[pmillett]

OK, I made a discovery - the microphonic "ping" at power-up has nothing to do with the relay engagement.  It is noise induced from flipping the power switch.
 
I tried leaving the power switch turned on, and instead turned the amp on and off by switching the AC to the adapter.  The relay pulls in after 10 seconds. No "ping" noise (well, none in HD600s - I'm sure there is still some noise in efficient headphones or IEMs).
 
I also tried moving the power switch very slowly... also no ping sound.
 
It is true that if the delay is lengthened, you won't hear it as much when you flip the power switch.  But I guess I don't quite understand why it matters.  Let's say it takes 20 seconds (from the time you hit the power switch) for the ringing to calm down.  You can make the power on delay more than 20 seconds and you won't hear it.  Or you leave the power on delay at 10 seconds (which is where it is now) and you hear a decaying ringing for 10 seconds.  It's not that loud...
 
Pete

 

[smellor]

Hey Pete,
 
I bought your board+nutube and am waiting to receive the parts from Mouser. Looking forward to build it. I am a new builder, so i decided i would use this build to try my hand at customizing the case. Having positive experiences with the cases from Hifi2000 (an italian vendor) who supplied the case for a recent diyaudio amp build, i bought a relatively large economy case from them, which i can play around with. One of the things i was thinking was to mount the nutube on the outside top of the case, since it looks so nice, and connect it through holes in the top via wires insulated with heatshrink tubing. I wonder if this would affect the performance to any great extent? My feeling says probably not, but i am a novice in this area so i would rather take the word of someone who actually knows what hes talking about Apart from the leads connecting it to the PCB, are there any exposed parts that carry voltages in the nutube?
 
Best,
Silas

 

[HiGHFLYiN9]

Any changes to the NuTube buffer based on discoveries with the headphone amp? I know most implementations will not have a volume knob, but there would still be the on-off switch.

 

[pmillett]

  Hey Pete,
 
I bought your board+nutube and am waiting to receive the parts from Mouser. Looking forward to build it. I am a new builder, so i decided i would use this build to try my hand at customizing the case. Having positive experiences with the cases from Hifi2000 (an italian vendor) who supplied the case for a recent diyaudio amp build, i bought a relatively large economy case from them, which i can play around with. One of the things i was thinking was to mount the nutube on the outside top of the case, since it looks so nice, and connect it through holes in the top via wires insulated with heatshrink tubing. I wonder if this would affect the performance to any great extent? My feeling says probably not, but i am a novice in this area so i would rather take the word of someone who actually knows what hes talking about Apart from the leads connecting it to the PCB, are there any exposed parts that carry voltages in the nutube?
 
Best,
Silas

 
I think the biggest risk would be noise pickup.  Try and keep the wires as short as you can.
 
There is no exposed metal on the Nutube package other than the pins, so they are all that needs to be insulated.
 
Pete

 

[pmillett]

  Any changes to the NuTube buffer based on discoveries with the headphone amp? I know most implementations will not have a volume knob, but there would still be the on-off switch.

 
I haven't tried to get the buffer to oscillate, but it might be unstable with very low DC resistance across the inputs.  This would apply only to the input buffers - the output buffers have a high resistance on the gate (the Rp of the Nutube is pretty large) so they should be fine.
 
To be safe, you can either change to 10k resistors on the source terminals of the 2N7000s - that seems to lower the gm enough that there's no oscillation - or change to 2N3904 + 10k like on the headphone amp.
 
Pete

 

[tedk84]

I just finished my build and ran into a problem:

The amp started, but one side of the headphone was quiet, while the other side sounded normal. one side of the tube isnt nearly as bright as the other one. I was/am able to set 11V for both test points. Can you guys help me? I used two jumper cables to work around the fact that i ordered the wrong relay (5.1mm instead of 2.54mm spacing) -.- ... When i cross the cables, the quiet side changes left<>right.

http://imgur.com/a/4VPl2

 

[pmillett]

I can't see anything wrong from the photo.  I assume you've double-checked solder joints for shorts and opens...
 
