Designing an Interstage Transformer

[popilin]

What!

No sliderules?

Ah, come on, not that hard !
Making the simplification

B = B(AC) = B(DC) = Bt/2

Bt = 12000 Gauss = 1.2 Tesla

We can write

N = L x 10^8 i(DC) / (S B)

[L] = Henry, = Ampere, = cm^2, = Gauss

The air gap is adjusted, measuring L, with the component i(DC).

 

[TonyTecson]

What!

No sliderules?

i still have mine from the 70's...

 

[TonyTecson]

Hi Tony
Nice to see you around here !

Look, I do what I can, I don't know what is a spreadsheet.

I always work with my old calculator.



What are you talking about ? Here valves are scarce and expensive, I have just a couple of projects for.

Once, I was tempted to get rid of my ECC82s, do you remember ?

After the overwhelming success of my last preamp and two class A monoblocks, I don't sell nor drunk.

BTW. All is here

http://www.diyaudio.com/forums/tubes-valves/181072-designing-transformers-j-c-maxwell.html

i must have mistook you for someone else...

 

[TheGimp]

Poplin, from your formula using L=50H , I=10mA, and core cross section is =0.88sq-in:

N = (50 * 10^8 * .01A) / (.88sq-in * 2.54 cm/in * 2.54cm/in *6000gauss)

N = 1467.8 turns

Then after winding the transformer coil, assembling the transformer and adjust the gap to achieve 50H with 10mA dc bias through the coil.

Is this correct?

 

[popilin]

i still have mine from the 70's...

I have two that belonged to my grandfather and then my father, but I have no idea how to use it, but keep it quiet...

i must have mistook you for someone else...

Well, it happens, the forum is big and there are many people.

I live in a small town where everybody knows everybody, and it happens to me every day, so if is for concern.

 

[TheGimp]

Dunno how you did your calcs there...

But I came up with 3800 turns. While that clearly is a high count ... it isn't outrageous. Just will take some time.

(I used u=2500, and H=432)

GoatGuy

GoatGuy,

Would you please post the formula(s) you used?

Thanks.

 

[TheGimp]

The comment on the slide rule was meant to be a joke. I still have two that work and one that does not (my fathers), although I've long forgotten how to use them. I vaguely remember using them in high school physics class in the late 60s.

 

[popilin]

Poplin, from your formula using L=50H , I=10mA, and core cross section is =0.88sq-in:

N = (50 * 10^8 * .01A) / (.88sq-in * 2.54 cm/in * 2.54cm/in *6000gauss)

N = 1467.8 turns

Then after winding the transformer coil, assembling the transformer and adjust the gap to achieve 50H with 10mA dc bias through the coil.

Is this correct?

Yes, this is correct.

As a rule of thumb, the wire (double or triple isolated) is selected for a current density of 3 A / mm ^ 2

You can play around with the numbers, maybe you can put more turns and approach the original target.

As insulation, you can use NOMEX, is very good insulator and for thicknesses of 0.08mm, has a dielectric constant lower than teflon !, very important to achieve low capacitance between windings.

 

[iko]

Calculating the number of turns this way does not take the gap into account. You can probably get away fumbling around with the gap after you wound your coil, for small currents like this example, but imagine you'll wind an output transformer for 100mA DC bias and as much as 260mA PP current. Then you might be sorry you didn't predict the number of turns, gap, and working flux density better. As they said, winding coils is labour intensive.

And even for this example, I'd do it right instead of playing with the core for the right gap. I'll be watching your progress with interest.

 

[TheGimp]

I was planning on using either Mylar or Kapton tape (.08mm) as I have both. However I'll start looking for Nomex.

I have double insulated magnet wire in sizes from 26 through 34 AWG. 34AWG is .02mm^2 so 60mA would be 3A/mm^2. Even 36AWG would be 38mA at 3A/mm^2.

EI-87 table shows a max of 5712 turns of 34AWG with 11 layers of .001" insulation. So I probably will get closer to 5000 turns total (2500 turns Pri, 2500Turns Sec) with .003" insulation.

