I am trying to build a MOKE system but due to some problems I can't afford lock In amplifier and PEM modulator so is there any alternative of these as my materials are both ferromagnetic and multiferroics?
This depends on the rotation of the linear polarization axis - we have measured Py nanostructure, Fe and Co layers etc. without lock-in and got a pretty good signal. Only for the nanostructures, focussing needs more care than in a setup with lock-in amplifier. The important point is that you measure with a diode bridge, not via simple intensity measurements.
If you don't know this principle, please have a look into the attached file ... and if the file is lost during uploading, what happens sometimes on this computer, please check for the APL 89,202512 in my "contributions".
If it is just a thin film sample with sufficient magnetisation, all you would need are a good photodiode, a stable laser and 2 polarisers for the optics. You would also need an ammeter or a voltmeter (along with a I-V converter) and ofcourse, interfacing with a computer.
One cheap option is to use a semiconductor diode laser with a modulated power supply, this eliminates the need for the PEM. Then, if you have an ADC converter with a sampling rate sufficiently high (compared to the modulation frequency of the laser) you can acquire the raw output data from the photodiode and calculate the FFT of your signal, retaining only the peak corresponding to your modulation frequency.
As mentioned above, you mainly need a good photodiode, a stable laser and 2 polarizers. The polarizers should have a very good extinction ratio of 1:10^6. Then you cross the polarizer and analyzer (90deg) and afterwards turn the analyzer slightly to get the best SNR.