MagneQuest Forum

RE: Thanks.

151.197.37.182

This Post Has Been Edited by the Author

In Reply to: RE: Thanks. posted by gluca on February 11, 2008 at 05:09:52

Hi GL:
you asked;
:::Is there any reason on your choice for the splitter (which I like)? I mean, why one should not use the input tranny to split the signal?:::
As Kevin stated in his response--- the input side of the voltage amplifier (6DJ8) reflects my design choices arhitecturally. I.E., that we do not use a phase splitting input trans.
There are a whole host of reasons that I chose the architecture that is shown. I probably won't have time to address all of the issues--- so will just try to give some sense of what I was looking at and for architecturally.
If our first active stage is going to be an LTP--- then why not let it do the phase splitting? It's there already (the LTP)--- and even if we did use a phase splitting trans at the input and derived a differential signal off of the secondary of the input--- then the rest of our circuit is going to be differential--- and our first active stage would still be the LTP.
so--- since the LTP can split phase--- and do it very, very well--- it just seems to me moreso elegant to let the LTP do it's job and to then use the simpler and (in my opinion) higher performance 10K:10K single-ended to single-ended input trans--- which gives you a 1:1 turns ratio. The benefit of the 1:1 as we use it is itself severalfold.
As we use the input trans--- you can hang a conventional series fed pot (like a P&G, PEC, Alps, etc) or a discrete series or ladder type attenuator off of the secondary--- which works quite well. If we wanted volume discrimination (i.e., control) off of the secondary of a phase splitting input trans--- we could not use the widely available solutions out there like the P&G RF-15 as just one example.
Other consideration is that phase splitting transformers are not a walk in the park to do well. By this I mean--- ideally (in a platonic world) we would want the two halves of the secondary to be exactly equal in amplitude but opposite in phase--- so that we would want exactly equal (i.e., balanced) capacitances on each half to the center tap as well as exactly equal coupling from the single ended primary to the differential secondary winding also in terms of coupling (i.e., that the leakage reactances be exactly identical)--- and we would want it to have excellent bandwidth (pick your poison on what range this bandwidth should encompass) as well as low disortion and etc. But suffice it to say--- this is a challenge to do really, really well to the extent that I would like to aim at---
Life could be simpler (if only the gods had been kind to us)---if our cd player (or other source) had a differential output then I would be spec'ing in a PP to PP input trans--- which by virtue of both sides being PP is actually easier to get the qualities I mention in the preceding paragraph (i.e., the equalities of capacitance and leakages) on both sides of the transformer than it is when the primary is single ended and the secondary is PP.
Now... someone keen will observe that we do make some phase splitters and that I have previously posted a circuit here with the EXO-173 phase splitter--- and ask if this phase spitter is "good"----
allow us to look at the context--- where I have shown the use of the 173 phase splitter was in a circuit with a single gain stage as the front end--- a lot of folks have as a preference that a single ended gain stage stage sounds better at the front end than a differential gain stage--- and I don't want to argue which is "better sounding"--- but just simply observe and state that this is a preference that some folks either hold or express---
The main point being that if your going to use a single ended (i.e., not a differential) front end--- and your building a PP amp--- then your going to have to split phase somewhere--- and that is where we use the 173--- and note that the 173 is a single winding trans--- which has implications in both the use and performance of the design and it's resultant technical and sonic characteristics. In this position the 173 works pretty well--- it is not platonic perfect--- as no piece of iron in this world is--- but it does have very good capacitive balance btwn the two halves as well as tight coupling. In my view, it is the best way to split phase if your using a transformer to do this job.
And--- we could design this same type of autoformer phase splitter for use as an input phase splitting transformer---but--- again--- the first stage of amplification would then have to be differential--- so why duplicate this ability to split phase with an input
trans if we have an LTP on the job?
Further along this line of reasoning--- I'm led to wanting to explore (down the road somewhere) the COMPACT amp---- which essentially recognizes the inherent possibility of using the PP output stage itself as the phase splitting mechanism. Again, it (the PP output stage) is there--- why not allow it to also function similtaneously as the phase splitter?
These ramblings are just a few kernels of thought that I've been mulling over for a long time--- please understand that my comments, ruminations, and zany ideas are not proffered as the big TRUTH or as matters of fact or the only "right" way of doing things--- but instead I offer these thoughts (none of which have been completely flushed out in this post) as a way to give some insight into why I chose to use the architectures (circuit topologies) that have been shown both in this thread and an earlier thread (the PP design with the EXO-173 phase splitter).
shew--- I better get to work---
Mike

Edits: 02/12/08 02/12/08 02/12/08 02/12/08 02/12/08 02/12/08 02/12/08

Follow Ups: