40+ watts SET, cost is not a problem

If you want something a bit less polite than Audio Note, the Atmasphere linestages and preamps are quite good--very clean, clear, and not overly warm, yet they don't sound harmonically bleached out.  

You did not specify what features you require in the linestage.  To me, remote control is not just a convenience, it is a necessity.  There tends to be a fairly small range of volume one wants to listen to music, and that setting can change even within a single recording.  It is quite hard to tell when the volume is right unless you can do it with remote control while sitting in the ideal spot.  With tube gear in particular, a balance control is essential.  The system will inevitably develop minor channel imbalances and one can go mad chasing down the cause or instantly cure the issue with judicious use of a balance control.

I liked what I heard and saw with the Backert linestages--nice sound, simple design, remote volume control, balance control, and serious people backing the product.

@larryi I dont have any particular need for the remote. Considering that we will go with a Gaku-on, what is critical is a well matched preamp which doesnt short change or create any sonic anomaly (due to character mismatch). Many people say it is best to get the pre-power from the same brand for this reason. Do you think it is okay to get a different brand of preamp without auditioning it with the power amp in question, that too when we are building it ground up? 

I have heard quite a few different brand set ups and I don't think it is necessarily a problem.  Any purchase without an audition in the particular system poses some risk.  It is not necessarily the case that a general sound of a component will carry forward into every system so any prediction on compatibility is a bit fraught.  Perhaps your best best is working with a custom builder who can "voice" the component/system after the purchase.  

I conquer with you completely. There is something so natural about the way SET present music. What it does with distortion is true and theoretical but it must be preserving something vital to make music sound so "right".

Having said that I am totally open to exploring PP amps. @atmasphere will your Novacron work well with a load like TAD CR1x ? Normally OTLs also need friendly loads. Hence I ask

@pani That speaker is not tube-friendly!! Being both 4 Ohms and sensitivity of only 86dB (meaning its efficiency of 1 Watt 1 meter is really only 83dB, which is borderline criminal), you need a lot of power! If you are only listening near field you might be alright, but if in a living room you’ll need 400 to 1000 Watts in most cases.

We discontinued the Novacron a couple of years ago- they were always limited editions. To use a set with those speakers you would need a set of ZEROs, which I think would be a good idea with any SET as well. But that won’t solve the power issue!

So I really think you need to rethink this, starting with a different speaker if you are committed to lower powered tube amps! FWIW, our class D might be alright on that speaker (I have reservations since it only makes 250 Watts into 4 Ohms at clipping); it has a distortion profile that is very similar to that of SETs although its quite a bit lower distortion overall.

And that brings me to something else you might want to understand:

The distortion characteristic of any amplifier is also the sonic quality of that amp. Put another way, distortion is far more audible than most people realize. When you compare amps (unless there is a frequency response error that might also accompany an amp of high output impedance) the differences you hear are the differences in distortion.

As a result it does not matter what the technology of the amplifier actually is; what matters a lot is how the amplifier makes distortion. The reason SETs sound the way they do is because of a quadratic non-linearity which results in a pretty prodigious 2nd harmonic, followed by a 3rd which is considerably less but also prodigious with respect to the higher orders, which, if the amp is designed properly, will fall off on an exponential curve as the order of the harmonic is increased. The 2nd and 3rd are able to mask these harmonics, resulting in a smooth sounding amp.

I hope you can see here that we really can measure what you hear and that it really is all engineering; the real issue is whether the designer of an amplifier understood this fact.

If you have a PP zero feedback tube amp that is fully balanced and differential from input to output, the even orders cancel, leaving the 3rd as the dominant harmonic (which is treated by the ear much the same as the 2nd), still able to mask higher orders. Because the even orders are cancelled, there is less distortion compounded from stage to stage in the amp so distortion is inherently lower. But you still have the exponential decay of higher orders, only now based on a cubic exponent, so they fall off at a faster rate than seen in an SET. That is why such amps are even more relaxed and also more transparent, since distortion obscures detail.

If you have a PP amp that uses single-ended input circuitry (like most PP tube amps) you combine the cubic and quadratic non-linearities, which tends to augment the 5th harmonic (and this has nothing to do with poorly applied feedback, which I touched on in an earlier post). I think this is (and understandably so) why the SET crowd disdains PP. But not all PP amps are built that way!

This is the tip of the iceberg!

In conclusion I really think what you are trying to do is a Sisyphean task. An SET simply will not work and all PP tube amps will be on shaky ground as well.

I know this isn’t what you were asking for, but your best bet here really is a class D, although you should keep in mind that class D amps vary in sound from one to another more than tube amps do- so the idea that you’ve heard one you’ve heard them all is false. The reason I say a class D amp is your best bet with that speaker is because you need the power, and some class D amps are smoother than any other kind of solid state and can be just as involving and spacious as the best SETs. I mentioned part of why this can be so in an earlier post. If you wish elucidation I can provide it.