[jobous]

Beautiful !!! Nice work.

 

[TravAndAlex]

Beautiful !!! Nice work.

Thanks! I fricking love DIY.

Just because I took a badass picture.

 

[jobous]

I'm trying to understand which pushbutton to choose. I mean which voltage, take into acount that there is already the R5 (10K).
The pushbutton is from Onpow (16mm_LAS2GQ series). Anyhelp here ??!!

 

[jobous]

I just finished my board, everything seems to be OK. Power on and the i see the light !!!!
Still some problems to solve , expecting enclosure this week to finish the amp.
Set bias to 11V, tried to listen but i was unlucky , the one of the two rca cables was faulty and i dont have another one.
I hear a constantly noise (increasing-decreasing using volume), but i dont know if is a ground issue....

 

[Pricklyears101]

Anyone knows the output impedance of these?

 

[TravAndAlex]

A few updates to my build. Copper lined the case to combat RFI. Installed a Khadas Tone Board DAC internally (which of course required fabricating a new rear panel). Added a yellow power LED for the DAC on the right side of the front panel to balance the amp power LED.

Overall I’m really impressed. Yup. Rings like a bell when the relay clicks on. Stock volume pot is noisy at either extreme but good at any reasonable listening levels. Sound really great with anything from PortaPros, to HD 650s and even plenty of power for 600 Ohm Beyer T1s. Color me very impressed.

 

[jobous]

Hi again. Now a have a major problem. I decide to change the cables from the pcb to the pot (smaller cables) and now the sound is very bad.
It sounds like it is very far , like a well and there isnt any bass at all !!!
I dont now why this happened. Has to do with the pot or the pcb (cause when i try to unsolder the old ones and solder the new cables , i have to say i didnt do a fine job there) ???
Or something else ??
I changed again with the original cables but there wasnt any difference.
Any help would appreciated. i'm very dissapointed cause except the ground issue the sound was amazing....

 

[jobous]

Problem resolved !!!!! The omp voltage was 10.85 V and when i correct it back to 11.00 V , the sound came back to normal. I wasnt aware that is so sensitive to such a small voltage difference (0.15 V).

Now all is in place !!! Amazing sound ,no ground issues , i will enjoy it for sure.



I have tested it only with my Sennheiser HD25. I'm looking forward to try with the HD58X ...

Thanks Pete for giving me the change to build my first amp.

 

[Desert Rat]

My NuHybrid Headphone amp makes a noticeable pop in the headphones when switching off the power. Power up delay works fine. Loud enough to be alarming. Build is in a DIYaudio store Galaxy chassis with jacks and input pot (Alps RK27, 10K audio taper) off board.
Is this unusual? Any suggestions? Pardon if this has been covered previously, I'm new to the forum.

 

[TravAndAlex]

Wrote up my build log with tips, tricks and info as an article for Headphonesty if anyone is interested. Designed to be a one stop shop for info.

If anyone thinks I missed anything important, just let me know!

 

[Kitchener]

Wrote up my build log with tips, tricks and info as an article for Headphonesty if anyone is interested. Designed to be a one stop shop for info.

If anyone thinks I missed anything important, just let me know!

Brilliant write-up.
Great work, beautiful amp.

 

[pmillett]

My NuHybrid Headphone amp makes a noticeable pop in the headphones when switching off the power. Power up delay works fine. Loud enough to be alarming. Build is in a DIYaudio store Galaxy chassis with jacks and input pot (Alps RK27, 10K audio taper) off board.
Is this unusual? Any suggestions? Pardon if this has been covered previously, I'm new to the forum.

It is normal to hear a pop when the power is turned off. It's easy to delay the operation of the output relay to kill the turn-on pop, but when you power off, the voltage collapses fast and it takes some time for the relay to drop.
I measure about a 1.5V pulse at the output that lasts about 10mS. I don't think that can damage any headphones, though it would be quite loud in low-Z headphones or IEMs.

