the possibility of a synergy issue.  I could try using some other DACs and amps, to rule this out, but everything I have is solid state.  Is it possible that tube amps are somehow more sensitive/responsive to varying input voltages than solid state amps?

I know that 12AT7s have three times the gain of a 12AU7, but I don't know that this will put me back into getting distortion when the ZSTAGE gain is set to maximum - with or without the 6dB of input attenuation.  I'm only guessing, but I suspect that as long as my ZSTAGE is only seeing a 0.5Vrms input when using the 12AT7, I won't hear any distortion at the far end - assuming, of course, that the ZSTAGE output doesn't exceed whatever the Burson Soloist can handle (which, according to Burson support, is something greater than or equal to 12Vrms).

For sure, I'll get a more rapid change per degree of arc as I adjust the ZSTAGE gain knob, than when using a 12AU7.

I will follow-up later, but thanks very much for your recommendations.  I'm stumbling along here, just trying to figure out how it is that other ZSTAGE owners are getting no distortion in the absence of 6dB attenuators.



Mike

Thanks very much for that detailed reply, Will!

I can say that your experience brings me some satisfaction in that it leads me to believe there's nothing wrong with my ZSTAGE - that it's reasonable to experience some distortion when the gain knob is at or near maximum with an input voltage near 2Vrms.

I'm especially assuaged by your use of software attenuation to accomplish pretty much the same thing as I've done with the Harrison Attenuators.  

You've also satisfied my curiosity about what to expect when my Mullard 12AT7 arrives.  It seems I might have to apply even more attenuation going in to the ZSTAGE, if I'm to enjoy a large adjustment range on the gain control.  Otherwise, I could find myself back to having a very narrow arc within which gain exceeds unity gain while not causing distortion.  I will post my findings later, once the 12AT7 has arrived and I've burned it in.

Answering your question, I'm convinced that the Harrison Attenuators are wonderfully transparent.  I've used them before and have never found fault with them.  They come in different "strengths," so they might not perform as well at higher attenuation levels, but the 6 dB versions sound very clean to me - with no coloration or noise of any kind. but I've never really dug deeply into A/B testing them at different frequencies, so YMMV.    Just to avoid any noise vulnerability along my interconnect cables between DAC and ZSTAGE, I've got them mounted at the input jacks of the ZSTAGE - not at the output jacks of my DAC.  

Quote:  How this software pre being before the source equates with an attenuater after the source I can't say.

I can't say with certainty, but I'm concerned that you might be reducing your bit depth by using software to reduce the output voltage of your DAC. I know that most DAC manufacturers prefer that you maximize the "volume" in your OS and/or player software (Foobar, etc.) because dynamics can be reduced with a reduction of bit depth.

Thanks again for you input!



Mike

Hi Will,

From the sounds of it, it's possible that your software is smart enough to avoid reducing bit depth when adjusting the "volume" sent to the DAC.  I really can't contribute much more to reassure you, but I do know that you'll want to keep your OS volume control turned all the way up, in any case, as that will definitely reduce bit depth, and therefore dynamics.

I suspect the Mullard 12AT7 that I've ordered from Brent Jesse is the same one you are using.  Here is his description of the tube:

Quote:


It should be here soon, so I'll be sure to follow-up with my findings.  I'm as much interested in the warmth it will bring to my somewhat cold Beyerdynamic T1 as I am interested in finding out how it behaves in the ZSTAGE, being that it is a 60 mu tube, with three times the gain of a 20 mu 12AU7.   I'm really encouraged by your use of this tube (or 12AT7's in general).

More later!



Mike