Compare: Modwright LS 36.5 & Audio Research LS26

"Can you explain the differences between the standard dual differential balanced circuit and the implementation Dan
chose for the LS 36.5? Do other designers use the same approach?"

Probably not well enough to do any good. LOL.

I just tried to write a nice, succinct explanation but, after rereading the several paragraphs a few times, I decided I was way out of my depth. In any case...

The 36.5 uses the Electraprint transformers as phase splitters to produce the balanced signal outputs. Using a dual differential circuit would have produced the same balanced output but then that would be a whole different beast of a pre, wouldn't it?

I'll assume Dan didn't want that for reasons of cost and complexity and purity of design etc...

But I'm guessing.

The reasons for having a transformer coupled output is pretty straight forward. You get a balanced output (for CMRR) and you get a lower output impedance for driving amps with low input impedances. Are there compromises to doing it this way? Probably, there are always compromises, but someone with more knowledge than I have needs to step in to explain what they might be.

As for others doing things this way, I assume it's the standard way to get a true balanced output from a SE circuit.

Would I could be more helpful...

We chose output transformers to create the balanced outputs (and serve the SE (RCA) outputs also), for the following reasons:

1) OPT (Output Transformers) allow us to provide a MUCH lower output impedance (100ohm), over the standard 600ohm output impedance of our SE circuit. Compare this to 1K + for most preamps. This allows for MUCH better integration with amplifiers of ANY kind, even those with 10K input impedance or lower.

2) Obviously the output transformers enable us to have differential outputs and phase switching. The same would be true had we used fully differential circuitry. The transformer provide other benefits.

3) Rather than attenuate the gain at the input of the circuit and limit bandwidth, we are able to operate the circuit at full gain and attenuate output signal via the OPT's. In this case, we chose 4:1 stepdown configuration, custom-wound for us by Electraprint Audio. In addition to attenuating signal at the output, noise is also attenuated noise by the same factor. This makes a quiet circuit even 4 times quieter.

An important clarification here. The OPT's in this circuit do not see DC voltage. The SE circuit is capacitor-coupled, hence the OPT's see purely AC signal, thus not suffering from typical bandwidth limitations inherent in many SET amplifier designs. An SET amp's OPT's block DC, thus causing typical bandwidth limitations, between 20Hz and 20Khz. The OPT's we use are flat to 60Khz and do not limit bass response to 20Hz.

So, what are the drawbacks?:

1) Lower noise floor.
2) Lower output impedance.
3) Appropriate gain levels.
4) BW from 20Hz - 60Khz.

Sorry, those aren't really drawbacks.... I don't feel there are drawbacks to this approach. Given the cost of good OPT's, it is hardly a cost-savings. I am sure that people have opinions about the use of transformers in the signal path. Bandwidth limiation is not an issue here, as it can be with SET amps. Also consider the fact that many consider transformer-based volume controls to be superior to any other form of attenuation. I would have liked to use them in the 36.5, but the impact on unit cost would have been prohibitive.

I am not trying to be sarcastic about the negative effects of transformer coupling. I simply don't see any in this design and the benefits are significant.

Our amplifier design (KWA 150) is actually transformer INPUT coupled AND fully differential. I.e. we use both a transformer and fully balanced circuitry. The reasons for this go beyond the depth of this discussion, but if I felt that transformers limited sonics, we certainly would not have used them in our amp design. One additional benefit of this approach to our amp design, is that both RCA and XLR inputs are treated as fully balanced internally, rather than forcing the amp to operate in SE mode due to RCA inputs.

I hope that this helps explain the reasons for our choice of balanced topography in the LS 36.5.

AS this thread pertains to the Red Wine design, I have not heard it and cannot comment on how our two designs compare. They are obviously very different designs. It sounds like Vinnie has done a great job and it has been well received.

Sincerely,

Dan W.

Thanks, Dan. Regarding item #3 above, why would it limit the bandwidth if you do attention at the input?

Dan,

Thanks for your explanation. I appreciate your response. I also want to share how your phono stage has increased my musical enjoyment. A masterful design.
Jim

Jax2, Thanx as I was reading along and was wondering how the Shill Plug for RWA came into play in this discussion? Good detective work on your part, and for out'g Jwarmbrand, as it was just Too Funny seeing all 14 of his tread posts being nothing but Shills! LOL