You just repeated what I already said - about 2 side isolation of toe lanimation. Does not matter what shape of the lamnation until alloy is right ro the specific aplication. And off course windings space...
cheers
This is not about undersizing core afainst neded power, in this example of iron core from the photos, the winding space almpst does not exixsts, it is for special purpose core capable to accept very small nuber of turns... 🙁
There are some first steps n the design process. Balancing parameters are not in the first steps area...
First step is to chosse a core aloy and consider window space.
Impedance transformation ratio is certanly not first thing.
Your goal has to be
transfer BW with minimum phase shift, permeability of lamination core, Primary inductance, Leakage inductance, winding capacitances, and so
then minimum histeresis distortion - induction of primary, air gap, magnetic lines lenght,,,
...
If You from the start chose core material with inapropriate characteristics
Does not have the much point is there a transfer impedance ratio is good.
...🙁
Guess what?
You cannot guarantee "minimum phase shift" in the high-frequency region without taking acount the transformer impedance ratio first. The impedance ratio will take part in the criteria of determining your geometry.
To realize things further, you need to be aware of your secondary load and your driving impedance.
You cannot guarantee "minimum phase shift" in the high-frequency region without taking acount the transformer impedance ratio first. The impedance ratio will take part in the criteria of determining your geometry.
To realize things further, you need to be aware of your secondary load and your driving impedance.
So your first step is chose a core with a permeability of 800? Really? You can only do that by adjusting the gap of the core. And even then there are fluctuations.
btw power transformers have not a permeability of 250-400, totally wrong.
manufacturers rarely give relative permeability data. Most likely that this data You will have to establish with measurements amd calculations. I explained the way how to fing it earlier. From my measurements of diffefent lamination from power xfrm lamination that is aprox value range. (Some C cores have larger about 800...)
...
What value did You find about power transformer lamonations relative permeanility?
Last edited:
That is not a huge addition of Magnetic lines length. And will not decrease Lp[Hy] as strong as lack of relative permeability.
For lm is more important to keep the ratio lm / air gap. And gap in this case will be slight wider. Leading to decrease of Lp.
That is why the permability should be consider seriously.
For signal transformers we are talking of extremly high permeability in the range of thousents. That is different types of the transformers.
Without any DC component trough the windings. For SE OT You need higher permeability than ususal is for power transformers.
.
I want to say that people using random cores without information of permeability value. Fill it with the layers of wire having in mind just famos impedance ratio. And with arbitary gap. And that is almost all. 🙂 On the schematics we dont have any other informations, about for instance Lp, Ls, LpRdc, Rdc-sec, for 30 seconds direct measurements.
Impedance ratio is a must and it coming from load line of the tube at first. So this is the some data thet You are considering when You design Tube stage. Having in mind Internal resistance of the tube, the first step to design transformer is to calculate Lp needed for LF.
.
Load line in LF region is not a flat line but some elipse. For insuficent Lp[Hy] that elipse is "fat" and tend to circle. The elipse shape is crucial for SE design at low F. because it should never touch X axis. Actually IF the Lp is low the elipsoid load lie at LF is deformed in that area touching X axis...
...
That is Because of reactive load aplied to the tube - nor resistive as it calculated in static representatios of anode curves. Even for the 1KHz the elipsoid shape is present with thiny elipse. Many people didt aware of this fact at all.
.
Secondary load is known variable. And amount of secondary turns (square function) is not changed a lot as amount of primary turns against the desired Lp(permeability, air gap...)
Last edited:
M cores are less desired because of fixad gap. But it could be fit. E cores have very flexibile control of air gap. And what is more important
could be find with wider window space, and a higher relative permeability. Within the 0.35mm laminations.
Other thing is that You can apply additional layer of insulation to the each lamination.
That window is 40x25mm=1000mm^2, it almost equals to EI114, 57x19mm. The biggest disadvantage of EI core is height of window that far too limited, only 19mm. It hard to contain any large coils. It's first time I customed the silicon iron directly from a factory. I amnot satisfaction with that, but it work better than most EI core I have. It can delivery 18W evan at 20Hz, without any observed distortion.
How many turns do you need for a coil?
I have many skills and method to increase GM or reduce the Ra of Power tube. By parallelling anohter one, or using CFB wilding and give some local NF to G2, I can highly reducing the Ra of power tube, for example 6L6G, from 2KR triode connection down to 300R. Even at 2.5KR for the RLa I can get a good damping factor, and keeping the primary coils is under 2000turns at same time. So, that windows is enough for me.
