I'm Baaack! Sorry for being absent lateley... lots of projects need to be placed under the Festivus Pole and I had a bout of the 'day job'. In any event I'm getting caught up and here is the first of a few posts that are in the queue.
One of the things i keep getting asked for is transformers with multiple taps and my stance has always been that multiple taps or the universal transformer us universally compromised. Sadly there is very little documentation out there on transformer behavior and it really took me a long time to come to grips with it.
Every transformer I measured with multiple taps had different frequency response for each tap so I simply chose to wind for the best bandwidth I could for a specific situation and people have been happy with the results. Then I get the email from Bill and Joe at Purity Audio Design for an output for their new flagship parallel 300B amp. They already use the Chapman autoformer modules in their top tier products. When they gild the lilly with silver (given a pair of silver chapman modules costs more than most equipment out there) I took their request seriously.
The deal breaker was two fold... since it was a commercial offering, they needed taps and the switching between taps needed to be user friendly. Traditionally the units with the worst variation from "tap to tap" are the ones that simply have a ground 4, 8, and 16 ohm tap. It is clean and easy but from a winding perspective it is far from ideal. The other traditional solution is to use various series parallel arrangements to allow for different speaker impedances but this is not user friendly. Since I bring out the winding wire for direct attachment to the binding posts, I do not consider any form of switch in the high current speaker loop to be acceptable.
This takes me back to the picture above and the broken eggs. That box holds about 45 pounds of scrap copper that this project created through failed experiments. Luckily the work paid off with a transformer that can deliver 20W (@ 50hz) into 4, 8 or 16 ohms that can be end user adjustable with minimal compromise. The first thing to look at is the high frequency performance where I get -1dB at better than 40Khz in all three positions and only slight variations in performance from tap to tap. I also show the results of the actual two transformers sent and am happy with how close they match each other.
The low end is much simpler and at 160ma + the 20W@50hz signal the inductance is still linear and will give a -1dB point of 10hz or slightly below when driven by parallel 300b's which will put any claims of amorphous cored transformers having weak bass to rest once and for all.
I'll leave you with some pics of the things that live under the "pretty cans" and then get back into character with a give away.
Wow... that was long and boring and totally uncharacteristic for me so I'll get back into character for the giveaway. Long time readers will remember the post about "ghost bunny" having respect for morgans book. Alas the bunny has been gone for nearly two years now and that very book he protected has been soiled by a cat. So if anyone wants a free copy (possibly with a cat) I'll gladly post it anywhere in the world for free.