Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › New Millett Hybrid Maxed Amp
New Posts  All Forums:Forum Nav:

New Millett Hybrid Maxed Amp - Page 172

post #2566 of 6401
12/7/07 at 7:40pm
Quote:
Originally Posted by amb View Post
tomb, I understand your enthusiasm for tubes but your statement above is flat out false in the case of the Millett Hybrid (MAX or otherwise). When the amp clips, it does so just as harshly as any SS amp. Have you looked at a spectrum analyzer and/or a scope to see what happens before you wrote that paragraph? Also, on what basis do you claim that SS amp clip more readily than the Max (given the same supply rail voltage)? I don't need to remind you that the Millett MAX design has a SS output stage that's lifted straight out of the PPAv2...
Specifically, I was talking about the soft clipping characteristics of a tube, not the output stage, and I was definitely in error when I used the specific word "readily." One might interpret "more harshly" as encompassing like characteristics of "readily," but perhaps that was too strong a term.

That a SS amp would clip more harshly is not something I made up. This characteristic can be found in most literature that ever compares the two. It seems to me that in the MAX, the tube will clip way before the output stage, since the tube is acting as the amplification device in the circuit - but perhaps I'm wrong on that account.

I was not comparing specific amps in that post and I'm not going to do that anymore, even though others have. I think the MAX competes very well in its price category and that was the essence of what I was trying to convey, albeit rather enthusiastically.

Rather than incur further objection from you, however, I'll refrain from the jargon in the future, too. The MAX's great success is due in large part from not only your e12 design, but also your MOSFET diamond buffer design. I would rather do anything else than make you think I've insulted your designs after what you've contributed to the community. That was the furthest thing from my mind.
Reply
post #2567 of 6401
12/7/07 at 7:47pm
I want to build a beta22 :P
Reply
post #2568 of 6401
12/7/07 at 7:56pm
Quote:
Originally Posted by ruZZ.il View Post
I want to build a beta22 :P
I can guarantee you that a Beta22 never entered my mind in any comparison with the MAX. Maybe I should've been more specific, but I'm in enough trouble already.
Reply
post #2569 of 6401
12/7/07 at 8:01pm
Quote:
Originally Posted by amb View Post
When the amp clips, it does so just as harshly as any SS amp. ... Also, on what basis do you claim that SS amp clip more readily than the Max (given the same supply rail voltage)?
Quote:
Originally Posted by tomb View Post
Specifically, I was talking about the soft clipping characteristics of a tube, not the output stage ... That a SS amp would clip more harshly is not something I made up. This characteristic can be found in most literature that ever compares the two. It seems to me that in the MAX, the tube will clip way before the output stage, since the tube is acting as the amplification device in the circuit - but perhaps I'm wrong on that account.
tomb, you are right that tubes clip in a more pleasing way than transistors. But, with the Millett, any clipping is likely in the CCS, which is solid state. In order to get tube type clipping, you would need to overdrive the grid. But, because of the way that the amp is designed, this won't likely happen until well after the CCS's room to swing has been passed.

As a (very) simplified example, with the 12AE6A, with .6mA of current and a 13V plate bias, the cathode is at ~0.5V. Assuming that the supply voltage is 26V, this means that the CCS has 13V to swing, if it can go right up to the rail ... which it can't. Optimistically, it might have 8V of swing in it. So, the tube can swing ~1V before clipping, and with a mu of 16, this would require the CCS to swing 16V, which it clearly can't do. To extract the full gain of the tube, and get to the tube's clipping, you would need a CCS that dropped more like 30V, which would mean a plate supply of 45V or so with that same 13V plate bias.
Reply
post #2570 of 6401
12/7/07 at 8:25pm
OK, I see, Doug. Thank you for the explanation.

