Variable Impedance schematic?

[clintrubber]

[quote author="rodabod"][quote author="clintrubber"]
As I hope and am reasonably sure the situation is like this:

Connecting them in series gives you a centre tap to which you can or won't be connecting phantom by what was it, a 3k4 ? Ratio will then be 1:N.
By connecting them in parallel the ratio will be 1/2:N, so 1:2N (and the impedance-ratio will scale fourfold).

As I understand bifilar, it's just that two windings are wound at once, so they're kind of twins i.s.o. one after the other (with 9 months delay :wink: ). For the principal functioning of the TX it won't matter I figure.
[/quote]

Hi Peter,

Is having a bifilar twin primary the same as having a split primary?

[/quote]I think the former is always the latter, but the latter not necessarily always the former I think. The difference is in how they're wound.

As you stated it, from 'twin' & 'split' both follows that the primaries are not already connected. If they were the word 'centertapped' would have been used somewhere.

You likely will have had a look already at various possible permutations in the Beyer catalog (the 4th digit I thought).

I asked what would happen when paralleling a split primary (because I thought it would halve the ratio) but this is what CJ said:

[quote author="CJ The Transformer Murderer"]Imagine if you used 100 strand litz wire to wind a primary. How many turns do you have if you wrap the litz wire around the core one time?

1 turn. Even though that turn consists of 100 individual strands of wire.

So if you parallel a primary, you are just doubling the strands in your wire, not doubling the turns.

So the ratio stays the same.

So, will the ratio not just be the same if you parallel these bifilar primaries? The turns number on the primary would just be the same, no ?

[/quote]Nice explanation, understanding this. Yes, the ratio will stay the same, as CJ said.

But how about impedance ? May get confusing since some 'fixed' relationship between ratio (N) & impedance-scaling (N*N) seems to get broken:
For each doubling of the amount of parallel windings the impedance when looking into the whole thing halves, correct ?
Say you got a 1+1:N and a sec. termination of 100kOhm. Just using 1:N gives an impedance when looking into that one primary of 100k/(N*N).
Add the other primary in parallel to the first primary, we maintain the ratio, but now get 0.5*(100k/(N*N)), correct ?

Of course, it is easy to double the ratio with the twin primary...

P.S, Peter, those puns are getting worse and worse....

[quote author="Peter The Comedian"]they're kind of twins i.s.o. one after the other (with 9 months delay :wink: )...

Please let us know how they turn(...) out !

[/quote]

Yes that was too funny, and that even in one posting. Sigh...
You're right, should be terminated.

 

[rodabod]

[quote author="Rod"]Is having a bifilar twin primary the same as having a split primary?[/quote]

[quote author="Peter"]I think the former is always the latter, but the latter not necessarily always the former I think. The difference is in how they're wound.

As you stated it, from 'twin' & 'split' both follows that the primaries are not already connected. If they were the word 'centertapped' would have been used somewhere.[/quote]

Yes, that makes sense now.

But how about impedance ? May get confusing since some 'fixed' relationship between ratio (N) & impedance-scaling (N*N) seems to get broken:

For each doubling of the amount of parallel windings the impedance when looking into the whole thing halves, correct ?

That sounds correct, each primary coil shows the same impedance, in parallel this impedance is halved.

Say you got a 1+1:N and a sec. termination of 100kOhm. Just using 1:N gives an impedance when looking into that one primary of 100k/(N*N).

Yep.

Add the other primary in parallel to the first primary, we maintain the ratio, but now get 0.5*(100k/(N*N)), correct ?

That sounds right to me.

I think my primary question is, is a (for example) 1:10 bifilar transformer actually a 1+1:20?

My secondary question is, why bother paralleling the two coils? If you put them both in series then the step-up transformer will be half the ratio of using a single coil. (1+1:N = 2:N in series).

