I`m on the way to finish my self wound 300B SE output transformers.
In contrast to the most, this design uses one big c-core with two sectioned coils.
My question now is, how to interconnect the sections, resp. which kind of connection would yield to the best frequency response?
Sections per coil are:
coil a) and coil b): p - s - 2p - s - 2p - s - p each
The secondaries are connected in parallel for all versions, so they aren`t shown any more.
For primary series interconnection, with respect to capacitance distribution, I think (also from comparison with existing designs), the best would be to connect them from one inner to outer coil and the other from outer to inner coil back.
This connection would look like this:
from inner windings coil a)p - a)2p - a)2p - a)p to outer windings in series with second coil outer b)p - b)2p - b)2p - b)p to inner windings again.
But still, there is the possibility to connect them alternating, which may be even better. This connection would look like this:
from the first inner coil a)p - b)2p - a)2p - b)p - a)p - b)2p - a)2p - b)p alternating back to the second inner coil.
This kind of connection would yield to some signal "mixing" ,instead of splitting between the coils, but I`m really unsure about the consequences or effects (advantages?).
Since I could`t find lot of information about this special topic on the net, I´m posting here. Hope that anyone out there has some thoughts, suggestions or probably experiences.
Maybe I´m completely wrong and the best thing is to connect them even another way round.
In contrast to the most, this design uses one big c-core with two sectioned coils.
My question now is, how to interconnect the sections, resp. which kind of connection would yield to the best frequency response?
Sections per coil are:
coil a) and coil b): p - s - 2p - s - 2p - s - p each
The secondaries are connected in parallel for all versions, so they aren`t shown any more.
For primary series interconnection, with respect to capacitance distribution, I think (also from comparison with existing designs), the best would be to connect them from one inner to outer coil and the other from outer to inner coil back.
This connection would look like this:
from inner windings coil a)p - a)2p - a)2p - a)p to outer windings in series with second coil outer b)p - b)2p - b)2p - b)p to inner windings again.
But still, there is the possibility to connect them alternating, which may be even better. This connection would look like this:
from the first inner coil a)p - b)2p - a)2p - b)p - a)p - b)2p - a)2p - b)p alternating back to the second inner coil.
This kind of connection would yield to some signal "mixing" ,instead of splitting between the coils, but I`m really unsure about the consequences or effects (advantages?).
Since I could`t find lot of information about this special topic on the net, I´m posting here. Hope that anyone out there has some thoughts, suggestions or probably experiences.
Maybe I´m completely wrong and the best thing is to connect them even another way round.
I am no expert but I have read up somewhat on the subject and done some OPT's with reasonable results. My 2 cents:
Since you are running SE you have the same signal in all primary sections so no real need for mixing. Leakage inductance is more or less set by the winding arrangement.
What you can influence is effective pri to sec capacitance. To minimise this you should arrange your windings to have the biggest signal swing in the winding with smallest area exposed to a secondary winding, i.e. the anode should be connected to the smaller inner windings and B+ to the outer windings.
Since you are running SE you have the same signal in all primary sections so no real need for mixing. Leakage inductance is more or less set by the winding arrangement.
What you can influence is effective pri to sec capacitance. To minimise this you should arrange your windings to have the biggest signal swing in the winding with smallest area exposed to a secondary winding, i.e. the anode should be connected to the smaller inner windings and B+ to the outer windings.
Okay, that mixing is not really necessary, is what I hoped (because of the added connection complexity).
But there`s still a question left. Following the rules of sec. to prim. capacitance, I should connect from inner windings of coil one series to outer windings of this coil and further (were signal swing is half) on to the inner windings of the second coil, series to this ones outer windings to V+.
What i saw on the Lundahl homepage (which is one of the few companies that use two coil c-cores), that they connect their less sectioned output trannies (f.e. the LL1623) series from inner section (tube anode) of coil one to the outer section of coil two, then to the outer sections of coil one and further on ending with the inner section of coil two for the V+ connection.
They mix the windings between the coils. Furthermore with the two coil arrangement, we have two points (inner windings), which have the smallest prim. to sec. capacitance. Think the Lundahl guys know what they`re doing.
Maybe there`s some kind of more uniform capacitance distribution, remembering that the V+ connection is earthy in signal terms.
It is probably a good thing to have in this direction, where the signal gets smaller (due to the series division), a decreasing capacitance again (from the outer to the inner windings of coil two) - instead of an increasing one.
The compromise could be a non-mixing arrangement from inner coil one to the outer sections, on to the outer sections of coil two, ending with the low prim. to sec. capacitance of coil two`s inner windings for the V+ connection.
What do you think out there?
But there`s still a question left. Following the rules of sec. to prim. capacitance, I should connect from inner windings of coil one series to outer windings of this coil and further (were signal swing is half) on to the inner windings of the second coil, series to this ones outer windings to V+.
What i saw on the Lundahl homepage (which is one of the few companies that use two coil c-cores), that they connect their less sectioned output trannies (f.e. the LL1623) series from inner section (tube anode) of coil one to the outer section of coil two, then to the outer sections of coil one and further on ending with the inner section of coil two for the V+ connection.
They mix the windings between the coils. Furthermore with the two coil arrangement, we have two points (inner windings), which have the smallest prim. to sec. capacitance. Think the Lundahl guys know what they`re doing.
Maybe there`s some kind of more uniform capacitance distribution, remembering that the V+ connection is earthy in signal terms.
It is probably a good thing to have in this direction, where the signal gets smaller (due to the series division), a decreasing capacitance again (from the outer to the inner windings of coil two) - instead of an increasing one.
The compromise could be a non-mixing arrangement from inner coil one to the outer sections, on to the outer sections of coil two, ending with the low prim. to sec. capacitance of coil two`s inner windings for the V+ connection.
What do you think out there?
Think the best way is to try both connection possibilities (without the complex mixing one).
The one following your comments and suggestions hemgjord (thanks a lot) which end in a series inner/outer to inner/outer coil connection.
And as an alternative the series inner/outer to outer/inner connection.
Comparison could be done by listening and measuring the frequency response (should work with my 24-bit/192kHz AD-converter and some software) of both versions.
The one following your comments and suggestions hemgjord (thanks a lot) which end in a series inner/outer to inner/outer coil connection.
And as an alternative the series inner/outer to outer/inner connection.
Comparison could be done by listening and measuring the frequency response (should work with my 24-bit/192kHz AD-converter and some software) of both versions.
- Status
- Not open for further replies.