I am sorry for my blatant bias and jargonistic suppositions. Please accept my apology for the trouble I caused, guys.
Reply
post #2571 of 6401
12/7/07 at 9:48pm
All good discussion TomB, Dsavitsk, Amb,ruZZ.il! This is what brings the characteristics of the Max into sharp focus. Really interesting stuff. Another post about 8 weeks ago, from someone I can't recall, talked about the output stage. He mentioned that when the output is overdriven, that rather than clipping it moved from Class A operation to Class B. Whats your understanding of that statement? Amb? Dsavitsk?
Reply
post #2572 of 6401
12/7/07 at 11:32pm
amphead, the push-pull output stage will transition from class A to class AB when it's asked to deliver more than double the (peak) current than its quiescent current. This may occur with low-Z loads well below clipping threshold.

You have to understand that clipping happens when the output voltage swing would exceed what's available from the supply voltage minus internal losses. It's a voltage swing phenomenon that would occur when you turn up the volume high enough, even without a load at its output.

The transition from class A to class AB, on the other hand, is load-dependent and quiescent-current dependent.
Reply
post #2573 of 6401
12/7/07 at 11:59pm
Thanks Amb, that makes sense! Higher output impedance couples more voltage swing, which could translate into the peaks being flattened hence clipping. And lower output impedance(with an overdriven signal), tends to degrade operation from Class A to Class AB, which can introduce crossover distortion. Two different ways to lose clarity, one when headroom is exceeded and the other when current flow pushes operation into the AB region.
Reply
post #2574 of 6401
12/8/07 at 12:24am
amphead, replace "output impedance" with "load impedance" in your post above and it would be mostly right. Clipping is nasty -- instant harshness in the sound. It should be avoided completely. On the other hand, class AB operation, despite onset of crossover distortion, may not be so bad, depending on the amp design. In an amp with global negative feedback (which the Millett isn't) such distortion would be corrected by the feedback, albeit not completely.
Reply
post #2575 of 6401
12/8/07 at 3:40am
Yeah, load impedance is the correct term. I stand corrected. The crossover distortion probably isn't much of a factor in the Max, due to its lower power(current density), compared to speaker driving amplifiers.(scratch this theory it doesn't match what I just read about the effect of biasing) ;-)
Reply
post #2576 of 6401
12/8/07 at 1:30pm
A work in progress on the MiniMAX -


(click for big pics)

The Lansing case is really cool except for the color. I made the mistake of thinking it might be neat to try white for once. As you can see, however, it's a refrigerator almond color, instead. I'm probably going to paint it and be sure to order a black one next time. The Lansing case bottom is the key to the whole case and is of very high quality. However, the top is thin gray steel covered with a grained-vinyl. It all seems to come together very nicely, though, and the overall effect is tight and compact.
Reply
post #2577 of 6401
12/8/07 at 2:24pm
Very cool Tomb!! The board looks very compact, I imagine the cap choices would be pretty limited.
Reply
post #2578 of 6401
12/8/07 at 4:03pm
hey that does look nifty! any circuitry changes? I think you mentioned it was pretty much the same..
Reply
post #2579 of 6401
12/8/07 at 5:07pm
Quote:
Originally Posted by ruZZ.il View Post
hey that does look nifty! any circuitry changes? I think you mentioned it was pretty much the same..
Yep, pretty much the same - just some tweaks here and there. The 1" height is a rigid constraint, though - and the Wima's and boutique caps are 0.1uf's. Those are some Sonicap GEN II's in there now, but I've got some 0.1uf K42's that will be going in. Colin is working on some more refinement.

It's basically a nice mixup with the big one, and the case is kind of interesting to work with - but we'll see when I go to punch holes in it.
Reply
post #2580 of 6401
12/8/07 at 6:43pm
What do you all think about this design for the top panel? If I send in my own metal to Front Panel Express, it's only $23.

I think it looks kind of busy and angular on the computer screen, with the stark black and white contrast. But laying down flat with some wire mesh under the heatsink holes, it should be more subtle.

Reply
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › New Millett Hybrid Maxed Amp