Marsh headphone amp from Linear Audio

Disabled Account
Joined 2012
SOTA performance

With the dual PS I have, caps right after the PS and then on to the board, excellent sound, by any standard. This with Hifiman HE-560, Foxtex TH900 recabled with Whiplash gold/silver and so on. Being fed by the Chord Hugo or the Sony ZX2.

Did you have any trouble matching the input devices? And, what is your power supply voltage? Did you try a higher voltage? Glad you find the sound excellent by any standard.

It varies from the original Hiraga in that he didnt realize the distortion of electrolytic caps. I changed it to this configuration to eliminate those caps. That was the new and preferred circuit after Hiraga. A few years later, i invented the dc servo concept and told walt Jung..... btw-- Walt Jung first introduced the dc servo via his DIY phono preamp. Now, no one needs to have caps as coupling or in the feedback signal path.

By selecting the jFETs as instructed, you get zero temp coefficient so not even a dc servo is needed. Ultra pure sound.

The sm versions of the jFEts may be a little harder to solder and handle but seems to be what will be most available when the nos is depleted from the market place. The only problem will be in matching them due to small physical size.

Glad you built it and are enjoying the music.


THx-RNMarsh
 
Last edited:
Did you have any trouble matching the input devices? And, what is your power supply voltage? Did you try a higher voltage? Glad you find the sound excellent by any standard.

It varies from the original Hiraga in that he didnt realize the distortion of electrolytic caps. I changed it to this configuration to eliminate those caps. That was the new and preferred circuit after Hiraga. A few years later, i invented the dc servo concept and told walt Jung..... btw-- Walt Jung first introduced the dc servo via his DIY phono preamp. Now, no one needs to have caps as coupling or in the feedback signal path.

By selecting the jFETs as instructed, you get zero temp coefficient so not even a dc servo is needed. Ultra pure sound.

The sm versions of the jFEts may be a little harder to solder and handle but seems to be what will be most available when the nos is depleted from the market place. The only problem will be in matching them due to small physical size.

Glad you built it and are enjoying the music.


THx-RNMarsh

I am running it at 15 volts. I know you said the sweet spot was 18 but I already had one supply at the 15 and just wanted to give each channel a supply.

Yes sm are harder to work with but a jig should be easy to make to do the testing. Soldering isn't that bad, just don't use to much solder, right. :^)
 
Hello jackinnj,

Been reading this thread lately... would it be possible to reduce the gain to 6db without affecting (much) the performance or changing the character of the amp?

also, how critical the PSU quality (unregulated vs regulated, feedback regulated vs series, smps vs linear) for this amp performance?

thanks, david
 
Disabled Account
Joined 2012
I did not try it at lower gain.... I can control level going into the HPA from source or a pot on input.

The PS is important. how critical? Depends on your expectations..... for best possible sound, use the best possible PS. I did not do anything more than simple regulator and it measures and sounds great.

If someone has tried more complex PS, we would like to hear your experience.


THx-RNMarsh
 
I did not try it at lower gain.... I can control level going into the HPA from source or a pot on input.

The PS is important. how critical? Depends on your expectations..... for best possible sound, use the best possible PS. I did not do anything more than simple regulator and it measures and sounds great.

If someone has tried more complex PS, we would like to hear your experience.
THx-RNMarsh

I would like to do a PS with inductor loading. As the field collapses around the inductor, there is a type of regulation going and you have the storage of energy in the inductor. I do this for tube supplies, 150 or higher H. I have 4 150H inductors. I should try it. I am happy with the sound of my amp though, with the dual PS's and added FC caps right after the PS. Very clean, natural, open, and dynamic.
 
Member
Joined 2007
Paid Member
I'm so busy at work I'm basically relegated to lurking here...sigh. However, I finally got the tracking version working that jackinnj hinted to me to try. I am using a 317 with a LT1963a, both for V+ and V- using a dual secondary trafo. (I know, Mongo say CT trafo go BOOM!)

Now I just have to go back to review what are the recommended trafo VA requirements and I should be able to wrap it up soon. I have the board 90% populated. I just need to set up the output transistors with heatsink and remove a biasing diode from each channel. Ohh so close.
 
I did not try it at lower gain.... I can control level going into the HPA from source or a pot on input.

The PS is important. how critical? Depends on your expectations..... for best possible sound, use the best possible PS. I did not do anything more than simple regulator and it measures and sounds great.

If someone has tried more complex PS, we would like to hear your experience.


THx-RNMarsh

Thanks Richard... how much maximum current that could be sent out by your amp (thus to calculate the maximum current supply needed for PS)?
 
Finally got the heatsink mounted and am burning my Marsh in. The sound isn't bad right out of the box. A little muted on the highs and lows but typical of new electronics (and headphone drivers). :p

WOW does that heatsink get hot! I can touch it but not for more than a second. I'll have to borrow a Fluke IR meter to make sure they are running safe. I am only running at +/-15V right now.
 

Attachments

  • Marsh Burn-in.JPG
    Marsh Burn-in.JPG
    319.2 KB · Views: 701
Yes, I removed one of the diodes and it runs MUCH cooler now! I then went from +/-15V to +/-18V and it still stays only slightly warm but sounds much better at the higher voltage. It seems to be taking much less effort now. The sound is more open and dynamic.

Now I just need a better set of cans than my SR60's! :p I was at AXPONA this past weekend and listened to a lot of different headphones. I am REALLY liking the Oppo PM-2 planars. They also just happen to be 32 ohm impedance as well! I still want to hear a set of RS1's before I pull the trigger though.
 
The boards are in! Offer remains the same -- $3 which is my cost (but you'll have to pay the postage).

Layout is a bit different -- I allow for a potentiometer to trim the value of R10, you can use the TO-92 or SMT versions of 2N5457/5460 or 2SK170/2SJ74. The optimal size for the trimmers is the Bourns 1/4 inch variety.

The new potentiometer arrangement, layout, etc:
 

Attachments

  • Marsh.jpg
    Marsh.jpg
    215.9 KB · Views: 583
  • Marsh_P2_Schem.png
    Marsh_P2_Schem.png
    16 KB · Views: 571
  • Marsh_P2.png
    Marsh_P2.png
    26.2 KB · Views: 529
  • Marsh_Input.png
    Marsh_Input.png
    61.3 KB · Views: 534
  • Marsh_Jumper.png
    Marsh_Jumper.png
    24.3 KB · Views: 265