Output transformers a cheap alternative?

[euro21]

Don't forget that power transformers, especially toroids, has tremendous primary-secondary capacitance. You must select manufacturers and types to use its for OPT.

It's possible to use its as OPT, but I think as try-out.

 

[popilin]

In the hope that this type of comment get's paraphrased - something like:
SE OPTs get more benefit from an introduced gap, or
SE OPTs achieve a better core utilisation when purposefully gapped, or
SE OPTs in general have a gap, or ....

Just saying, its not a black and white situation for a diy audio forum.

A gap in a SE OPT is mandatory, DC magnetic field is given by

Bdc(max) = [4 π μ(eff) Np i(DC)] / (9 l)

μ(eff) depends on the gap, without it you will saturates the core, and on the long term you would have a nice magnet instead of an OPT.

Indeed it is a black and white situation.

 

[kodabmx]

Agreed, this is why reactor chokes have a big gap, but it's only mandatory for core efficiency. You'd need a BIG transformer, but even an air core transformer would function, just not well. This is the main reason I haven't build anything SE. I'd need to either use parafeed or a dummy load on one side of a PP OPT.

 

[trobbins]

Indeed it is a black and white situation.

i agree, all transformers have a 'gap', as ueff is never infinity. Every power transformer has a 'gap', ipso facto any transformer can be used as an se transformer - black and white.

 

[kodabmx]

A physical gap? Or the gap that means I'm 99% empty space?

 

[popilin]

i agree, all transformers have a 'gap', as ueff is never infinity.

No, I mean

μ(eff) ≈ l μ / [l + l(gap) μ]

For l(gap)=0 is μ(eff)≈μ

Every power transformer has a 'gap', ipso facto any transformer can be used as an se transformer - black and white.

Most toroidal power transformers have no gap.

E I power transformers have a natural gap, but by design it is very small in order to maximize μ(eff), it is also difficult to measure and you will end up with a mediocre OPT, or a nice magnet useful as a paperweight.

 

[kodabmx]

Is the gap just the varnish? Or is it designed into it? Sorry, but I have very limited knowledge of magnetics.

 

[popilin]

No, the gap is 2x distance between E and I around the magnetic circuit when used properly, i.e. all E stacked together and all I stacked together.

 

[kodabmx]

So the insulation provides that distance? Does this have a translation to a "tape measure" (IE wound in a coil, not a formed core) style toroidal? Does this mean instead of cutting a gap you could stack the E/I plates with a thin film between to achieve the gap?

 

[popilin]

Does this mean instead of cutting a gap you could stack the E/I plates with a thin film between to achieve the gap?

Exactly.

Edit: Just to clarify, in equation on post#26 l(gap) is evaluated twice, so you must do l(gap)/2 for the real gap, that's why I said 2x distance between E and I, just to be consistent with the equation.

 

[nigelwright7557]

Does this mean instead of cutting a gap you could stack the E/I plates with a thin film between to achieve the gap?

That's what I used to do with fly back mode SMPS transformers.

 

[trobbins]

E I power transformers have a natural gap

black and white - all E I transformers have a gap.

 

[PRR]

... ... the same if the transformer was substibuted for a 2k power resistor and 360V B+. And signal could be taken from a small 1:1 - 8 ohm in, 8 ohm out - transformer placed in series with the resistor. ...

That is a terrible mis-match. The 2K resistor will absorb 99.6% of the output power. And you still have DC current in your transformer, though because the low Z is few-turns *maybe* not enough to hurt.

 

[popilin]

No worse than most EI transformers for capacitance. You'll see in practice that the strays are a non issue at audio frequencies unless working at really high impedance, or when trying to use them as interstage coupling. Last time I ran frequency response measurements on several I found flat response out from 10Hz to 100khz, so I think that should be more than suitable

My calculations and 10KHz square wave test say the otherwise, distributed capacitance and leakage inductance do matter; but if you have measurements, please post them, you will save a lot of money to most of us.

 

[Lingwendil]

The trick is to run them with higher current tubes in most cases, although small toroid's will work well at low currents.

I no longer have access to advanced test equipment so I can't get any recent measurements for you. Previous testing was enough to prove the concept for me.

 

[popilin]

The trick is to run them with higher current tubes in most cases, although small toroid's will work well at low currents.

No idea about this, I must think a little.

I no longer have access to advanced test equipment so I can't get any recent measurements for you. Previous testing was enough to prove the concept for me.

You claim flat response from 10Hz to 100KHz, as the laws of physics cannot be broken, stray capacitance and leakage inductance must be very low, or your measurements are wrong.

I was stumped by low strays of a toroidal power transformer measured by Morgan Jones (see Valve Amplifiers Fourth Edition, page 411)

Cp-s=540pF, Ll=6mH, parallel capacitance of primary unknown however.

Very difficult to obtain, even by design...

 

[tikiroo]

Recent tests with a 115+115:15+15 120VA Multicomp toroid from Farnell showed open loop frequency response <10Hz to 65 kHz -3dB. Note this was driven from a hybrid triode/mosfet PP pair with very low equivalent Ra. Standing current was ~150mA per side.

 

[Shoog]

To answer the original question. I would be concerned at capacitive losses multiplying up. This would likely produce unpredicable ringing at high and low frequencies and also a response that barely made it out to 1K.

Shoog

 

[smoking-amp]

I am wondering how some of these toroid freq. response measurements are being made. If you drive the toroid with a 50 Ohm signal generator and no 8 Ohms load you -will- get stupendous results. But that is not how the measurement needs to be made. A series source resistance equivalent to the driving tubes is required (nominally similar to the primary Z rating), and an appropriate speaker like load R is required. And the low freq. end needs to be driven to full voltage (power) swing with some measurement of magnetizing current to insure saturation is not being entered. (current probe and scope)

There was a commercial vendor selling -purpose made- toroidal OTs, but apparently wound with the usual random wind technique for power xfmrs. With advertised specs that looked quite good. Until someone actually tried to use one for a tube amp, and it only worked up to 2 KHz!

These power toroids generally are lacking in turns quantity for the Hi Z, Hi V environment of tube amps. So using high current, LV tubes (aka TV Sweep tubes, paralleled) is a good idea. If you can verify that the toroid winder uses a progressive (single sweep chevron layer'd, $$$) winder machine, rather that a (typical) random wind machine, then you will be rewarded with much lower distributed capacitance. But brute force, high current, tubes can overcome the high dist. capacitance otherwise.

 

[kodabmx]

I've measured mine and full output at 50kHz is down 2-3db from 1kHz. Dummy loaded with a 6R resistor, fed from a DDS, measured with a scope. 1W output is within 1db from 6Hz to 50kHz. And with gNFB, 50kHz square is a lot more "square" than I expected... Without gNFB the same signal looks like a triangle/sawtooth hybrid.

I've used VPT series as OPTs with the following tubes: 6N13S PPP (VPT48--), 6P1P PPPP triode (VPT12||), KT120 triode (VPT18--), EL34 PP (VPT18--) regulated screens, 6P43P PP (VPT12||) triode, 6P3S PP (VPT12--) tetrode, and a 12AV5 PP triode, (VPT12||) all with excellent results. (-- = secondaries in series, || = parallel)
As far as commercial purpose built coils, Plitron from Toronto, Canada and Toroidy in Poland come to mind...