Toroidal transformer input and output waveforms.

[neazoi]

Hello,
I am trying to build a simple form of single bit ROM (core rope memory) for illustration purposes.Before proceeding with the design, I am making some tests on a transformer.

The picture shows a pulse comming out of a 555 timer, into one turn on the primary.
The secondary is connected to an oscilloscope.
Every time a pulse occurs on the primary, I get this strange waveform on the secondary. It is a sine wave and it's amplitude degrades over time.

I need a brief explanation of why is this happening?
Also is there any way to practically find at what point a square curve core is saturated?

 

[jiripolivka]

Hello,
I am trying to build a simple form of single bit ROM (core rope memory) for illustration purposes.Before proceeding with the design, I am making some tests on a transformer.

The picture shows a pulse comming out of a 555 timer, into one turn on the primary.
The secondary is connected to an oscilloscope.
Every time a pulse occurs on the primary, I get this strange waveform on the secondary. It is a sine wave and it's amplitude degrades over time.

I need a brief explanation of why is this happening?
Also is there any way to practically find at what point a square curve core is saturated?

You cannot input into "one-turn primary" from the 555 as it shorts the 555 output. To inject a pulse into such extremely low impedance like one turn, you must use a high power, low output impedance amplifier. Try an IRF540 MOSFET, and check the current pulse magnitude: it can be up to 5 A, from a 5V DC source, but as you do not indicate the wire gauge of the primary "one turn", you can melt the wire by a too high current.

- - - Updated - - -

FRom your picture the core seems not to be saturated (the input pulse is too weak). You do not indicate the load o the secondary which is also very important.

 

[neazoi]

You cannot input into "one-turn primary" from the 555 as it shorts the 555 output. To inject a pulse into such extremely low impedance like one turn, you must use a high power, low output impedance amplifier. Try an IRF540 MOSFET, and check the current pulse magnitude: it can be up to 5 A, from a 5V DC source, but as you do not indicate the wire gauge of the primary "one turn", you can melt the wire by a too high current.

- - - Updated - - -

FRom your picture the core seems not to be saturated (the input pulse is too weak). You do not indicate the load o the secondary which is also very important.

Yes I initially sorted the output of the 555, so I have tried to put a 100nF capacitor at the output so not DC passes through. The pulse was able to pass through the capacitor and induce the waveform shown in the diagram on the secondary.

Then I ingreased the number of turns on the secondary from 10 to 30, so now the primary was 1 turn and the secondary 30 turns. I got a much greater amplitude output and the same waveform shape, as it was expected.

The load in both cases were just an oscilloscope connected at the output. I have also tried to connect a diode accross one of the secondary windings. This diode rectified the waveform and let only the top side (positive) to pass through the oscilloscope.

In a ROM configuration, core saturation is not desired, since in every prescending pulse you need to get an output. If the core has been saturated by a positive pulse, the next positive pulses would have no output effect I think.


By the way, have you got any ideas, of how could I detect if a core is saturated or not? Maybe by applying more pulses and notice if there is a waveform on the output or not?

Also, what is the saturation effect of using smaller ferrite cores, will they need less turns or power to saturate?

 

[jiripolivka]

Yes I initially sorted the output of the 555, so I have tried to put a 100nF capacitor at the output so not DC passes through. The pulse was able to pass through the capacitor and induce the waveform shown in the diagram on the secondary.

Then I ingreased the number of turns on the secondary from 10 to 30, so now the primary was 1 turn and the secondary 30 turns. I got a much greater amplitude output and the same waveform shape, as it was expected.

The load in both cases were just an oscilloscope connected at the output. I have also tried to connect a diode accross one of the secondary windings. This diode rectified the waveform and let only the top side (positive) to pass through the oscilloscope.

In a ROM configuration, core saturation is not desired, since in every prescending pulse you need to get an output. If the core has been saturated by a positive pulse, the next positive pulses would have no output effect I think.


By the way, have you got any ideas, of how could I detect if a core is saturated or not? Maybe by applying more pulses and notice if there is a waveform on the output or not?

Also, what is the saturation effect of using smaller ferrite cores, will they need less turns or power to saturate?

Please be kind to the poor 555! I once blew one when I misread 50 Ohms for 500 and the load was also coupled thru a capacitor. For single ohm load like yours (even 30 Ohms is seen by a 555 as a short) you need a good pulse power amplifier!
The saturation means the core does not grow more magnetization with more drive. On the secondary you should see on the scope a flattening of the pulse. Now you can see a sharp rise without the flat top, so no saturation.
I worked with RF ferrite cores and have no experience with effect you study. MOst of problems are due to the core material, something not too well defined in specifications.
You can try to see some saturation effects if you bring a ferrite or neodym magnet close to your ferrite ring. If it stays afterwards, be ready to demagnetize by an open-core AC transformer (earlier used to erase magnetic tapes.

 

[Venkadesh_M]

Is that matches a secondary of a CT open circuited??

- - - Updated - - -

open-circuit-in-a-ct-secondary

If the primary circuit has current flowing the secondary circuit should never be opened. This can cause very high voltages to occur due to the Ampere-Turns of the primary that start magnetizing the core. This results in the core being saturated through a large portion of the cycle. When a transformer core is saturated, it is a nontransformer. Consequently, the transformer acts as a transformer only during a short period above and below the zero to maximum in the opposite direction. While it is acting as a transformer it will cause very high voltage peaks. The voltage may not appear very high when the measurement is taken with an RMS Volt-Meter but a Peak Volt-Meter would indicate a much higher peak value. The peaks can be sufficient enough to cause death.

- - - Updated - - -

 

[neazoi]

Ok,
This is the circuit I have made: **broken link removed** no text description yet.
By coupling the 555 directly to the transformer primary, I get the waveform described in post #1. But by coupling the 555 with a 100nF to the transformer primary, I get the waveform shown in the last picture in the website.
The top line is the output pulses as comming out from the coupling capacitor (which is connected to the primary).
The bottom line is the output pulses from the secondary.

I think this makes sense now. A rising edge of the positive pulse (from 0 to 5v), causes a positive pulse on the transformer secondary. Then, as the pulse top is at 5v, no voltage is at the secondary. Finally when the falling edge of the pulse comes (from 5v to zero), it causes a negative pulse on the transformer secondary.

I hope this is true.

 

[Venkadesh_M]

Yeah it seems like that... but how do you saying It is a ROM and how do u demonstrating??

 

[neazoi]

Yeah it seems like that... but how do you saying It is a ROM and how do u demonstrating??

I am writing the description of the page now, so I will make it available in a few hours for your comments.

 

[Venkadesh_M]

Ok let us see...

 

[neazoi]

Ok let us see...

Ok here it is **broken link removed**
I will include a small video of the test setup later this afternoon.
Your comments are welcome.