Square loop toroid switching help

[neazoi]

Hi, I have a TN9/6/3 3R1 toroid and I would like to find out of I can switch it's magnetization using my RF signal generator operating at 7MHz AC.
I follow the procedure in this page https://meettechniek.info/passive/magnetic-hysteresis.html and especially the part entitled "Measuring arrangement with a digital oscilloscope". The scope is the HP 54520A.

I switch my generator on (50R around 7dBm) and I use the "Versus" function (pp173 in the manual) of the scope (X-Y mode).
With this setup, I see a curve very similar to this https://meettechniek.info/extern/images/square-ferrite-3r1.png (the black one). However in my experiment, the very top and the very bottom lines of the curve are completely parallel to the scope grid lines, in contrast to this picture.
Also, when I reduce the generator power, gradually, I start seeing a waveform more like an elipse (which I assume means that the core cannot be switched) and not a line like the blue curve of this picture https://meettechniek.info/extern/images/square-ferrite-3r1.png

I am confused in the description of the procedure in the page.

First it says:
"When a digital oscilloscope is used the coil voltage could be digitally integrated. Of course the scope must therefore have the math options including integrating signals..."

And then it says:
"In the X-Y mode the field strength H is plotted on the X-as, and the flux density B on the Y-as."

So do I do the right thing to switch the scope to the X-Y mode, or do I need to switch it in the integrate function? Btw, when I switch it to the integrate function a very low power sinewave is shown. It is only in the X-Y mode (versus) that I see a curve similar to the hysterisis loop.

Apart from this, I am not sure if the core is switched, or if the very top and bottom lines are just some kind of clipping from the scope. But wouldn't this "clipping" occur then at all sides of the loop (top left and bottom right as well)?

Please give me some advice on these issues.

 

[schmitt trigger]

I've been using this technique for many years with analog scopes. Digital scopes really made this measurement very simple, and actually had an article in Electronics World where I described the use of a digital scope's integrating function to see the hysteresis curves.

It is very simple procedure:
1- On the Channel 1, the current sample, which has been converted to voltage via a resistor or a hall cell transducer, is displayed and centered.

2- On the Channel 2, you first integrate the waveform with the scope's math functions. Also center the display.

3- Then, when the vertical amplitude of both waveforms are well within the scope's vertical scale, and at least a couple of periods are displayed on the horizontal, then you apply CH1 to X and CH2 to Y.


To me, the mistake most rookies make is going directly to step 3, without making steps 1 and 2. They are important, because it will allow you to double check if your waveforms in the time domain are clipping or not. This may explain how come the top and the bottom of the curves are flat.

Please post your scope images.

 

[neazoi]

I've been using this technique for many years with analog scopes. Digital scopes really made this measurement very simple, and actually had an article in Electronics World where I described the use of a digital scope's integrating function to see the hysteresis curves.

It is very simple procedure:
1- On the Channel 1, the current sample, which has been converted to voltage via a resistor or a hall cell transducer, is displayed and centered.

2- On the Channel 2, you first integrate the waveform with the scope's math functions. Also center the display.

3- Then, when the vertical amplitude of both waveforms are well within the scope's vertical scale, and at least a couple of periods are displayed on the horizontal, then you apply CH1 to X and CH2 to Y.


To me, the mistake most rookies make is going directly to step 3, without making steps 1 and 2. They are important, because it will allow you to double check if your waveforms in the time domain are clipping or not. This may explain how come the top and the bottom of the curves are flat.

Please post your scope images.

I would love to see your article, if this is available somehow.

Regarding point 1, The page showed, uses a resistor, if that is what you mean. So I use this way.

Point 2 is the big question. My scope has an integrate function, and another one called versus (X-Y). I cannot use both at the same time. You mean to use the integrate, to center the waveform and then deselect this and use the X-Y mode on step 3?
Please help me on this.

Regarding point 3, I first ensure that my waveforms on ch1 and ch2 in the time domain, are not clipping (actually well below the scope's clipping point), then I switch to the math functions to the X-Y mode.
Is that what you mean?

 

[schmitt trigger]

-Yes, Use a resistor. The Hall Cell is less intrusive, but a resistor is just fine.

-Unfortunately, different digital scopes have many different ways of doing things. I used a Tek scope, forget the actual model number, it was a TDS-something...the one with a floppy disk on front.
But the key is to apply the math function while you still have Ch1 and Ch2 active, you will have three waveforms displayed. Then de-select Ch2 so it is not viewable, and then you only have Ch1 and Math...... then change from YT display to XY display. That function must be there, somewhere in a menu.

- And yes, the purpose of displaying the waveforms as YT first is to ensure that they are not clipping.

See my attached article.

 

[FvM]

The intended setup requires integrate function to be useable in x-y mode. That's possible with some oscilloscopes (e.g. LeCroy Waverunner) and not possible with some others (e.g. not my Agilent 6000). Apparently it's not possible with yours.

In this case, you need to use an analog integrator for the induced coil voltage to display the hysteresis curve.

 

[neazoi]

The intended setup requires integrate function to be useable in x-y mode. That's possible with some oscilloscopes (e.g. LeCroy Waverunner) and not possible with some others (e.g. not my Agilent 6000). Apparently it's not possible with yours.

In this case, you need to use an analog integrator for the induced coil voltage to display the hysteresis curve.

That was exactly what I was wondering and thanks for pointing this out.
So I have to make the circuit with the external integrator, like if I had an analogue scope only available.

I used 50R for the R, is that ok?
Also any other opamps to use? TL071 TL081 for example?
I want to test this on HF frequencies 1-15MHz or so.

 

[FvM]

The integrator must work as such in the frequency range of interest. TL081 may be used up to a few 100 kHz maximum, for MHz range you need a RF OP or use a passive integrator.

 

[neazoi]

or use a passive integrator.

Please tell me what is it and how to build one?

The opamp cirquits I have seen use a split rail PSU which is not available here. A passive integrator would be fine, I just need to see if the core is switched or not I do not need precision.

 

[FvM]

A passive integrator is a RC low-pass dimensioned for a cut-off frequency e.g. 0.05 to 0.2 of the signal frequency. Series R according to the source impedance, e.g. 1k to 100k.