[bruce108]

Thanks, Tom, that helps.
 

 

[scootsit]

Tomb,
Some time ago, there was a discussion between you and dsavitsk (I think) about low impedance cans and you mentioned the output impedance of the millet mosfet. I found it once and now, no matter what I search, I cannot seem to find it! What is the output impedance of the millet mosfet?

 

[tomb]

Quote:

Tomb,
Some time ago, there was a discussion between you and dsavitsk (I think) about low impedance cans and you mentioned the output impedance of the millet mosfet. I found it once and now, no matter what I search, I cannot seem to find it! What is the output impedance of the millet mosfet?

The output resistors would easily override any impedance at the buffer's output.  So, take that value (RB14) as the output impedance - 22R.
 
The important focus of the discussion was that Dsavitsk and I do not necessarily agree with the drive to super-low output impedance in an amplifier.  It's one of the things that can cause Grado's to sound harsh, for instance.  Dsavitsk thinks it has to do with some headphones having a high inherent damping.  Apply such a headphone to an amp with very low output impedance and the result could be over-damping, which can lead to more distortion, not less.  That may be the key to why everyone seems to think Grado's need tube amps (inherently higher output impedance).  It has certainly proved out empirically for both of us.

 

[Beefy]

Quote:

The important focus of the discussion was that Dsavitsk and I do not necessarily agree with the drive to super-low output impedance in an amplifier.  It's one of the things that can cause Grado's to sound harsh, for instance.  Dsavitsk thinks it has to do with some headphones having a high inherent damping.  Apply such a headphone to an amp with very low output impedance and the result could be over-damping, which can lead to more distortion, not less.  That may be the key to why everyone seems to think Grado's need tube amps (inherently higher output impedance).  It has certainly proved out empirically for both of us.

 
Indeed, and it is entirely dependent upon the phones. IMHO the low impedance Audio Technicas sound far better from a low output impedance amp. The bass gets all flabby if the output impedance is high.

 

[scootsit]

I remembered being really interested in the debate. I just couldn't find it. That's really interesting and I'm very curious now to build something with really low output impedance and hear the difference. Thanks for the answer, very interesting.
 
I just picked up some cans with 4.4kohm impedance, there is actually audible sound out of the Millet, which is pretty impressive!

 

[scootsit]

In particular, this reminds me of the claims regarding the O2, and makes me really curious if those claims hold water given real-world, low-impedance cans. BUT, this thread is probably not the place for those considerations.

 

[tomb]

Quote:

In particular, this reminds me of the claims regarding the O2, and makes me really curious if those claims hold water given real-world, low-impedance cans. BUT, this thread is probably not the place for those considerations.

Well, one of the things I had confused for awhile was equating high inherent damping with low impedance phones.  A low impedance headphone does not necessarily mean high inherent damping (it just happens to be true for Grados).  However, high-impedance phones automatically mean that the damping factor will be higher for any amp: DF = Zload/Zout.  The real issue occurs with low impedance phones, which seem to be more prevalent today.  For 300 ohm headphones, the Zout of the amp is always very small compared to the headphones' Z, but at 32 ohms, small changes in Zout can have significant effects on the DF. 

 

[scootsit]

I can imagine that it would depend on the sensitivity and the impedance.

 

[ModMax]

Regarding the Millett MOFSET-MAX amp, RA2L/R and RA4L/R call for a 1Kohm Metal Film Resistor.  Is this a requirement, or can a 10Kohm Metal Film Resistor be placed here?  What about a 1Kohm Carbon Film Resistor instead of Metal Film?

 

[tomb]

Quote:

Regarding the Millett MOFSET-MAX amp, RA2L/R and RA4L/R call for a 1Kohm Metal Film Resistor.  Is this a requirement, or can a 10Kohm Metal Film Resistor be placed here?  What about a 1Kohm Carbon Film Resistor instead of Metal Film?

I answered this in your e-mail, but just in case someone else is interested in this post:
 
RA2-L/R serves as the capacitance isolator in the voltage supply line (+) between the tube signal section and the output buffer.  The value is not that critical, but 10K is probably way too much - better to stick in the 1K range.
 
RA4-L/R serves as the bleeder resistors for the output caps.  You can also assume that they serve to apply current to the output caps to charge them in case there's no load connected (headphones).  You could probably use the 10K here, but you might have to adjust the timing on the relay-delay.  Again, better to keep it at 1K to prevent having to change some of the other things on the amp that have been tested out and tweaked over the years.
 
Bottom line, neither resistor is in the signal path, nor are they critical if they vary by 10-20% or more.  So, your 1K carbon will do fine.

 

[preproman]

+1

 

[ModMax]

My next step in building my MOSFET-MAX is installing the VitQ caps I picked up from Beezar (http://beezar.com/catalog/product_info.php?cPath=24&products_id=68).
 
(1) My 1st question is:  Do the VitQ caps have a polarity?  I don't see any markings on the cap case nor have I been able to detect that either lead is connected to the cap case as suggested by other forums online.  Does it matter which direction they are installed?
 
Also, on the subject of boutique components, my build is as follows:
 
CA2/CA7 = Elna Silmic II 1000uF 35V
CA8 = VitQ .22uF 100V
CA9 = Wima .22uF 250V
 
(2)  My 2nd question is:  Should I leave CA9 blank with my setup or go ahead and populate with the Wima cap?

 

[scootsit]

Quote:

My next step in building my MOSFET-MAX is installing the VitQ caps I picked up from Beezar (http://beezar.com/catalog/product_info.php?cPath=24&products_id=68).
 
(1) My 1st question is:  Do the VitQ caps have a polarity?  I don't see any markings on the cap case nor have I been able to detect that either lead is connected to the cap case as suggested by other forums online.  Does it matter which direction they are installed?
 
Also, on the subject of boutique components, my build is as follows:
 
CA2/CA7 = Elna Silmic II 1000uF 35V
CA8 = VitQ .22uF 100V
CA9 = Wima .22uF 250V
 
(2)  My 2nd question is:  Should I leave CA9 blank with my setup or go ahead and populate with the Wima cap?

 
1) No.
 
2) I think so, but Tomb would be better to answer that.

 

[tomb]

1) No.
2) Yes - populate CA9 with the Wima.  The Elna's play well with the Wima's.  It's only with Black Gates at CA2 should you leave CA9 blank.  Also, don't try to put a boutique film cap back there, either - it will most likely end up cutting out some bass - that includes VitQ's as well.  Up front at CA8 VitQ's are the preferred choice (Sonicap Gen II's are good), but any other boutque film cap that will fit is appropriate, too.

 

[ModMax]

Many thanks, tomb and scootsit.