It is odd that you can still adjust the bias to get 11V and there is a difference in  the glow.  That indicates a big difference in the current through each triode.
 
Can you check the voltages on the tube pins (to ground)?  Starting at the top (towards the input jack), you should get:
 
1 - 0.7V
2 - 0.7V
 
3 - 2V
 
4 - 12V
5 - 0V
6 - 0V
7 - 12V
 
8 - 2V
 
9 - 0.7V
10 - 0.7V
 
Voltages are approximate, some variation is expected.  Let me know what you measure, that may give a clue.
 
Pete

 

[tedk84]

Oh god, is was really late last night....
I mistook L_out for GND and GND for L_out. After i checked this morning, GND>L_Out was 11V, and GND>R_out was ~22V (...). So i fixed that and checked everything - works.

Thanks anyway Pete

 

[pmillett]

Ah, OK, good.  That makes sense.  Was there a short on the plate, or just bias adjustment?
 
Pete

 

[tedk84]

no, just bias adjustment, no shorts

 

[mitsu763]

Received my board today. Not sure when I'll get to it but this will be my first build.

Is the Bill of Material up to date on Mouser?

 

[pmillett]

Is the Bill of Material up to date on Mouser?

Yes, it should be good.

Pete

 

[ericj]

I swear i saw a thread either here or diyaudio where someone asked Pete if he'd make boards for the NuTube buffer + Class-D amp referred to on the website, Pete said it has an SMT chip so that would be problematic, etc. 
 
Can't find it. 
 
Anyway. I humbly raise the existence of the TPA3122D2N. DIP-16 package, 2x15w or 1x30w into 4 ohms per chip (with typical class-D power vs. thd curve) at 28v single supply, $2.88 each at Mouser. Though of course there's the output network to consider, which adds a few bucks. 
 
A couple of these in a PBTL configuration are louder than the classic ta2020 T-Amp with no surface mount parts or heatsink required. Plenty of oomph for near-field, dormitories, office-go-boom, etc. 
 
Not that i have actually heard one. But it's substantially similar to the other current TPA3xxx chips, which are generally well regarded in the low-volt low-dollar arena. 

 

[pmillett]

  I swear i saw a thread either here or diyaudio where someone asked Pete if he'd make boards for the NuTube buffer + Class-D amp referred to on the website, Pete said it has an SMT chip so that would be problematic, etc. 
 
Can't find it. 
 
Anyway. I humbly raise the existence of the TPA3122D2N. DIP-16 package, 2x15w or 1x30w into 4 ohms per chip (with typical class-D power vs. thd curve) at 28v single supply, $2.88 each at Mouser. Though of course there's the output network to consider, which adds a few bucks. 
 
A couple of these in a PBTL configuration are louder than the classic ta2020 T-Amp with no surface mount parts or heatsink required. Plenty of oomph for near-field, dormitories, office-go-boom, etc. 
 
Not that i have actually heard one. But it's substantially similar to the other current TPA3xxx chips, which are generally well regarded in the low-volt low-dollar arena. 

 
I think it was on DIY Audio...
 
Yup, I did think of that class D as a possibility.  Pretty wimpy though
 
Another option would be to re-design the board I have to make everything through hole except the class D (MP7770).  50 watts is nice.
 
My thought would be to pre-solder the IC onto the board (I have a small reflow oven, and placing only one part is pretty easy).
 
Thoughts?
 
Pete

 

[Heffa]

   
Another option would be to re-design the board I have to make everything through hole except the class D (MP7770).  50 watts is nice.
 
My thought would be to pre-solder the IC onto the board (I have a small reflow oven, and placing only one part is pretty easy).
 
Thoughts?

 
I myself wouldn't mind a SMD-variant either...
SMD soldering isn't that hard once you get the hang of it unless you go for REALLY tiny 0201 or such, and the boards usually looks really nice (to me at least).
0603 package and up isn't that hard to do manually imho.
 
(and I do mean imperial sizes, 0603 metric I have never tried 

 )