help with saturable reactor

[neazoi]

Hello,
I am making a saturable reactor using toroids. I am using a 220v-24vac transformer as a power source for the load and the load is a miniature 24v incandescent lamp. I have tried different sizes of cores, square loop ones and normal ones but non of them is working. I have tried different configurations for the primary (700T, 200T, 100T, 50T) and for the secondary (100T, 50T, 25T, 10T).
The reactor does not seem to work at all. I always see the bull at full brightness. Even if I bring a powerful magnet close to the core, no change in the bulb brightness.

What am I doing wrong?
Maybe the cores I have used cannot respond to the 24V 50Hz AC ?

 

[neazoi]

Ok, I just noticed something weird is happening.
I have tried some 220v to 12v transformers and they did not work. But when I connected them the other way round (220v winding to the load and 12v winding to the DC control) it worked! Although the control current had to be raised to 2-3A to control the load.
But I think this is not the correct way to connect them, the DC winding has to have more turns than the load wiring, am I wrong?

Could it be the case that the load winding must have lots of turns too (but less than the DC one) so that it presents a significant reactance (which will be varied).

 

[FvM]

As a simple point, a saturable inductor has highest reactance without DC bias. If you use it as series power control element, the reactance must be designed large enough to reduce the output power to the intended minimal level. Apparently this isn't the case in your design. But without quantitative design data it's just a guess.

Some basic magnetic amplifier concepts:
https://archive.org/download/MagneticAmplifiers/MagneticAmplifiers.pdf

As additional comment to your latest post: Using a single transformer as saturable inductor brings up serious problems to isolate the control circuit, respectively the risk to short the reactance through the control winding.

 

[neazoi]

Yes I am aware about these issues, I just wanted to see at what point the core saturates.
Thank you for let me know about the fact of the load reactance that must be designed large enough to reduce the output power to the intended minimal level. This means that the load winding must have quite large number of turns too I guess, so that without the core saturated, it must provide minimum current to the load. hm... a bit try and see may be needed. I am using 3R1 material ferroxcube cores, but I cannot see in the datasheets such info.

 

[FvM]

Realistically, a 50 Hz saturable inductor will mostly use laminated or tape wound iron cores rather than ferrite. 3R1 is a dedicated mag-amp material, but more suitable for > kHz frequencies.

 

[neazoi]

Realistically, a 50 Hz saturable inductor will mostly use laminated or tape wound iron cores rather than ferrite. 3R1 is a dedicated mag-amp material, but more suitable for > kHz frequencies.

An audio magnetic amplifier must have a core that will accept both ultrasonic as well as audio. Literature states 1:5 difference at least, which brings the RF frequency to 100KHz for a maximum of 20KHz (although the responce of the amp could not be so high). Thus the core must cope with low audio frequencies as well as 100KHz or so. I think a compromise is done somewhere increasing the loss at some frequencies.
I have tried 700T for the control and 100T for the load wiring at the biggest size 3R1 core. Still no success. But without any power to the control winding there is no reduction to the bulb brightness, so I guess more turns are needed for the load wiring.
There may be a case that the load wiring saturates the core (unwanted situation).
The solution to this as suggested in the literature, is to use opposite phase windings for the load, but with a single load wiring, I cannot figure if such a case is happening to be honest.

 

[FvM]

I don't know how "audio magnetic amplifiers" are related to the original post. Their operation (modulating a DC power source) is obviously quite different from a AC power controller.

Generally speaking, in both cases the actual core and winding data have to put into the calculation.

As I already mentioned, using a single core/single magnetical core for the magnetic amplifier, in constrast to usual transductor or dual core topologies, results in tight, possibly unrealistic requirements for the control circuit, e.g. need for a current source with large voltage compliance. But I presume, this details can be well calculated.

 

[neazoi]

I don't know how "audio magnetic amplifiers" are related to the original post. Their operation (modulating a DC power source) is obviously quite different from a AC power controller.

Generally speaking, in both cases the actual core and winding data have to put into the calculation.

As I already mentioned, using a single core/single magnetical core for the magnetic amplifier, in constrast to usual transductor or dual core topologies, results in tight, possibly unrealistic requirements for the control circuit, e.g. need for a current source with large voltage compliance. But I presume, this details can be well calculated.

If a center tapped load wiring is done the second core can in theory be avoided, as the two windings fields produced will cancel inside the core.
This has the advantage that with a single core, the control winding is not affected by the load winding and also the load winding does not saturate the core accidentally. But this is true only in magnetic amplifiers with full rectification, not saturable reactors.
I will try something like 200T for my saturable reactor to see how it goes.

 

[neazoi]

The 700T has about 20R resistance and this is not proven efficient when "regulating" 50Hz, when no saturation.
On the other hand, another transformer I have shows good results with a 120R winding.
But to get 120R I have to wind 7000T on the toroid or so. Is this realistic?
Maybe the REACTANCE has a different value and this is what plays role in a magnetic audio amplifier (100KHz PSU frequency)

What I need to ask is that the load winding may have a low resistance, but it may present a much more reactance at 100KHz, effectively regulating this frequency, whereas with the same number of load turns it will be ineffective on 50Hz.

Is that the case?
If yes, I may better start to test my cores using a 100KHz signal instead of a 50Hz one.

 

[neazoi]

The 700T has about 20R resistance and this is not proven efficient when "regulating" 50Hz, when no saturation.
On the other hand, another transformer I have shows good results with a 120R winding.
But to get 120R I have to wind 7000T on the toroid or so. Is this realistic?
Maybe the REACTANCE has a different value and this is what plays role in a magnetic audio amplifier (100KHz PSU frequency)

What I need to ask is that the load winding may have a low resistance, but it may present a much more reactance at 100KHz, effectively regulating this frequency, whereas with the same number of load turns it will be ineffective on 50Hz.

Is that the case?
If yes, I may better start to test my cores using a 100KHz signal instead of a 50Hz one.

Yesterday I tried if with a 100KHz sinewave oscillator connected in series with the control winding, just to see if there will be any reduction in output voltage, from this 700T winding (20R). There was no reduction.
What am I doing wrong?