 

[popilin]

I was planning on using either Mylar or Kapton tape (.08mm) as I have both. However I'll start looking for Nomex.

I have double insulated magnet wire in sizes from 26 through 34 AWG. 34AWG is .02mm^2 so 60mA would be 3A/mm^2. Even 36AWG would be 38mA at 3A/mm^2.

EI-87 table shows a max of 5712 turns of 34AWG with 11 layers of .001" insulation. So I probably will get closer to 5000 turns total (2500 turns Pri, 2500Turns Sec) with .003" insulation.

Go ahead with 0.02mm^2, is in the limit for manual winding.

Dielectric constant of Mylar is too high, I don't know for Kapton.

Try to use no more than 80% of the window, unless you want to go crazy.

Soon formulas to obtain capacitances...

I forgot to mention interleaving, to reduce L disperse...

 

[popilin]

Calculating the number of turns this way does not take the gap into account. You can probably get away fumbling around with the gap after you wound your coil, for small currents like this example, but imagine you'll wind an output transformer for 100mA DC bias and as much as 260mA PP current. Then you might be sorry you didn't predict the number of turns, gap, and working flux density better. As they said, winding coils is labour intensive.

And even for this example, I'd do it right instead of playing with the core for the right gap. I'll be watching your progress with interest.

Wrong !
μ is implicit in L, and air gap is implicit in μ

L = 4 π μ S N^2 / (9 l x 10^8)

All equations are consistent.
From what I see, you have not seen the attachment, there is all explained in detail.

 

[popilin]

Before I forget, between layers, I use 0.05mm polyester, cheap as it is used to insulate roofs (with paint), also is flexible and allows more wire in windings.

It is also important to interleave the windings, to increase magnetic coupling between primary and secondary, and reduce L disperse, eg 3 primaries and 3 secondaries is a reasonable start point, be in mind that if you increase this numbers, Ld will decrease, but Cd will increase...

With a 1:1 turns ratio, the capacitance between windings is not too critical, but just in case, a way to calculate.

Neglecting edge effects, we can obtain a reasonable approximation to the value of capacitance C.
For two parallel plane plates of area A separated by a distance d, filled with a dielectric medium of dielectric constant ε, from Maxwell's equations

∇.D = 4πρ

D = ε E

Integrating in a volume V

∫∇.D dV = ∮D.ds = D A = ε E A = ∫4πρ dV = 4πQ

Then

E = (4πQ) / (ε A)

Also

E = - ∇φ = dφ/dx

Integrating

Δφ = ∫dφ = E ∫dx = E d

By definition

C = Q/Δφ

Then

C = (ε A) / (4πd)

A is the average winding area, and d is the insulation thickness.

Note that [C] = cm, but no problem, I love cgs units.

1 cm = 1,113 pF

Try first do all dry, unvarnished, because of the dielectric constant.
Do not expect to get it right on the first try, patience and perseverance.

Good luck !

 

[TheGimp]

I was considering bi-filar winding.

This brings up a question about winding. Is capacitance between layers less if I wind all from left to right?

This opposed to winding one layer left to right and the subsequent layer right to left.

That is should all layers be wound in the same direction across the bobbin, or is it acceptable to wind alternate layers in opposite directions?

 

[popilin]

I was considering bi-filar winding.

This brings up a question about winding. Is capacitance between layers less if I wind all from left to right?

This opposed to winding one layer left to right and the subsequent layer right to left.

That is should all layers be wound in the same direction across the bobbin, or is it acceptable to wind alternate layers in opposite directions?

1) Bifilar winding, significantly increase the capacitance, but greatly reduces Ld.

2) Vertical sectioning, also reduces capacitance, requires winding in opposite directions in each vertical section. Must be accompanied by interleaving.

As a humble advice, try to make a standard winding, with say, 3 P and 3 S, wound in the same direction, eg one layer from left to right, the second layer from right to left, an so on, always maintaining the same direction of rotation.

Place a good insulator between each P and S (NOMEX or Teflon) with sufficient thickness to prevent arcing and reduce capacitance between windings.

