[CatalinM]

Hi Pete,

I did that actually. The last 10-15 hrs of the burn-in (and listening afterwards) were done with bias set around 8.25V.

I think the lack of bass on the AT headphones makes me push the volume up, which accentuates the highs, but either way it's still a bit glassy. Is there anything I could change? Maybe R8/9 or the output coupling caps? I do want to get some Burson v6 opamps in there anyway, although I'm not sure that would improve the situation.

 

[pmillett]

You can try increasing R8/R9. Not sure if it will help or hurt with these headphones. I know with Grados I like to put 50-100 ohms in series or they sound too harsh to me...

Opamps probably won't make much difference.

 

[CatalinM]

I switch R8/R9 from 2.2 to 100 ohms and it had some effect (less on the highs, more on the volume I'd say... now I have to turn the volume pot higher to get the same loudness, which is an improvement in its own). This didn't fix the issue entirely, so I tried HiGHFLYiN9's advice and took a dampener sheet from the Samson headphones and inserted it in the Audio Technicas - surprising improvement.

All in all, it's usable now, event though I'm not entirely satisfied with the solution. I'll go ahead and continue the work on the external power supply and upgrade the opamps.

Doing some critical listening, I'd say that - in my particular combination - it has a pronounced bump in the 2-3khz range, which I'd be happy to get rid of, without sacrificing what's above that. I feel that particular range brings out the cymbals too much and causes fatigue.

 

[Evgeni]

Try moving the bias point off center a bit - set it for maybe 9V or 8V instead of 11V, This will introduce more 2nd harmonic which tends to smooth out the highs...

Pete

That’s an interesting suggestion. Since I’m mostly after “fun” sound rather than neutral/uncolored, what’s the bias with most 2nd harmonics so that I can listen to “tube” sound? Is it as low as possible or the 8-9V is the sweet spot? I understand what’s fun depends on the listener but anyway, what would you recommend as the most “fun” bias?

 

[caenlenfromOCN]

@pmillett here is a diy sort of liek starving student with some modifications that a friend made me, it sounds equal or better than Lyr 3 imo. I know you probably don't care, just thought I'd share since your technically like the person that started this whole thing years ago. wiring is done a lot different, I don't know much about this stuff, but you can see for yourself, pics below. let me know what you think, cheers! It had some noise issues at low volume at first, but then he repaired it for me and now 0 noise floor and no channel imbalance that I can tell, and sounds great with all my cans, even planars. my only complaint is that the vlume knob gains power too fast, so i end at 9am for most of my cans on the knob. i need to get some 600 ohm cans of some kind and see if they sound good with this, im sure they would.

 

[CatalinM]

I've replaced the OPA551's with a pair of Burson Audio V6 Classic and I think they improved the sound quite a bit. On first impression the bass was tighter and soundstage was slightly larger, with more clarity and separation between instruments... everything seems more layered, in a very balanced way. I let them burn in for about 5 days straight, but the evolution was linear and I can't really say how it changed now without doing an A/B.

I'm now curious to replace the C3/C6 from Silmic II to Audio Note Kaisei... but they're not available in 35V and the 63V version is too big. Would it be safe to put in 25Vs ?

Here's a pic of the current state of affairs:

 

[pmillett]

25V caps is close to the limit. If your power supply really measures 24V or less, it will work.

Pete

 

[CatalinM]

I’m working on a highly regulated supply, built around an Amb Labs σ78. It’s probably overkill, but it’ll be a fun build and I’ll be sure it outputs 24V.

However, speaking about C3 ans C6, shouldn’t they have only the bias voltage (when idling), plus minus the signal amplitude? So I’d imagine it should be way lower than 24V unless you max out volume. Please correct if I’m not getting this right.

 

[pmillett]

Right - during normal operation, C3 and C6 are charged to wherever the bias is, normally around 11V. But you also have to consider what can happen at power-up, power-down, and if the output gets shorted. In some cases you may get more than that voltage applied. So the safest thing is to keep their rating greater than the maximum possible voltage that could get applied, even in a fault condition.

All that said, you're right, if your supply is 24.0V then a 25V cap should be OK here.

Pete

 

[CatalinM]

Makes sense. Thanks, Pete!

 

[gdogpwns]

I am building this right now and I am having a peculiar issue. I am not getting any sound out, and when I probe the test points, the Nutube begins to flicker. I am new to this type of thing, where should I begin diagnosing?

If it helps, I was confused and soldered the volume knob to where it bridges to those pads immediately (almost touching) above them. Could this be the problem?

The two glowing sections (is my noob showing yet?) of the NuTube seem to be glowing at different brightnesses and the side closest to the headphone jack flickers very slightly if watched.

Thanks for all of your work by the way. Can't wait to hear this.

EDIT: All of a sudden, the tube begins to glow when I turn it on then fades out. The LED remains lit. I accidentally began probing with my multimeter set to 200V AC instead of DC. Could I have broken something in the process?

 

[CatalinM]

Having your multimeter set to 200V AC shouldn't break anything there. From what you're describing, it seems the filaments are not properly powered - have a look around C12 and C15. Maybe the caps are bad or there's a short somewhere. Or a cold solder joint.

 

[gdogpwns]

Having your multimeter set to 200V AC shouldn't break anything there. From what you're describing, it seems the filaments are not properly powered - have a look around C12 and C15. Maybe the caps are bad or there's a short somewhere. Or a cold solder joint.

I gave a good look around C12 and C15 and resoldered those points but nada.

I've noticed a new symptom. Previously, I would get no audio at all. Now my headphones crackle on power on as well as a high pitched ringing (not microphonics I believe). My right-bias side of the NuTube is not glowing at all now, and the readings from the test point do not change on it no matter how much I turn it. It also pops on power up, and I believe that there is a circuit in place to prevent that.

Thanks again for your help CatalinM.

EDIT: I am now getting audio. I think that I just did a bad soldering job (went out and got a new iron, my last one was falling apart.) But the same part of the Nutube is flickering. While reading the test points of the right bias, I noticed that it climbs from about 1.1V to 2.1V (or thereabouts) then drops back down. I guess this means the capacitor (C15?) is charging then discharging. Any ideas?

 

[CatalinM]

Hmmm, it seems there's still something wrong with your power supply. Use your DMM to measure between the positive supply of the opamps (pin 7 on either of one) and the GND test point. It should read a solid 24V. If that's ok, start measuring the filament supplies, as the general power supply is OK. Use the GND test point and the plus pin on C12 and C15. This should read 700mV, as per Pete's documentation.

 

[gdogpwns]

I gave a good look around C12 and C15 and resoldered those points but nada.

I've noticed a new symptom. Previously, I would get no audio at all. Now my headphones crackle on power on as well as a high pitched ringing (not microphonics I believe). My right-bias side of the NuTube is not glowing at all now, and the readings from the test point do not change on it no matter how much I turn it. It also pops on power up, and I believe that there is a circuit in place to prevent that.

Thanks again for your help CatalinM.

Hmmm, it seems there's still something wrong with your power supply. Use your DMM to measure between the positive supply of the opamps (pin 7 on either of one) and the GND test point. It should read a solid 24V. If that's ok, start measuring the filament supplies, as the general power supply is OK. Use the GND test point and the plus pin on C12 and C15. This should read 700mV, as per Pete's documentation.

Okay, I found something strange. Reading from the GND test point (I measured this from the back of the PCB if that matters) and C12 and C15, it shows ~250mV for each but it also seems to sometimes fluctuate all the way down to ~120mV. Both filaments are not glowing either now. Any ideas?