It is possible to speed up the delay circuit and improve the pulse a little bit, maybe dropping it by half. But it will still be pretty loud. You can do this by changing R2 to 150k and D2 to a 20V zener, 1N5250.

Otherwise, about all you can do is unplug the headphones...

 

[Desert Rat]

I'd like to share my mods to this headphone amp. I don't seem to have permission to add photos to this post, so here is a text only description:

NuHybrid Headphone Amp Mods

After building and listening to the Nelson Pass B1 Korg preamp and comparing it to the Pete Millett's Nutube Buffer preamp, I found I liked the sound of Nelson's Korg tube setup a bit better. I figured out how to make Pete's Buffer sound more like Nelson's B1 Korg and then it was onward to the NuHybrid headphone amp.

Not obvious is that all 3 circuit designs invert the signal. This happens at the NuTube itself. This is easy to correct with loudspeakers by reversing the loudspeaker/amplifier connection wiring but not easy with headphones unless you want to take them apart and rewire them. Listening tests confirmed that I preferred the correct polarity with loudspeakers, so I decided to add input transformers with reversed primaries to the NuHybrid to also correct that circuit.

I first experimented with some inexpensive Triad TY-250P transformers I had around just for evaluation purposes. Imaging was very much improved but their lack of shielding, excessive phase shift in the low bass and higher THD proved unacceptable. I ended up purchasing and installing a pair of Jensen JT-11P-1 1:1 high quality line input transformers and installing them with the primary winding connections reversed.

Here is what I did to modify the NuHybrid Headphone amp to compensate for the polarity inversion and to have it sound more like the Pass B1 Korg. The build was done in a DIYaudio Galaxy chassis with all jacks, power switch, LED and volume pot mounted off the PCB and on the chassis panels.

1. The Jensen input leads are configured as unshielded twisted pairs connected directly to the isolated RCA input jacks. Input polarity is reversed with respect to the transformer output. The RCA jack shells (grounds) are tied together and connected to circuit ground to eliminate hum when a source is connected. The output leads of the transformers are configured as unshielded twisted pairs and connected directly to the volume pot inputs. The transformer can shields and internal electrostatic shields are connected to circuit ground at the input side of the board. I provided a direct chassis ground connection at that point a with the series resistor/cap to chassis ground connection shown in the Jensen data sheet. A dedicated ground wire was run from the volume pot ground terminals to circuit ground at the input side of the board. Since I was using a 10K pot, the Jensen data sheet damping network was not required. Even though my pot is an Alps RK27, it has excessive noise at the extreme CCW rotation, which seems to be a problem with 10K pots in general. If I end up changing it to something like a 50K, I'll add the recommended damping network.
   
   
2. I bypassed the on-board PCB volume pot connections altogether and went straight from the pot to some slightly larger Kemet 220 nF input caps that were then connected to the PCB at the C20 and C21 locations. Shielded coax connections from the pot to the caps input sides are grounded only at the PCB end.
   
   
3. I changed the NuTube plate resistors R13 and R14 to 332K.
   
   
4. I set the plate voltage at TP1 and TP2 to about +9.75 volts. I found this is best adjusted while listening to music and set to your preference. I also found that anything over +10.5 volts made the magic go away.
   

So how does it sound? With the resistor and plate voltage mods, I think it sounds better than the stock 475K plate resistors with voltage set closer to half-supply. Smoother, with no loss of dynamics or detail. Correcting the signal polarity with the transformers made a very significant improvement in detail and imaging. Yes, they cost as much as the PCB, tube and board components combined but once you hear the difference, there is no going back.

 

[gaz2613]

I'd like to share my mods to this headphone amp. I don't seem to have permission to add photos to this post, so here is a text only description:

NuHybrid Headphone Amp Mods

After building and listening to the Nelson Pass B1 Korg preamp and comparing it to the Pete Millett's Nutube Buffer preamp, I found I liked the sound of Nelson's Korg tube setup a bit better. I figured out how to make Pete's Buffer sound more like Nelson's B1 Korg and then it was onward to the NuHybrid headphone amp.