I am not looking for a perfect core, just a good one especially that grands of silicon iron is still producing. And make a good wilding for that core with a resonable price.
Last edited:
This is not definitive. Especially for low primary Z impedance transformers.
The priority of interest at the elliptical loadlines is entirely up to you. Some designers do not stress it so much. I usually aim for a low-signal unloaded secondary cutoff of -3dB of 5-7Hz in my SE transformers, and a full power capability of 25Hz before saturation knee.
A lower reflected primary transformer does not necessarily get smaller capacitance than higher reflected primary impedance.
In fact, the lower you go in primary impedance, the more you can increase capacitance while decreasing leakage inductance.
Lower pimary impedance transformers need more interleaving to keep leakage HF inductance roll-off when transfering power.
On the countrary, you need less interleaving with high primary impedance transformers, where you must decrease capacitance and accept more leakage inductance.
An audio transformer doesn't need to be necessarily interleaved, for example a 30k tube Rp output transformer.
Here's an important tip however. The more you increase leakage inductance, the more careful you need to be with capacitance distribution. Beginners that get slapped by a few dipping resonances rely on paralleled equal turn secondaries, this is your safest bet. Low primary impedance transformers are also easier, they are more forgiving into neglection of capacitance distribution due to their lower leakage inductance primary sections resulted from lots of interleaving. They are also easier on capacitance requirements.
Remember that capacitance decays exponentially from the anode, where at the anode end it is the highest and is equal to the static capacitance, if we assume the secondary is at ground potential, as is the case with most OPTs. At the B+ end you basically get zero capacitance due to the lack of voltage difference between layers there. Remember that to charge capacitance, you need a difference in voltage potential. So the first rule of thumb is keeping the anode far away from ground as possible. Later you can learn tricks like "capacitance dumping", I came on my own with this term. You can learn to deliberately dump capacitance in regions where it's less offensive to the HF response.
In high-z interstage transformers it gets trickier and you need to be more careful with their design. Windings direction and connection are more critical, where if done wrongly, you can have big packages of Cs and Ls resonating into dips, sometimes dangerously close to the audio range.
In fact, the lower you go in primary impedance, the more you can increase capacitance while decreasing leakage inductance.
Lower pimary impedance transformers need more interleaving to keep leakage HF inductance roll-off when transfering power.
On the countrary, you need less interleaving with high primary impedance transformers, where you must decrease capacitance and accept more leakage inductance.
An audio transformer doesn't need to be necessarily interleaved, for example a 30k tube Rp output transformer.
Here's an important tip however. The more you increase leakage inductance, the more careful you need to be with capacitance distribution. Beginners that get slapped by a few dipping resonances rely on paralleled equal turn secondaries, this is your safest bet. Low primary impedance transformers are also easier, they are more forgiving into neglection of capacitance distribution due to their lower leakage inductance primary sections resulted from lots of interleaving. They are also easier on capacitance requirements.
Remember that capacitance decays exponentially from the anode, where at the anode end it is the highest and is equal to the static capacitance, if we assume the secondary is at ground potential, as is the case with most OPTs. At the B+ end you basically get zero capacitance due to the lack of voltage difference between layers there. Remember that to charge capacitance, you need a difference in voltage potential. So the first rule of thumb is keeping the anode far away from ground as possible. Later you can learn tricks like "capacitance dumping", I came on my own with this term. You can learn to deliberately dump capacitance in regions where it's less offensive to the HF response.
In high-z interstage transformers it gets trickier and you need to be more careful with their design. Windings direction and connection are more critical, where if done wrongly, you can have big packages of Cs and Ls resonating into dips, sometimes dangerously close to the audio range.
Last edited:
My Vacuumschmelze databook is full of those permeability curves. I have also data from other manufacturers from all over the world. It’s not so rare as you say. But skip the EI cores.
Hi 50AE
Thank you. I appreciated your detialed explanation.
You siad " So the first rule of thumb is keeping the anode far away from ground as possible." Let's using Mr. Turner's drawing as an example. If Point No1 connect to anode, I should avoid make Point A be at ground potential. Am I right?
If a singal swimmng 70Vrms on primary wilding, secondary is only few volt. What the different if I let Point A at ground.
Attachments
Last edited: