Andre Jute wrote:
Patrick Turner wrote:
I use about 65 consecutive steps when designing a PP OPT.
Is that all? I think if you tried to write them out you would find that
it is more complicated than that.
Each of the 65 steps may have several considerations, and no I won't list
them on this group, its several pages, and a few diagrams.
I am working with a computer IT student who is trying to
create a useful working program to
create decent transformers.
I don't know anything about
transformer design (thanks, Menno and others but I just don't have a
lifetime to give to something so complicated) but I once tried to write
out all the steps involved in designing even a simple SET amp and after
two days gave up when the list was more than a hundred pages long.
But you are a writer, you do have a tendency to use a pile of words
where the more technical and literary challenged among us will
simply say do this, do that, divide z by m squared, and so on until we are
done.
The guy working on the program is also into tubes, and has built
some gear, but he found it almost impossible to get beyond the basic
turn count and core size. Anyone can do that, but its
woking out how to fill the window efficiently, keep the leakage and shunt C
low,
and losses low and get a range of load matches without wasting sections or
having
unequal wire current densities that take a lot of hard work.
The wire sizes have to be factored in to suit the wanted turns per layer on
standard available bobbins and the tendency that the theoretical
turns per layer won't fit easily in practice must be allowed for, as well
as winding
bulge during wind up; ie, the avoid all the swearing and cursing by first
timers,
the design must finish up absolutely true and bullet proof for anyone to
wind.
It must be extremely easy for anyone to understand the final bobbin cross
section
diagram that tells folks what to do, so that the design can be exactly
described in one simple paragraph, just like the description of the
Williamson
OPT in RDH4.
If only other makers were able to describe so simply and with such easy
precision
and economy of words.
( But they'd rather keep ppl guessing about exactly what is inside the
bobbins. )
I am not aware of any program for winding trannies that is any good.
Hence I am attempting to do this myself.
The IT guy is very slow because he gets distracted with other projects.
But I am not going to hurry him, i don't do that.
It is indicative that Gordon Rankin, not notably a man to waste time
unless you paid him by the minute, once admitted that his *post-design*
checks on an amplifier circuit amounted to a hundred steps. In other
words, after he committed all the necessary work to create the design,
he took another hundred steps to be certain it would work right.
Well, most of what you need to build a decent OPT is at my website.
Anyone with 3/4 of a reasonable brain could just read, understand, and
they'd get 65 steps like me, and the checks to make sure it is right are
included.
A few ppl over the years have emailed me to help them with their attempts.
They all seemed happy.
A colleage in Sydney where far more customers for tube gear lurk
is quite hopeless designing OPTs, so he always comes to me
when there is something fancy, such as an SET OPT for GM70.
I diagrammed the whole amp and OPT, he built it, and the audio club
liked it, and that's a nice result.
If you consider each turn to be something to worry over and write a page
about,
you get thousands of pages.
OPT building isn't rocket science, its just a part of electronics that is a
sum of a whole brew
of simple concepts bundled together like the rest of electronics.
Once youv'e wound a few trannies, you wonder what all the fuss is about
and wonder why ppl think its so hard and complex and not easy to do.
The practical aspect does take natural skill with tools, something
not all the population has, some guys I know are quite dangerous
with any sort of tool in their hands. They will never
be able to do anything, and will have to rely on
the skilled to get them through life.
One in particular comes second in the local speed chess competions.
He comes about 3rd in the major tournements, and I will never beat
him at chess.
I don't understand his encyclopedic knowledge of chess games,
I don't know whay I lose or win sometimes, since analysis
6 moves deep is beyond me, so horses for courses.
He
shared some of the steps and it was all meaty stuff for which the
reasons for any check would take several pages to write out
comprehensibly. It isn't rocket science to conclude that, if the
post-design check is 100 steps, the actual work is much, much more.
I doubt its as difficult as you say.
I have a few checks for F response, primary L, leakage L,
shunt C, and testing, and that's all that's needed.
Its all in RDH4 on just a few pages which most ppl just cannot understand.
The working out of total lumped shunt C of an interlaved OPT
is very difficult though if one tries to use the table for it in RDH4, so I
don't try,
and there is no need, because empirical ideas for insulation thicknesses
will always give low enough shunt C, and what's important is the
C looking into each end primary end with the secondary grounded.
The reactive nature looking into the primary ends with an interleaved
tranny
is never just pure C or L, but a complex arrangement of a low pass multi
order filter which is usually far too difficult for anyone to accurately
measure, calculate, or model
with a program and nobody has ever done it correctly.
Engineers hate such things that cannot be designed with predictable
precision.
So its suck it and see, creating much nervousness in anyone worried by the
actions of a maker who does not appear to know much.
But there are some things that cannot be known.
But after following a few known golden rules, its dead easy to get
a huge E&I OPT to go 15Hz to 270kHz at 250 watts for tubes or any other
devices,
and with no horrid series or parallel resonances withing the pass band.
Patrick Turner.
Andre Jute
Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review