Spot on. Take note of the typical arrangements for paralled secondaries as many other manufacturers have similar multi sectioned balanced windings of 8+8 or 16+16 as with my custom with other similar vintage Partridge /Williamson configurations. I found the expected secondary performances not quite so perfect when the MF/HF end shows up minor phase differences. It is impossible to get a large transformer which is so nearly perfect in other aspects to be correct for the parasitics, when windings are at calculated positions of both pri and sec sections. These sections are brought out to be part of another series or parallel arrangement and from there may be very slight phase errors.
There is nothing easy, and also the high influence of vacuum impregnation i.e increased dielectric capacitance can mess with many leakage issues. Hence the best quality windings on the market are those with only Kellogs paper used as layer insulation materials and not that dreaded yellow insulation tape. However, caveats here; those output transformers sporting low primary impedances and lower prim to sec ratio's are better off with parasitics, as much less copper is exposed in area when compared to a 10K primary Z. so fewer layers. So it is reasonable to expect a lower primary Z transformer performing better than it's 5-10K Z counterpart. This is correct. Manufacturers will get away with what they can for ease.
My bench approach to find out the best secondary configuration with a multi sectionalized specimen is to use a known solid state amp, if possible with a bridge configuration that can sport a good quality 1-5Khz square wave at low level connected to the whole primary, with secondaries not terminated with a load. The bridge replicates to a near push pull. This may be somewhat tedious, but after a while becomes a picture of which one individual secondary produces the best squarewave.
Don't be fooled in every secondary is identical, certainly not and that winding labelled No 1 is closest to the core; (could be) but resistances of others will be low all round or it could be simply imply a manufacturers start point. Color code all secondaries, and tabulate results accordingly.. Parasitics love being damped and masked, so an open circuited section will display all. Examine the quality of each secondary winding, in every possible connection combination. with a 4 sectioned transformer as an example with 1-2 Ohm load all are paralled and the 15/16 ohm all in series, the troubles start appearing with the awkward 8 ohm load, this usually requiring two sections in series and two paralleled, the result all in series. Playing with the combinations can create interesting square wave results. In one case with another transformer manufacturer, the results were so varied, that when the time came to connect the real tube circuit, it was found the GNFB small capacitor step values were way out to the claim of "so-called drop in replacement". Murphies Law !
This is tough A-B going, but we are are dealing with a component that is ultimate in determining amplifier quality. An enormous number of manufacturers during the 1950's to present day have struggled to claim theirs is the best. Not always true.
A sectionalised 8+8 can sound just as good as a 16+16 sectioned job, but size for size the HF bandwidth response is a give away to the quality.
Keep at it !
rJ
actually it does ...
search - advanced - find posts from - any date - newer than / older than
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.