I think I'm confused now.... What is going on here?:

http://www.lundahl.se/pdfs/datash/1578_8xl.pdf

It says the primaries for this 1+1:10 transformer gives 1:5 in series, and 1:10 in parallel..... Have I gone mad?

I'm asking too many questions - sorry.

All of these current transformer questions are just winding me up.

 

[clintrubber]

[quote author="rodabod"]I'm asking too many questions - sorry.[/quote]
No, please continue. We learn from this - I do.
I'm a newbie at all this TX-stuff, but I trust the big boys here will correct us if we're really going wrong.

I think my primary question is, is a (for example) 1:10 bifilar transformer actually a 1+1:20?

I'm not sure how to imagine a 1:10 bifilar transformer. The A*P*I-output-TX is a bifilar-wound TX since it has four equal windings, of which a few happen to be considered Prim & others Sec. 'All are 75'.
But for a TX with a turns-ratio unequal to 1 one of the 'twin-wires' will run out before the other is finished. So it can't be a bifilar as I understand it.

I do ask myself though whether a TX which is said to be 1:10 and has split primaries (winding methode not relevant here now) is a 1+1:20 or a 1+1:10. I figure its a 1+1:20.
Maybe we're wondering the same but phrased it differently ?

My secondary question is, why bother paralleling the two coils? If you put them both in series then the step-up transformer will be half the ratio of using a single coil. (1+1:N = 2:N in series).

Might be a Dutch habit - since we've paid for that other primary winding we want to use it. But serious, good question. Just using one primary and leaving the other primary unused could still give you 'any' impedance you want by adjusting the secondary termination. Using both looks more elegant though, but I suspect there's a good reason for it we both don't yet see.
(If it's unused it should be taken care of of course that it doesn't pick up nasty signals)

I think I'm confused now.... What is going on here?:

http://www.lundahl.se/pdfs/datash/1578_8xl.pdf

It says the primaries for this 1+1:10 transformer gives 1:5 in series, and 1:10 in parallel..... Have I gone mad?

It looks good to me. The ratio of each primary to the secondary seems 1:10. Two primaries series in fact gives you 2:10, so let's call that 1:5.

I'm asking too many questions - sorry.

All of these current transformer questions are just winding me up.

You don't mind to sneak a few in as well, are you ? :grin:
(Let's hope that phrase doesn't have another meaning as well... )

Bye,

Peter

 

[rodabod]

I do ask myself though whether a TX which is said to be 1:10 and has split primaries (winding methode not relevant here now) is a 1+1:20 or a 1+1:10. I figure its a 1+1:20.
Maybe we're wondering the same but phrased it differently ?

Yes. That's what I was thinking.

I think I'm confused now.... What is going on here?:

http://www.lundahl.se/pdfs/datash/1578_8xl.pdf

It looks good to me. The ratio of each primary to the secondary seems 1:10. Two primaries series in fact gives you 2:10, so let's call that 1:5.

Ok, so each primary coil gives a ratio of 1:10 to the secondary.

Therefore, in series we have 2:10 which as you say is 1:5.

But, in parallel, we still get 1:10 as it says in the datasheet - so, is the winding ratio still considered the same?

I thought what we were saying earlier was that the ratio would halve since the impedance of each coil would be in parallel?

 

[clintrubber]

[quote author="rodabod"]

It looks good to me. The ratio of each primary to the secondary seems 1:10. Two primaries series in fact gives you 2:10, so let's call that 1:5.

Ok, so each primary coil gives a ratio of 1:10 to the secondary.

Therefore, in series we have 2:10 which as you say is 1:5.

But, in parallel, we still get 1:10 as it says in the datasheet - so, is the winding ratio still considered the same?

I thought what we were saying earlier was that the ratio would halve since the impedance of each coil would be in parallel?[/quote]
The winding-ratio must still stay the same, still 1:10, despite more primaries in parallel - and like CJ's example indicated.
But when determining the impedance-'scaling' we need to take into account that there are now more primaries joining.

Bye,

Peter