"Millett Hybrid" redux: NuHybrid headphone amp, using the Korg Nutube

[vixr]

  If you wire it off board, I don't see any reason that a Blue Velvet wouldn't work.

so, the higher Ohms value isn't an issue?

 

[Sodacose]

Maybe Pete will comment, but I don't think there would be major input capacitance with the buffer that follows it to create a hf roll off.

 

[vixr]

Maybe Pete will comment, but I don't think there would be major input capacitance with the buffer that follows it to create a hf roll off.

Thanks for the speedy answers, I have gone the other way and used a 50k Ohms pot in place of a 100k pot with great success. It was done on the 6DJ8 "Headwatt" amp by Bruce Heran...

 

[Sodacose]

I bet the lower pot value was a parts availability choice. Hard to find dual gang audio taper in 50k-100k values from digikey/mouser. At guitar parts values (250k+) then you do usually have to start taking into account what it is feeding.

 

[vixr]

I bet the lower pot value was a parts availability choice. Hard to find dual gang audio taper in 50k-100k values from digikey/mouser. At guitar parts values (250k+) then you do usually have to start taking into account what it is feeding.

I actually did try to find a 100k DG AT pot for a few hours on the interwebs, the closest I could get was two singles that allowed the shaft from either pot to be inserted into the back of the other pot. Seemed dubious at best...didn't buy them.

 

[tomb]

  Haha, aw man.  I just finished a hybrid inspired by your original:
 
http://www.head-fi.org/t/838886/my-take-on-the-starving-student-el-estudiante
 
Now I've got to build another?  
 
Thanks for sharing, Pete!

I hate to keep playing Polly Prompter in this thread, but "original" and "starving student" do not belong together.  They are separated by about 6 years, at least.

  Pete is probably being too polite to correct this, but he seems to be stating that the NuHybrid is a redux of original Millett Hybrid, first designed and published by Pete in Audio Xpress magazine - all the way back in 2002.  Here's a copy of that article:
Build A Low-Voltage Tube Hybrid Headphone/Line Amp

 
The Millett Hybrid was then followed by the revMH Millett Hybrid, designed by Drew Dunn and Nate Maher and released in 2005-2006.  AMB actually was the first to use the OPA551 on Pete's Millett Hybrid design instead of the BUF634 (with the revH Millett Hybrid PCB).  A little later on, Steinchen designed a discrete buffer for the revMH Millett Hybrid that became known as the MH Diamond Buffer Boards:

 
The MH design was later followed by the Millett Hybrid MAX, designed by Colin Toole using Pete's original design with a maxed out power supply and discrete, heat-sinked diamond buffer.  Then there was the Millett Hybrid MiniMAX, and finally, the Millett Hybrid MOSFET-MAX, which is still available on PCB today.  All of these were based on low-voltage, space charge tubes - the 12AE6, 12FM6, and 12FK6.  They also used 24V walwarts, although most were tweaked to 27VDC (see earlier post in this thread) through the use of 24VAC walwarts.  All of this is documented in great detail on diyforums.org and can be gleaned from looking on Pete's site at the different headphone amp designs.

 
Pete will correct me if I'm wrong, but the Starving Student was somewhat of a departure from the Millett Hybrid family.  It came along in 2008, with Pete's and Nate Maher's posted as the first two builds.  It originally used point-to-point construction, high-voltage tubes (they started out at $1-$2!), a 48V surplus Cisco VoIP telephone power supply, and uniquely - used the tube heaters to bias the output buffer MOSFETs (none of which existed in any of the Millett Hybrid variants).
 
A little later after Pete came out with the Starving Student, Dsavitsk designed the first PCB for the Starving Student.  We sold batches of kits a number of times, referring to the design as the Starving Student Millett Hybrid, or SSMH.  Those kit batches are now rare, because the tube itself became rare - the 19J6.
 
Anyway, I've finished my NuHybrid except for soldering in the NuTube itself.  I want to rinse and clean up the board in the morning, and I'm not quite sure whether the Korg NuTube is moisture-proof.  I'll report back when I get it hooked up and running.

 
 
 
P.S. At some period along the timeline described above, Tyll Hertsens actually sold a version of Pete's Millett Hybrid as a HeadRoom headphone amplifier.  I've never had the pleasure, but I know Nate had one and it was probably closest to the Millett Hybrid MAX in design and used the same tubes, power specs, etc.:

 

 

[Sodacose]

Thanks for the clarification and reading, tomb. I was referring to the original Starving Student design as found on Pete's website:

http://www.pmillett.com/starving.htm

 

[tomb]

Thanks for the clarification and reading, tomb. I was referring to the original Starving Student design as found on Pete's website:

http://www.pmillett.com/starving.htm

Yes, but you referred to the Starving Student as "a hybrid" that was Pete's "original," which is why I clarified.

 
Maybe I'm just anal about it because I've been following, building, selling, and documenting Pete's hybrids for quite a few years, now.

 

[tomb]

Here's mine so far:
 

 
As I said, I'm waiting to clean up the bottom of the board before installing the NuTube, in case it's not moisture-proof.  The smallness of the PCB surprised me.  I was actually expecting something on the order of the revMH Millett Hybrid PCB, but this thing is about half the size.  I used some left-over MiniMAX caps that are really too large for the space, but I got them in somehow.