Not obvious is that all 3 circuit designs invert the signal. This happens at the NuTube itself. This is easy to correct with loudspeakers by reversing the loudspeaker/amplifier connection wiring but not easy with headphones unless you want to take them apart and rewire them. Listening tests confirmed that I preferred the correct polarity with loudspeakers, so I decided to add input transformers with reversed primaries to the NuHybrid to also correct that circuit.

I first experimented with some inexpensive Triad TY-250P transformers I had around just for evaluation purposes. Imaging was very much improved but their lack of shielding, excessive phase shift in the low bass and higher THD proved unacceptable. I ended up purchasing and installing a pair of Jensen JT-11P-1 1:1 high quality line input transformers and installing them with the primary winding connections reversed.

Here is what I did to modify the NuHybrid Headphone amp to compensate for the polarity inversion and to have it sound more like the Pass B1 Korg. The build was done in a DIYaudio Galaxy chassis with all jacks, power switch, LED and volume pot mounted off the PCB and on the chassis panels.

1. The Jensen input leads are configured as unshielded twisted pairs connected directly to the isolated RCA input jacks. Input polarity is reversed with respect to the transformer output. The RCA jack shells (grounds) are tied together and connected to circuit ground to eliminate hum when a source is connected. The output leads of the transformers are configured as unshielded twisted pairs and connected directly to the volume pot inputs. The transformer can shields and internal electrostatic shields are connected to circuit ground at the input side of the board. I provided a direct chassis ground connection at that point a with the series resistor/cap to chassis ground connection shown in the Jensen data sheet. A dedicated ground wire was run from the volume pot ground terminals to circuit ground at the input side of the board. Since I was using a 10K pot, the Jensen data sheet damping network was not required. Even though my pot is an Alps RK27, it has excessive noise at the extreme CCW rotation, which seems to be a problem with 10K pots in general. If I end up changing it to something like a 50K, I'll add the recommended damping network.
   
   
2. I bypassed the on-board PCB volume pot connections altogether and went straight from the pot to some slightly larger Kemet 220 nF input caps that were then connected to the PCB at the C20 and C21 locations. Shielded coax connections from the pot to the caps input sides are grounded only at the PCB end.
   
   
3. I changed the NuTube plate resistors R13 and R14 to 332K.
   
   
4. I set the plate voltage at TP1 and TP2 to about +9.75 volts. I found this is best adjusted while listening to music and set to your preference. I also found that anything over +10.5 volts made the magic go away.

So how does it sound? With the resistor and plate voltage mods, I think it sounds better than the stock 475K plate resistors with voltage set closer to half-supply. Smoother, with no loss of dynamics or detail. Correcting the signal polarity with the transformers made a very significant improvement in detail and imaging. Yes, they cost as much as the PCB, tube and board components combined but once you hear the difference, there is no going back.

Are you saying that for the line in and line out connections on the board the positive is actually the negative and the ground is the positive? That's my understanding of reversed signal polarity. If that's the case then that would mean the signal would be connected to ground if the negative is actually the positive surely or am I missing something?

 

[tomb]

Personally, I think someone's getting mixed up, here (could be me, though!).

AFAIK, inverting the signal is actually what allows negative feedback. If the signal was not inverted, a feedback connection back to the input would blow up, become unstable, runaway, etc. It is the inversion of the signal that allows the feedback loop to close, and thereby work. In other words, the music signal is inverted by the OPA551 opamp in the NuTube Millett Hybrid circuit. That signal becomes negative. It's then fed back into the input side of the opamp as a negative music signal. Now the opamp has two signals that are input, one positive and one negative. The opamp is then able to compare/correct to minimize the voltage offset between the two. That minimizes distortion and noise. As I understand it, that's how negative feedback works.

I am not familiar with Nelson Pass's Korg preamp, but if it is simply an amplifier circuit without feedback, I'm guessing that trying to eliminate the signal inversion at the input to the NuTube hybrid is not the way to go about creating a zero-feedback circuit.