[Jaypetermen]

I showed the PPA v2 and MMM board schematics to a friend to get his advice on which board he would recommend. He thought both designs were very good.

He was interested that the MMM channels are not balanced as the PPA are. What is the design concern in this type of amp?

Jay

 

[amb]

Quote:

Originally Posted by Jaypetermen He was interested that the MMM channels are not balanced as the PPA are. What is the design concern in this type of amp?

Can you elaborate on what he thought was unbalanced with the M³?

If he was referring to the fact that the opamp seems to drive the N-channel MOSFET directly whereas the P-channel is "below" the Vbe multiplier biasing circuitry, this is done to minimize the number of parts while accomplishing several functions. The opamp sources the current through the Vbe multiplier for adjustable MOSFET biasing, and this current is pulled by the cascoded JFET current source which also keeps the opamp operating in class A. The whole output buffer has only one BJT, two JFETs and two MOSFETs. In this amp it is a virtue, because performance is not compromised to achieve the simplicity.

During the early experimental phase of the M³ project, I built a prototype with a diamond buffer output topology virtually identical to that of the PPA v2, except that the output devices were MOSFETs rather than BJTs (and biased accordingly). However, I measured higher intermodulation distortion in that configuration than the simpler circuit that we now use.

 

[Jaypetermen]

Quote:

Originally Posted by amb Can you elaborate on what he thought was unbalanced with the M³?

He stated that the FET's are not exact symmetric electrical components, to get the circuit exact one would need to adjust the "bias" for each individual device separately.

 

[amb]

Quote:

Originally Posted by Jaypetermen He stated that the FET's are not exact symmetric electrical components, to get the circuit exact one would need to adjust the "bias" for each individual device separately.

If he is referring to the MOSFETs, yes they are not perfectly complementary, but neither are NPN and PNP BJTs. As for needing separate bias adjustments for each side, he is not correct.

 

[Jaypetermen]

He replied that; in differential mode (push pull) the 0 crossing needs to not have a gap. The bias of each is only method to control this. Maybe its not enough difference but driving the two gates with a single constant current source the bias is set for ?, I guess an average of the two parts. Do not forget the bias is very very important in controlling/adjusting for min distortion.

 

[morsel]

There is no gap. Default bias is 80mA, plenty for Class A operation under all but the most extreme conditions. The output stage is a well understood complementary MOSFET pair similar to other amps such as the SDS Labs and Hafler DH-200 . We went through an extensive design phase with public review before releasing M³. Invite your friend to read the entire M³ Project Announcement thread, and come here to speak for himself if he is still inclined.

 

[ablaze]

its been a while, has anybody built/listened to both yet?

 

[MrYman]

......yet?

 

[n_maher]

Here you go...

http://www6.head-fi.org/forums/showthread.php?t=119766

Nate

 

[MrYman]

ohh yes, i saw this earlier today, i wanted to post the link here myself but then forgot...

splendid review... both you and ian... you also made me very curious about Millett Hybrid Tube amp, which you both like a lot... did you write more about Millett's amp somewhere? most interesting for up to $400 to me are now: ppa, m^3, millett hybrid, millett ha-1 & wna....

 

[n_maher]

You can find a pretty decent review of a Millett Hybrid here , it's actually an amp that I built so I voiced my opinion in the thread. Ian also spent some time with it and put his thoughts in there as well.

There isn't a bad choice on your list that I can see. I don't like trying to figure out which amp is "better" as the only measuring device that I have is my ears, which I can guarantee are not the same as yours. I can listen to a Millett, M3, or PPA V2 for hours on end and be perfectly happy.

HTH,

Nate

 

[MrYman]

Thanks again Nate! Yes, I agree that all are very good, with perhaps just bit different 'flavours'... though I cannot have all of them as you do

What about WNA amp? I didn't yet notice if you've already built it or not? If not, considering to build one? Maybe heard it?

On the side note, what's your opinion of NJM4556AL chip used as headphone amp? (Two days ago I've got Focusrite Saffire device, which uses that chip... Saffire sounds better and louder than Audigy2 Platinum to my ears... but I'll listen bit more before posting impressions on 'my' portable audio devices thread...)

 

[n_maher]

Quote:

Originally Posted by MrYman Thanks again Nate! Yes, I agree that all are very good, with perhaps just bit different 'flavours'... though I cannot have all of them as you do

Do yourself a favor and don't! Waaaaaay too much $$.

Quote:

Originally Posted by MrYman What about WNA amp? I didn't yet notice if you've already built it or not? If not, considering to build one? Maybe heard it?

Haven't built it and don't think I will as I'm quickly coming to the end of my crazy amp building days. I've got a couple ideas on some of Peter Millett's other designs (like the HA-1 you referenced) but I've found that while I enjoy building amps I really enjoy listening to them more.

Quote:

Originally Posted by MrYman On the side note, what's your opinion of NJM4556AL chip

Can't say I have one. Haven't heard one and I'm not familiar with that chip. It's certainly an interesting bit of kit you've got there, perhaps you should let us know what you think of (wink, wink, nudge, nudge).

Nate