output transformers

[Yvesm]

Hi Larry

You are right stating that it is impossible to made two sections equal each other, but.... williamsons OPTs were made that way, and how could be significant the difference between the two sections compared to the tubes tollerance? Splitting the windings in two section there could be a lesser stray capacitance at least 1/4 times the same windings on one section, as is stated on chapter 5 of the Radiotron. This is why I'm experimenting it. Till now I winded all my transformers with 5 to 9 interwinding layers but in one section only and they are working very fine. Now, I have to find somenting else to make things worse

Sounds right to my ears !
This technic is comonly used in "modern switch mode supplies" wich run at more that 100Khz square.
Bobins with 4 or more sections are commonly used, as long as Litz wire is.

So you are the expert.

Bleahhh ! I recently made a "****" for a 300B.
It was on a EI96 core and showed very poor hi freq behaviour (less than 25Khz -3dB point)
Too low interleaving I bet !! (4 pri, 5 secs)

Resident is right here

Yves.

 

[resident]

Yvesm,
Nice machine!What kind of meter is this?

Notice that turns must be the one exactly next to the other.
And you must wind like this:layer-insulation paper-next layer-insulation-next layer-insulation-etc
I can't see from the pic if you do it like this but seems that turns are not exactly side by side.

Larry,

Splitting the windings in two section there could be a lesser stray capacitance at least 1/4 times the same windings on one section, as is stated on chapter 5 of the Radiotron. This is why I'm experimenting it. Till now I winded all my transformers with 5 to 9 interwinding layers but in one section only and they are working very fine. Now, I have to find somenting else to make things worse

So...just inform us with the results.It'll be interesting!

Thanks for your help about the bobbin construction.
Not an easy job but I think it's worth it.
Have you done any tests btw your bobbins and a common one (like the bobbin that Yvesm uses)?

Nevertheless I think, sometime in the future, I'll try a double C core from Metglass, but is still not very easy to find that material in Saudi Italy .

Why don't you buy double C-cores from Italy?

So I find the use of extra permeability cores expensive and not so convenient.

I agree with this. But some fellows think that strange exotic cores are better for OPTs.Or silver wire...
Are the same guys that use gold plated sockets mounted on PCBs or matched pair tubes on amps without bias.

 

[Yvesm]

Hi Resident,

Notice that turns must be the one exactly next to the other.
And you must wind like this:layer-insulation paper-next layer-insulation-next layer-insulation-etc
I can't see from the pic if you do it like this but seems that turns are not exactly side by side.

True !
I try hardly !
I don't remember what I was winding when I took the shot !
From the original photo (uncompressed) turns look neat, just some light reflections.

The counter is the "turn counter" actuated by a door magnet (the white thing barely visible on the pulley) passing near of a Hall effect sensor.
It also activates the small stepping motor that drives the wire "positionner" by mean of a thread.

It is an evolutive tool

Yves.

 

[resident]

Nice job!
I think I'll start to wind my OPTs,too.Needs too much time,though.And a winding machine.

One more hint:leave some space left and right of the bobbin.Say 3-5mm.It's a hint that my local winder do.

Larry,
I just asked my local winder about double C-cores and he told me that there is a company in Italy called Femag.
By the way he also liked your bobbin!
Maybe I'll build one one day and give it to him to wind me an OPT.

 

[Larry Lomax]

Hi Yvesm, hello, Resident,

The success of my bobbin is outstanding.... no, just found somebody in the world who like it a bit Thank you!

Well, I think there is not a sensible difference in using a such bobbin compared to commercial ones. I made it just becouse:

The material is very hard to break.
It has a high voltage dielectric rigidity.
It is resistant to high temperature for soldering.
It is easy to work with a milling machine to good precision.
And over all it is of the correct size I need for the core.
And.... ush... it looks vintage!
Nevertheless, there are other products even more expensive than that I used, and even with better electric and mechanic characteristics. Same tufnol same RS-Components.com

I use to wind myself my transformers becouse I'm a fanatic of making things well, so, when I wind somenthing I take maxymum care on laying turns side by side and taking care not to overlap them. After winding a layer a bit of insulating spray (for the final ones but not for the prototypes) and over a layer of insulating sheet, like mylar, over and over.
When the transformer is performing well, I take it out of the amp and I submerge it on a soucepan filled with melted beeswax, I allow it to seep the wax for a time, and finally I take it out letting it to cool up. This prevents the "thing" to sound. I do not let space aside the winding becouse of the use of the insulating bobbin and the mylar insulation that fits greatly to it, so no chances are left for the turns of the adiacent layers to come in contact.

Off course I will post the results and all the data collected after finished to make you know.

Just don't talk about Italian companies.

Exothic cores are really better of ordinary ones, if the use is correct.
A trash winded, exothic transformer doesn't sound any better than a good winded ordinary core transformer.

Gold is also a very good plating material for connections but it doesn't sound any different from silver or chromium. It only lasts for a lot more time before showing oxydation.

Materials by themselves do not make a good amplifier, but they help smart people in making them.

Best regards and nice weekend to all of you!