If this is your first transformer, not complicate, make bifilar windings with a wire so thin is a nightmare.

In order to keep the magnetic field homogeneous and prevent short circuits, wires should not cross each other, and that is difficult, even in a normal winding.

 

[GoatGuy]

1) Bifilar winding, significantly increase the capacitance, but greatly reduces Ld.

2) Vertical sectioning, also reduces capacitance, requires winding in opposite directions in each vertical section. Must be accompanied by interleaving.

As a humble advice, try to make a standard winding, with say, 3 P and 3 S, wound in the same direction, eg one layer from left to right, the second layer from right to left, an so on, always maintaining the same direction of rotation.

Place a good insulator between each P and S (NOMEX or Teflon) with sufficient thickness to prevent arcing and reduce capacitance between windings.

If this is your first transformer, not complicate, make bifilar windings with a wire so thin is a nightmare.

In order to keep the magnetic field homogeneous and prevent short circuits, wires should not cross each other, and that is difficult, even in a normal winding.

Good heavens! What on earth do they teach the physicists of Cordoba Agt'a, these days anyway? My hat is doffed to you, Sr. Physicist ... for your invoking not only Maxwell's Laws into the capacitance estimate, but also inline special characters for both integration and circular integration!. Remarkable. And to think, I was under the misaprehension that Argentina was full of ... gauchos and ill-tempered beef cattle.

L O L

GoatGuy

 

[popilin]

Good heavens! What on earth do they teach the physicists of Cordoba Agt'a, these days anyway? My hat is doffed to you, Sr. Physicist ... for your invoking not only Maxwell's Laws into the capacitance estimate, but also inline special characters for both integration and circular integration!. Remarkable. And to think, I was under the misaprehension that Argentina was full of ... gauchos and ill-tempered beef cattle.

L O L

GoatGuy

Wow, Thank you !

I think do not deserve it, but I'm flattered.

Not only on the Internet, there are falsehoods, also in some books, and it hurts a lot of people who want to learn.

I'm just a TV repairman, but it's nice that my humble contribution is valued.

 

[TheGimp]

What to use to wind the coils?

I've looked on ebay and an occasional old winder shows up.

There are new Chinese units but I'm not thrilled with them.

It looks like a reasonable unit goes for around $200, which I prefer not to spend.

That leaves me with making one. Sewing Machine motors are cheap, and speed can be controlled with a foot pedel.

I've also got a 240rpm motor (geared) that I am inclined to use. I've tested it with a variac and I can control the speed over a limited range.

If I use pulleys and a belt I can configure it for 120r/m which seems reasonable for a hand feed unit.

I've got 1/2" stock and a set of four 1/2" id bronze bushings I could use for bearing.

Drill and tap the end of a shaft for 1/4" All-thread for the spindle end.

Drill the opposite end radially and use a photo detector through a flip-flop to drive a counter.

The wire tensioner would have to be worked out.

It looks like a full weekend project.

What do y'all use?

 

[popilin]

I had the same idea, I took the old sewing machine from my mom, I put a screw in the wheel, a magnet, a reed switch connected to a cheap calculator, et voilá, a winding machine.

When the little motor begins to pick up speed, the wire comes to life and does what it wants.

Solution: Move the steering wheel with one hand and hold the wire with the other hand, one turn every four seconds !

One layer of epoxy, one layer of wire, one layer of epoxy, one layer of polyester, and so on, finished with five layers of NOMEX.
As a newbie, I calculated the transformers for a 95% window, bad idea, after finishing each winding, put presses until the extent necessary, allow the epoxy to cure for three days, and then proceed to the next winding.
More than eighty windings for five transformers, and about three months of hard work.

Today I can proudly say that the result was sublime.
A preamp and two class A monoblocks, that even now I can not explain why they sound so good.

 

[bear]

Not sure you need the higher temp Nomex for an IT transformer. Ought not be much heat at that power level? Kapton would seem to be sufficient?

_-_-


PS. I'd suggest to expect the first couple of winds might not work out, so expect to maybe do a few bobbins, wind, test, wind, test, etc...