MC step up primary winding turns

UTC A10 & HA-100X for reference

UTC_A10_HA100X_blueprint.jpg
 
layers

Is it possible to connect the green primaries in parallel and the red secondaries is series to yield maximum step up transformation?

Yes.

Any idea how many layers primary secondary will work ok

In the blueprint it states the secondary has 25 layers each winding so total 50 layers. But no mention of primary.

I think the drawing itself illustrates 6 layers per winding so total 12 layers, if I interpret the drawing correctly.

=================

Posting both A20 & A10 side by side, just for reference side by side.
In the UTC A-20, a 600Ω:600Ω transformer, it also has 6 layers per winding so total 12 layers.

PS, Both blueprints are courtesy of CJ in groupdiy forum.

Have fun building!

UTC-A20-Blueprint.jpg UTC_A10_HA100X_blueprint.jpg
 
Last edited:

PRR

Member
Joined 2003
Paid Member
> both A20 & A10

These are 500 ohm windings. An MC needle wants 1/5th to 1/10th of that, 1/2 to 1/3 the turns.

OTOH these are relatively high-level cores compared to MC levels. You should use a smaller core. Then you need more turns because less iron inside.
 
Last edited:

PRR

Member
Joined 2003
Paid Member
Yes, self-capacitance. Also cost (high-mu metal is expensive).

I have not worked enuff with MC to know if the C really matters. (Certainly not on the MC side, but the 47k side is high-Z for a transformer).

And of course cost is relative to desire and available parts.
 
The C really matters at 47k and is why there are so many interleaved winding layers.
It is a challenge keeping the resonances safely above 20 kHz

The transformer is presenting a load resistance of about 100 Ohms to the MC and a source resistance of around 2k to its 47k loading, so fortunately not actually a 47k transformer
Turns ratio is maybe 1:20
 
well my first transformer not too bad, 1:10 Ratio

Core E I Super oriented grain 30mm x 25mm x12mm 0,23mm

Sec 40k
primary 400ohm
15hz -3db
20hz -2db
30hz-20khz 0db
55khz -3db
test with 300mVrms in ( minimum of my generator )

Layers : Primary 3 Secondary 4 1 coil



Any idea to do better at 20hz looses
 
47k load applies for moving magnet cartridges. It is typically set by a fixed resistor at the input of the phono preamplifier. You can modify it at your convenience. A typical MC cartridge needs 10 to 100 ohms load. It all depends on your MC cartridge. Nothing prevents you from adding a resistor parallel to the secondary, this way reducing the reflected load on the primary. Your MC cartridge will be even happier. And the lower cutoff frequency will go down, because:

T = L/R, T = 1/ω, T = 1/2*π*f

If you decrease R, T goes up and f goes down. No need to increase the number of turns, because the interwinding capacitance will increase, and the upper cutoff frequency will go into the audible band.
 
47k load applies for moving magnet cartridges. It is typically set by a fixed resistor at the input of the phono preamplifier. You can modify it at your convenience. A typical MC cartridge needs 10 to 100 ohms load. It all depends on your MC cartridge. Nothing prevents you from adding a resistor parallel to the secondary, this way reducing the reflected load on the primary. Your MC cartridge will be even happier. And the lower cutoff frequency will go down, because:

T = L/R, T = 1/ω, T = 1/2*π*f

If you decrease R, T goes up and f goes down. No need to increase the number of turns, because the interwinding capacitance will increase, and the upper cutoff frequency will go into the audible band.


Thanks to explain, will check with 20k sec load
 

PRR

Member
Joined 2003
Paid Member
> test with 300mVrms in ( minimum of my generator )
> Any idea to do better at 20hz looses


Losses? Or saturation? Can you see the shape of the output?

Control of audio level is THE BASIC skill of the audio engineer. Don't let your generator tell you how low you can go. Also for fake-MC you want a lower source impedance than most generators. This will cost you two resistors. 1k and 10 Ohm will make 300mV into 3mV at 10 Ohm source. 3mV is still fairly high for MC.
 
> test with 300mVrms in ( minimum of my generator )
> Any idea to do better at 20hz looses


Losses? Or saturation? Can you see the shape of the output?

Control of audio level is THE BASIC skill of the audio engineer. Don't let your generator tell you how low you can go. Also for fake-MC you want a lower source impedance than most generators. This will cost you two resistors. 1k and 10 Ohm will make 300mV into 3mV at 10 Ohm source. 3mV is still fairly high for MC.


Thanks PRR, the sinusoidal wave at 20hz is very good not deformation by saturation core , my generator is 50ohm, I measured with the generator in the transformer out, 300mV in to 30mV out and no changes the looses.
 
The DC resistance of windings in a signal transformer should be a lot less than the AC impedances involved, otherwise they are a major source of loss/attenuation. And also for low level signals the resistance is a source of extra Johnson noise. Something like 1 ohm would be a great primary DCR for an MC transformer, given some MC cartridges are a few ohms themselves. That's a tall order, so there will be compromise.

Similarly the winding inductance (with other windings open circuit) should have much more impedance than the signal impedances, in order to avoid low-frequency losses. Even 1H of inductance has only 125 ohms of impedance at 20Hz. For an MC transformer that reflects 100 ohms back to the cartridge you'd like the primary inductance to be 10H or so, making the 20Hz inductive reactance > 1k. Again that is a tall order - some compromise will inevitably happen here too.

These issues are why audio signal transformers often use permalloy or other ultra-high permeability cores to get winding inductances high without needing high turns-counts (which would create too much winding resistance).

The end result is that an MC transformer has to be surprisingly large and expensive to perform well (like any good audio signal transformer) even though the power being converted is typically nanowatts to microwatts - its not saturation that's the issue at these power levels, its basically the frequency response at the low end that is difficult to achieve without lots of copper and iron/nickel, and the shielding to keep out hum - I suspect toroidal cores are often used too to reduce hum pickup, making for an expensive precision wound part.