  I also will install a couple of Vitamin Q's for the output bypasses - C4 and C7, but they'll have to go on the bottom of the board.  There's no room at all with those caps.  I also jumpered R8 and R9.  We'll see how that works instead of 1R resistors, but I always preferred the sound when jumpering the output resistors on the Millet Hybrid, et al.  We'll see how it goes.

 
 
P.S. This took me about an hour to get this far, but I was working fairly fast.  It should be pretty easy for most to put together.

 

[Sodacose]

Hi Pete (if you're still here). Can you talk at all about why you decided to use the Nutube as a voltage gain device? It's an interesting "tube" with some challenges to overcome, so I'm just curious about the overall design goal and how the NuTube fit into it.

Seems it has some size/efficiency advantages. And maybe it just sounds good (I know this isn't your first amp using it).

 

[pmillett]

  would it be possible to use an Alps "blue velvet" 50k ohms pot with this amp? I have a few spares, I could wire one off board...

 
Sure, it will work fine.
 
The resistance here isn't very important.  Since it is driving a source follower, it has very little loading capacitance to contend with (though the gate capacitance of a MOSFET is high, the gate and source move together, so the resulting capacitance is very low).
 
I picked 10k mostly because a lower input resistance seems to help with noise pickup.  But 50k or 100k are fine.
 
Pete

 

[pmillett]

  As I said, I'm waiting to clean up the bottom of the board before installing the NuTube, in case it's not moisture-proof.  The smallness of the PCB surprised me.  I was actually expecting something on the order of the revMH Millett Hybrid PCB, but this thing is about half the size.  I used some left-over MiniMAX caps that are really too large for the space, but I got them in somehow.

  I also will install a couple of Vitamin Q's for the output bypasses - C4 and C7, but they'll have to go on the bottom of the board.  There's no room at all with those caps.  I also jumpered R8 and R9.  We'll see how that works instead of 1R resistors, but I always preferred the sound when jumpering the output resistors on the Millet Hybrid, et al.  We'll see how it goes.

 
 
P.S. This took me about an hour to get this far, but I was working fairly fast.  It should be pretty easy for most to put together.

 
Wow, you're fast!
 
I think the Nutube should be fine with cleaning solvents or water... it's all glass and metal, and is sealed.
 
You are spot on about R8 and R9 - some people will like zero ohms better than 1 ohm, and some might also prefer 10-33 ohms there.  It depends on your headphones and listening preferences.
 
I'm anxious to hear your listening impressions, especially as compared to the other Millett hybrid variants!
 
Pete

 

[pmillett]

Hi Pete (if you're still here). Can you talk at all about why you decided to use the Nutube as a voltage gain device? It's an interesting "tube" with some challenges to overcome, so I'm just curious about the overall design goal and how the NuTube fit into it.

Seems it has some size/efficiency advantages. And maybe it just sounds good (I know this isn't your first amp using it).

 
Good question.
 
Part of the reason is just "because it's there".  It's new, and in interesting part.
 
Compared to normal tubes, it can be operated at very low power - 700mV @ 17mA for the filament, and you can use B+ of 12V with no problem.  The same is true of a few of the "hearing aid" tubes, but they have gotten difficult to find - at least the good ones.  So I see the Nutube as being used in places like portable amp and guitar pedals, places where normal tubes don't work very well.
 
The biggest issue with the Nutube is that it has very low transconductance.  Coupled with the fact that it works best in A2, you really need to put a buffer in front to drive the grid.
 
Within those constraints, the Nutube is quite linear.  And it acts just like any other directly heated triode.  You can tune it by changing bias, plate load, and feedback (if used) to change the resulting characteristics. So, you can build a hybrid amplifier (headphone, power, class D, etc.) that winds up sounding like a single ended triode amp.  I guess that's the "sounds good" part.  Though SE triodes certainly are not everybody's favorite sound.
 
By the way, I've got an article coming up in next month's audioXpress that talks more about it.
 
Pete

 

[Sodacose]

Killer, thanks for the insight.  I'll be watching for the audioXpress article!

 

[pmillett]

TOMB (I guess the first person finishing an amp) has noticed a couple of issues with the amp that I want to make everybody aware of...
 
First, there still can be a bit of a thump when turning the amp on.  This is caused by the time it takes to charge the output caps.  You can eliminate this by changing R3 and R4 from 10k to 1k.
 
You still get some noise when you power the amp off.  I don't think it is enough to damage anything.  If you are worried about it, I guess you can always unplug the headphones like with the old original hybrid and starving student...
 
There is also noise when setting the volume control to zero.  The faster you turn the knob, the worse the noise.  I thought that this was just a cheap volume control pot, but it turns out there is actually some instability that causes some HF oscillation when the volume control is near zero.  Until I figure out the best fix, either move the knob very slowly, or avoid turning it all the way down.  I know, it's not ideal, but I will figure out how to fix it within a few days.
 
Thanks, TOMB!
 
Pete