New PassDIY Headphone Amp (now available)

[flocchini]

XP-10 / Specifications
Overall Gain
-73 dB to +10 dB
Volume Steps
83
Input Impedance
96k bal, 48k single-ended
Output Impedance
1000 Ohm bal, 150 Ohm single-ended

Using SE outs

Input is vinyl with XONO
Gain 40dB
Max Output 20V RMS
Output impedance 300 ohms

Gain of this headphone amp - 6L6 or Meldano please help

2 pico - I have tried other op amps with same results - will look at surrounding components - quite possible I put the same incorrect component in both channels given my board stuffing habits

Thanks

Bob

 

[2 picoDumbs]

Also check for brainfart mistakes eg wiring of rca connections as well as checking quality of all interconnecting signal cables.

What dc offset do you have on the headphone amp?
Just gathering as much information as possible.

 

[flocchini]

pico

I just completed my M2 so I substituted it in the chain - works well

Checked input rca connectors and 4 pin XLR - no shorts

DC offset left channel is is 175mV and right channel is 50mV - I expected better.
Anyone who has built this have measurements?

I have tried other op amps with same results.

Will start checking all resistors.

Thanks and best

 

[2 picoDumbs]

175mV seems quite high for a headphone amp. I'd personsally want below 10mV.

Are you using 2sk2013/2sj313 on output?

 

[flocchini]

175mV seems quite high for a headphone amp. I'd personsally want below 10mV.

Totally agree

Are you using 2sk2013/2sj313 on output?

Yes

 

[AndrewT]

Some sensitive headphones could be driven to near their Xlim stops with an offset of 175mVdc.

I would be expecting <2mVdc of output offset for a headphone amplifier.

 

[WalterW]

Did you use these output stage:

I found out that the R6, R15, R23 an R24 which are all 10K, must be real close, to have an almost zero output offset. Especially R6 and R24 if I remember correctly.
Also the V+ and V- of the PSU must be real close and stable.
Variations between v+ and V- give offset.

When I measured a few 10K resistors and picked the closest together, I now have output offset around 0.8 - 1mV at both channels.

 

[flocchini]

Thanks Walter

I will look at my PSU.

I did change out the 2013 and 313 to no avail.

Away for 10 days or so - will attack it when I get back.

Best

Bob

 

[2 picoDumbs]

You can fix dc offset by changing R6 to 9.8k and add 500 Ohm trimpot.

That doesn't help you fix the main issue though.

 

[flocchini]

You can fix dc offset by changing R6 to 9.8k and add 500 Ohm trimpot.

That doesn't help you fix the main issue though.

thanks pico - nice solution to that problem

 

[tg123]

The project is not dead, due to life outside this forum I kicked the can down the road quite a bit further than I intended... Any delay (I.E., the last 6-7mo or so...) is 100% mine.

I saw your comment and realized that it was languishing on my shelf, dusted it off and started messing with it again over the last 2 days.

Long story short, and not trying to offer any excuses for my dropping of the ball with this project, I think I've finally found a configuration that is bulletproof and sounds great.

So hang in there, if it all works I'll have a substantive update soon.

Any updates?

 

[RobLK]

Well since I am waiting for boards for this amp and also only have a few 2sk2013 and no 2sj313, I removed about half of the passdiy headphone circuit and made it work with the one transistor. Well at least it works in simulation. I wonder if it will work in the real world?

 

[RobLK]

I've been using ltspce for only a very short time now so I don't know it very well but I will take a guess that the FFT analysis for this circuit is pretty optimistic. The Passdiy amp looks even more optimistic. I assume that accurate FFT circuit simulation can't be done. If not I'll just have to wait and measure the passdiy board with the Arta program, I'm sure it will show a more realistic
picture.

 

[JimT]

It will ultimately depend on the accuracy of the models, but I do suggest you retry it with a more realistic 2.83V AC on the input.

Jim

 

[RobLK]

About the best the circuit can do and still remain a sine wave is 9v peak to peak with 1.8 in.

 

[meldano]

Some impressions

 

[meldano]

Discrete OPA from generg
Compare to OPA627

 

[carsten.witt]

And what did you think?

 

[SomekPoland]

Any progress on diyaudio.com PCB for this project?

 

[meldano]

And what did you think?

In my setup (AKG K812) the OPA627 is the better one.