Looking for a new project. Anything similar to the Schiit Lyr?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
There are many similiar designs posted on the internet, basically a tube gain stage with a mofset output. The good thing is a high Vrms output since the classic OPT is replaced by a mosfet output, the problem is dc servo or caps (pick your poison) and tube speaker amp level distortion. Do you really need that much power? If you do I would recommend Beau's hybrid aikido thread in the tube forum, he has tackled mainy of the problems with hybrid amplifiers.
 
The nice thing about Beau's method using the Tubmos is Lateral Mosfets, you aren't going to find those on any store bought headamp. Also see DDE's page DDDAC 2000 (look for the tubmos page). Also someone had a DHT-Mosfet headamp design posted here last year that simulated nice don't know if anyone built it though. Its an interesting subject 'cause you gain a ton of power by substituting silicon(mosfets) for iron (output transformers) and you save money.
 

rjm

Member
Joined 2004
Paid Member
I was pretty happy when I test drove the Lyr (LYR Schiit Lyr 6W Hybrid Headphone Amplifier [] - $449.00 : Schiit Audio, Headphone amps and DACs made in USA.) and it was nearly exactly what I was looking for except not DIY friendly.
Any tips as to what I should search for?

So other than having tubes and MOSFETs, any hints about the actual circuit?

I see two twin-triode tubes and four MOSFETs and what looks to be two power transformers (high, low voltage most likely). Blurb says 6 W / 32 ohms, that means push-pull output, and it also says zero feedback. Something called "Dynamically Adaptive Output Stage" which could be a number of things ranging from very clever to the pathetically simple but end result is a transition from single ended at low output power to push pull operation at high power.

Burns 30 W, can output 40 V p-p (~+/-24V rails on the output stage)

So what are you looking for exactly? Since a tube voltage stage and a MOSFET output stage are two independent circuit blocks, you can take any MOSFET buffer (singled-ended or push-pull) and any configuration of tube voltage stage and roll your own at will.

You can be be a little clever and weld the two together, like this , but to be honest you don't gain all that much from doing so.
 

Attachments

  • lyr_06.jpg
    lyr_06.jpg
    190.6 KB · Views: 540
Last edited:
I pike up a Lyr, its a nice design. I think a DIY version would cost considerably more but the reward could be great. I think the tube stage could be improved upon with B+, better tube operating point, etc. And obviously a better power supply for the mosfet stage. Obviously this will cost much more money that the Lyr, so I wouldn't want you to think of a diy Lyr as being a means to save money.

Basically following the hybrid aikodo topology. Where we have a near perfect preamp stage followed by a push-pull lateral mosfet buffer. The tubmos is one option but a better one may be to use the Thor output (using the jfets just to bias the mosfets) , the jfet should be quiter than the TL431 in the tubmos.

I think the main goals of the project should be better Power supplies/transformer, betterdrive from the tube stage (operatng point topolopgy), better tubes than the 6Dj8's and lateral mosfets. Would be great to get in all on one PCB. A lot of fertile ground for the ortho owners.

My time is lmited but I could put together a schematic and simulate with LTspice. I'm still on a lot of OT. But desiging board is too time consuming with my work schedule.

If there is interest I think we should start a new thread.

Lets face it lowish distortion >2W into 38 ohms (non balanced) is a project worth pursuing for the group and the hard work has been done by beau with his speaker hybrid.
 
I did a little probing on my new Lyr, the tube stage is a simple common cathode, resistive loaded plates with the shared cathode ccs, operating point with my less than fresh amperex orange globes is 110V anode, 2.7V cathode, which gives me about 3mA per tube . Its a good design, not some enemic chinese SRPP run 100V B+. Its quiet with Grados so the B+ regulation must be good stuff. The p-p mosfet output is well laid out. The coupling cap in front of the output stage is a decent Wima 4.7 MKP4. If I hear a room for improvement the bass seems a bit too snappy/harsh at high high power levels (fine at normal levels).

I guess a DIY version could be done for about the same price without a whole lot gained. Would do a little better with a more linear operating point, more like 125V/5mA.

I would be more interested in an no expense tube hybrid headamp. Maybe tube rectified/Salas HV B+ to 4p1l (coleman reg), to lateral mosfet p-p type buffer. Would be expensive and time consuming but could be worth it. I am surprised there isn't much interest in building a DIY ortho amp, everyone seems content to plug them into antique Sansui receivers:rolleyes:
 

rjm

Member
Joined 2004
Paid Member
Tube+MOSFET hybrid headphone amps are bit ... uninteresting. They are basically two fully independent sections, one high voltage, low current, the other low voltage, high current. Since the don't ideally share a common power supply, the circuit design quickly devolves into a tube preamplifier and a MOSFET headphone buffer that just happen to share a common chassis.

As for the Lyr, I'm sure it's a solid-by-DIY-standards design, given it's heritage. I'm most interested to know how the dynamic biasing is done:
It works by sensing the current flowing through the output stage, and dynamically reconfiguring from Class-A single-ended to push-pull Class AB.
I've seen amps with this general design claim surface a few times over the years, but never seen a circuit schematic to tell me exactly what is being done.
 
Last edited:
Tube+MOSFET hybrid headphone amps are bit ... uninteresting. They are basically two fully independent sections, one high voltage, low current, the other low voltage, high current. Since the don't ideally share a common power supply, the circuit design quickly devolves into a tube preamplifier and a MOSFET headphone buffer that just happen to share a common chassis.

As for the Lyr, I'm sure it's a solid-by-DIY-standards design, given it's heritage. I'm most interested to know how the dynamic biasing is done:
I've seen amps with this general design claim surface a few times over the years, but never seen a circuit schematic to tell me exactly what is being done.

Not all hybrids are unitnteresting, with these new orthos we need to swing upwards of 17Vrms, that equivalent to a 35watt amplifier for speakers (without has much current delivery). If you had looked at the EHHA it is not a tube stage plus separate buffer. Its a differential input-VAS-currentbuffer all shared powersupply with global feedback, just has a tube differential input instead of jfet.

Even the lower model orthos need a solid 6Vrms swing into 38 ohms which is a challenge for SET tube folks.


I think the dynamic biasing just means the output stage is push-pull biased around 100mA, and when the volume/load high enough it goes into class B just like any heavily biased class A push-pull stage. But I didn't look at the mosfet stage, the Lyr does fit your tube preamp + mofset stereotype.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.