The Lock In has a 5v output but it may be that at 5kHz it does not produce sufficient current to generate a significant magnetic field via the coil. This is dependent on the characteristics of your coil.

I would be tempted to drive the coil directly via a power supply that can cope with inductive loads (if they are large). Kepco do some. You can then measure the input signal via a shunt.

What field do you get?

Dear Rao,

While doing the ME voltage, the sample should be kept at center in the magnetic poles. Some times its depends on the field we applied and also the nature of the samples. Are you using bulk or think films. By using the Lock in Amplifier, we can get the signal with out any noise. How are you measuring the out put voltage.

Is that ME setup built in your lab or purchased from some where?

I worked on my composites, I have given the ME setup pic in my papers. Please check some of my papers. May be you will get some idea.

What composites are you preparing. Wishing you good luck.

Hi all,

I also have the same exact problem mentioned in the question.

I have pellets of BTO/CFO prepared in different compositions and also pellets of pure BTO and CFO.

In lock-in based measurement method, we run Helmholtz coil using keithly sourcemeter and for driving dc magnetic field, we have kepco bipolar power supply as source. At first, we were getting high bias voltages and noise due to the shielding and grounding issues. But, now even after eliminating those problems, we get the same small signal for all samples. We were suspicious whether the ME coupling coefficient of my samples are low due to the unwanted high conductivity of the samples. However, with BTO and CFO showing the same behavior and signal as the mixed samples, it seems that there is something wrong with the measurement, apart from having not coupled enough samples.

I highly appreciate any help to figure out the reason.

Paul Andrew Bartlett , thank you for your comments. For an ac voltage of 5V at

5 kHz, I obtain a coil current of around 55 mA, which corresponds to an ac magnetic field of about 5 Oe. I have studied the frequency dependence of the inductive load of the coil, and also the coil specification. Both of them indicate that the inductive loads are tolerable up-to 10 kHz, above which the current decreases drastically. I will repeat the measurements using an external source, but I feel the observed noise is due to some other factor I may be missing.

Venkata Ramana Mudinepalli , thank you for the comments. I have tried moving the sample along the length of the coil, and I obtain the maximum voltage at the center of the coil. This means that the voltage I am observing is definitely from the coupling from the ac field. But, I am not able to figure why such large noise is observed! I basically want to measure the ME coupling coefficients of my thin films, which are single phase multiferroics. However, to verify the ME measurements, I have carried out the measurements on both bulk single crystals, as well as thick films and thin films. In all cases, (even when the sample is not multiferroic), I obtained the ME voltage in the range of 100 - 200 micro-volts! So my calculated ME coefficient would then be solely dependent on the thickness of my sample. For thin films, the calculated ME coefficient thus comes to be about 10 V/cm.Oe, which I feel is too high! There is obviously some large noise present in all of my measurements.

Farnaz Safi Samghabadi , thanks for your inputs. Even I am starting to feel the noise is arising from improper shielding or grounding. I used BNC cables to connect the samples, but maybe this shielding is insufficient? Also, I used the grounding available within the lock-in amplifier. Could you elaborate on how you overcame the shielding and grounding related issues? Thanks.

Hi Badari,

Actually, right now every element in our setup is grounded, from ac and dc coils to even sample. Since, all of them can induce a magnetic field and an EM wave in the presence of the other and affect the measurement. Regarding shielding, our dc coils and sample are shielded. At first, we were using a Faraday shielding cloth which we found out it was magnetic itself! So, we simply use aluminum foil now.

Hope it helps!

Dear Rao,

Definitely, the ME voltage is depends on the thickness of either bulk materials or thin films, also coupling between the compounds used in the composite. Have you taken the cross-sectional TEM of your thin films. What you get the 10V is wrong I think. Could you please see ME setup in my research papers.

Venkata Ramana Mudinepalli Could you please name or share the paper having ME setup explanation? I looked for it in some of your articles, but could not find it.

Thanks!

Dear Safi,

Please see the attachment file.

Hello,

we had the same problem with ME measurements from Terfenol-D/PZT and PZT thin films until we saw that a commercial capacitor and a single loop put into a coil gave exactly the same results. We then realized that the presence of an ac magnetic field can generate parasitic voltages (EMF, eddy currents...) with values in the order or higher than the ME voltage you want to measure. We developed an electrical model to get the real ME voltage.

Here are the 2 articles we published from this work.

Christophe Cibert This is exactly the kind of answer I was looking for! Thank you very much for explaining the problem. I will try the technique mentioned in your article and verify.

You might want to check if you have a ground/earth loop.

What I always have is a BNC cable with the shield mesh disconnected from one of the connectors. Cut a gap of ~3cm just leaving the core and insulator exposed. That way you can use it to try to stamp out any ground loops on data cables. It sometimes works.

A ground loop can dump all sorts of rubbish into your signal if not careful...

https://en.wikipedia.org/wiki/Ground_loop_(electricity)

Badari Narayana Rao