Hi all
I now understand the meaning of your words in concrete terms.
I need to review the assembly of my secondaries of my circlotron OPT.
I 'll post pics before and after review
I now understand the meaning of your words in concrete terms.
I need to review the assembly of my secondaries of my circlotron OPT.
I 'll post pics before and after review
Basically, I found out the best results come when combining capacitance dumping towards the primary together with the secondary connection rules. I cannot prove yet, academically in details, exactly why this works, but for the moment I'm using my own rope-pull imagination/philosophy. Let's review the latest interleaving example for an explanation attempt.
---------------S------------
////////////////////////////// - Cpsf = negl.
B+------------P1----------g
f--------------P2----------g
---------------------------- - Cpf = 0,25
d-------------P5----------c
b-------------P6----------c
b-------------P7----------a
Anode-------P8----------a
---------------------------- - Cpf = 0,25
d-------------P4----------e
f--------------P3----------e
///////////////////////////// Cpsf = 0,1
---------------S-----------
By having a dominant Cpf factor of 0,25 between P8 and P4, basically, more capacitive currents are pulled there, so less current flows between the bottom Cpsf of 0,1 between P3 and the bottom secondary. This, in my findings is always beneficial for fighting the dip resonance.
I also found out that by aiming to equalize the Cpf to Cpsf, dip to peak resonances tend to flatten against each other, so one can get a close to flat frequency response.
On the other end, if for example, the Cps get close to zero and Cp becomes dominant, you'll go towards a dominant peak resonance. The other way around, a dominant Cps gets you closer to a dip resonance.
Connecting the wrong secondary layers in series moves the dip resonance closer to the audio band.
Quoting the same schematic again, with an example of self screening secondaries. Add two additional outer secondaries that you can connect in series. The parallel secondaries screen the series ones.
---------------Sseries-------
---------------Sparallel-----
////////////////////////////// - Cpsf = negl.
B+------------P1----------g
f--------------P2----------g
---------------------------- - Cpf = 0,25
d-------------P5----------c
b-------------P6----------c
b-------------P7----------a
Anode-------P8----------a
---------------------------- - Cpf = 0,25
d-------------P4----------e
f--------------P3----------e
///////////////////////////// Cpsf = 0,1
---------------Sparallel-----
---------------Sseries-------
---------------S------------
////////////////////////////// - Cpsf = negl.
B+------------P1----------g
f--------------P2----------g
---------------------------- - Cpf = 0,25
d-------------P5----------c
b-------------P6----------c
b-------------P7----------a
Anode-------P8----------a
---------------------------- - Cpf = 0,25
d-------------P4----------e
f--------------P3----------e
///////////////////////////// Cpsf = 0,1
---------------S-----------
By having a dominant Cpf factor of 0,25 between P8 and P4, basically, more capacitive currents are pulled there, so less current flows between the bottom Cpsf of 0,1 between P3 and the bottom secondary. This, in my findings is always beneficial for fighting the dip resonance.
I also found out that by aiming to equalize the Cpf to Cpsf, dip to peak resonances tend to flatten against each other, so one can get a close to flat frequency response.
On the other end, if for example, the Cps get close to zero and Cp becomes dominant, you'll go towards a dominant peak resonance. The other way around, a dominant Cps gets you closer to a dip resonance.
Connecting the wrong secondary layers in series moves the dip resonance closer to the audio band.
Quoting the same schematic again, with an example of self screening secondaries. Add two additional outer secondaries that you can connect in series. The parallel secondaries screen the series ones.
---------------Sseries-------
---------------Sparallel-----
////////////////////////////// - Cpsf = negl.
B+------------P1----------g
f--------------P2----------g
---------------------------- - Cpf = 0,25
d-------------P5----------c
b-------------P6----------c
b-------------P7----------a
Anode-------P8----------a
---------------------------- - Cpf = 0,25
d-------------P4----------e
f--------------P3----------e
///////////////////////////// Cpsf = 0,1
---------------Sparallel-----
---------------Sseries-------
Things starting to make sense to me about windings capacitance many thanks to you.
How about Leakage Inductance; is it anything to consider other than interleaving, keeping the total winding height low and neat winding?
How about Leakage Inductance; is it anything to consider other than interleaving, keeping the total winding height low and neat winding?
Leakage inductance is simpler and has normal, textbook rules. The good news is, it is only dependent on geometry and turn count. To minimize it, keep the 1/2 outer layer rule and respect as identical layer primary/secondary lengths as possible. Keep dielectrics tensioned, try to not to have air pockets. If you have to wind on a tall, short coil, you can compensate with more interleaving or lower turn count.
Some folks believe leakage inductance is bad and tend to minimize it as low as possible due to worsened resonances, but my findings say that is the high leakage inductance + bad capacitance distribution that does this. One can wind high leakage transformer with smooth frequency response. I usually aim for leakage roll-off of 100kHz and 60kHz for big, 100W SE transformers. Of course, there are some exception where you should highly take leakage inductance into consideration, such as interstage step-up transformers, where the Ls will resonate with the next stage Miller capacitance.
Some folks believe leakage inductance is bad and tend to minimize it as low as possible due to worsened resonances, but my findings say that is the high leakage inductance + bad capacitance distribution that does this. One can wind high leakage transformer with smooth frequency response. I usually aim for leakage roll-off of 100kHz and 60kHz for big, 100W SE transformers. Of course, there are some exception where you should highly take leakage inductance into consideration, such as interstage step-up transformers, where the Ls will resonate with the next stage Miller capacitance.
Thank you again 50AE, one last question if I may; what insulation do you use between P-P layers and also P-S ?