what are the advantages of double core output transformers vs EI ?

[bembel]

Hi all,
my electronic "corner shop" has, at last, found a supplier for tube output transformers. they are offering Ei cores as well as double C.
What are the advantages ? any datas ?
here is the list (for PP OTs), are the prices fair (for real price with taxes, multiply this prices by 1.2 )

[cerrem]

The double C-core would be more efficient and all grain in the same direction means max permeabilty for same given flux density...
The main difference between C-core and EI lams is the incremental inductance curve....
The EI lamination stack has 2 gaps.....the small gap is from E to E and is roughly twice the thickness of the oxide coating for one lam... When this gets saturated, then it jumps over to the second gap which is from E to I spacing which is larger gap.... WIth the standard EI scaling dimensions...the peak inductance occurs at approx .53 of the peak permeability of the steel which is roughly 13,000 Gauss for M6 core material... So if you designed at 100W OPT to be at 13,000 gauss at 100W output ..you would have peaked your inductance at roughly 53 Watts....and the inductance would be going back down the curve and at lower inductance at fullpower output at 100W...
Now if you designed a double C-core for 100W power at 13,000 gauss... You only have one effective gap, therefore your incremental inductance woul PEAK at fullpower output at 100W...
What this all means is that the EI lam will give you more inductance for the same given conditions when the amp is at half power range...which to me makes more sense, since the peak inductance is in a more usuable range... How often do you play yur Hi-Fi amps at full power output...

But a smart OPT designer will ensure that the lowest inductance on the curve still meets the minimum amount needed for the desired low frequency pole...

Chris

[refference]

CHRIS , I agree with you , at least for me
( and I can see that for you ,too ) , makes more
sense to use a E I lamination , if you want the
max. bass response at a half of total power .

BEMBEL , in my humble opinion , the prices are high
only a bit or two, for example , I can get a VERY GOOD output
transformer ( similar to the model EPP 3550 , with UL
taps ) with a local manufacturer , that always
have done jobs for me , by aprox. 110 euros ( 305 real , the
local currency , obviously , and all taxes included ) .

Regards for all ,

Carlos

[sparky813]

Not sure how much difference it makes, but I have cut cores with steel that is .002 thick compared to .006 to .014 for most laminations. These cores were designed for high frequency, 5kw inverters and to have low loss at 6khz or so. Each set of two matched cut cores weighs about 20 lbs.

The cut cores are also easy to gap for SE applications using a glastic spacer and a stainless steel strap. I havn't wound an audio transformer with them yet, but I plan to , and I will compare to others I have built.

Any thoughts on the advantages of the thinner lams?

Sparky

[bembel]

Thank you very much for your answers I can see more clearly now !

In addition I've got two unknown salvaged "double C" transformers, one for power(unfortunately low voltage, see attached picture at the left) they is an ECC83 in the middle for size comparison, and one for output, that are not yet tested (the right side is OT PPUL).
Is there something special relative to doubleC to keep in mind during measurements estimating them ?

[sparky813]

One other minor item: you can peel away some of the inside layers of a c-core if you need more bobbin space or build up the bobbin and find that it has become too large. Just start at the flat surface with an exaco or other sharp knife, and peel away a few layers at a time. I also use needle nose pliers to sort of wind up the unwanted layers, after I get them started.

Sparky

[cerrem]

As the other poster mentioned... The C-cores can come with much thinner strip....and you get more efficiency and higher effective permeabilty, which has it's benefits...
Sometimes the double C-core is more appropriate for a design...
For example, when I make MacIntosh unity coupled outputs...like for a MC30, MC60..ect.. i wouldn't use anything but the double C-core like the original ones did...because it's part of the sound... and you don't want to change the winding geometry to fit it for EI lamination..then the leakges and winding capacitance will all change due to the different winding width...
Even though for SIngle Ended outputs, the C-cores are mucho easier for gapping.... I prefer EI lams in a SIngle Ended output, since inductance is at a premium with SE ouputs...So I want my inductance peaked by mid power level...Thats why I believe the C-core Single Ended outputs sound thinner..since they don't have the full inductance until full power output..and since music is variable amplitude, hence variable output power....at very low levels the C-core amp sounds thinner...

Chris

[Dave Cigna]

Originally posted by cerrem
I prefer EI lams in a SIngle Ended output, since inductance is at a premium with SE ouputs...So I want my inductance peaked by mid power level...Thats why I believe the C-core Single Ended outputs sound thinner..since they don't have the full inductance until full power output..and since music is variable amplitude, hence variable output power....at very low levels the C-core amp sounds thinner...

I understood your explanation of inductance peaking to apply to EI cores with interleaved laminations - the two different gaps and all. Maybe I misunderstood, but if I didn't, then it seems that it wouldn't apply to SE cores because they are gapped; the lams are not interleaved.

-- Dave

[Tweeker]

Thats why I believe the C-core Single Ended outputs sound thinner..since they don't have the full inductance until full power output..and since music is variable amplitude, hence variable output power....at very low levels the C-core amp sounds thinner...

An SE amp is typically operated at the knee of the B-H curve, it has as much inductance is its going to have a low signal, with a large signal it may drop as the core saturates.

[cerrem]

Dave...sorry... I meant PP not SE.....my bad...

Tweeker... Your point is valid..but in all the years I designed SE OPT's , I NEVER go near the knee of the BH loop...not even close..not even with power transformers...
The tip of the BH loop is significantly way past the peak permeability point.... That would be dis-respectfull to the poor little output transformer.....
The core distortion at those high flux densities renders the design not suitable for Hi-Fi in my opinion....
I usually set SE flux density at 12,000 Gauss @full power@ 20Hz...The inductance is another story..I will use enough L to get my -3dB point to 2Hz in PP designs and for SE designs way below 20Hz, like 8 to 12Hz worst case....
Not trying to be arogant...just don't see the point in making OPT's unless it's done right...

Chris