Lock-in Amplifier Quick Start

An introduction to lock-in amplifiers for students in the undergraduate Advanced Physics Lab at the University of Toronto.
0:42 Basic mathematics of what a lock-in amplifier does
1:42 Hands-on Exercise 1 looking at lock-in capabilities and features.
6:01 Hands-on Exercise 2 for testing and calibrating the lock-in.
If you find the audio hard to understand, turn on Subtitles / Closed Captions by clicking on the icon at the bottom of the video. If the video is too fast, try changing the Speed to 0.75 or pause the video as needed. (Both Subtitles/CC and Speed controls can be accessed by clicking on the gear icon at the bottom of the video.)

[@homemadelemonai7243]

excellent video, thank you for making and sharing

[@ag1272]

This is a great video, thank you. It would be easier to understand if the pace was a bit slower, though.

[@julianandresangeljimenez9152]

Excelente explicación del lock in

[@anandpatel8779]

The input impedance of the srs830 is 1 mega ohms will it damage my 50 ohm generator?

[@scooby45u]

Great video. Just repaired an Ortec 9503 lock-in amp yesterday that I found on e-bay. Started working with it earlier today then found your video. Pretty much confirmed that I knew what I was doing never being exposed to the use of a lock-in before. Experiments in my basement lab await. I hope to see more on the use of lock-in amps from U of T in the future. Thanks

[@torontoapl6987]

+scooby45u I'm glad it was helpful. Best wishes.

[@va4409]

Hello Scooby45u: I have one small query regarding Ortec 9303E Lock-in. How do you tune the reference phase knob? It only goes from -4.5 to 5.5 degrees. I am a bit confused with that function.

[@cyrusIIIII]

Could it be used in signals other than AC or sinusoidal nature signals? For example if I have a sensor and hypothetically it generates an Analog pA current or nV voltage (DC or AC) in its output, a lock-in amplifier could measure that? What would be the range of acceptable frequency? If not just curious what measurement devices could measure that? I mean for example nanovoltmeter can also amplify and measure such small voltages, what would be the advantage of lock in amplifier to nano voltmeter? sorry for too many questions :)

[@torontoapl6987]

+cyrusIIIII Lock-ins do not work with DC signals. Our lock-ins typically work from 10Hz to 100kHz, but I have seen more expensive Lock-ins that work down to 1mHz or up to 100s of MHz. One can often choose whether the signal is AC or DC. In our undergraduate lab, we measure small low-temperature resistances by applying an AC current so the voltage can be measured with a lock-in. These resistance measurements could be done with a modern nanovoltmeter and low noise DC current source, but they cost more than the low-end lock-in and function generators we already own. What works best depends on the nature of the signal you are trying to observe, the noise and drifts it is subject to, and your available equipment and budget. You might want to look at articles such as www.evaluationengineering.com/articles/200812/nano-measurements-need-mega-care.php or www.tek.com/sites/tek.com/files/media/document/resources/Lock-In%20WP.pdf. Cheers.

[@user-kd4pr1ql1q]

thanks a lot for your video! everything is very clear and understandable! I am looking for information on working with optical detectors, tracking phase shift, I am writing through a translator, I hope it does not spoil the meaning much

[@tahhansadaf9953]

Great video. It was really helpful.I am working with SRS830 DSP lock-in amplifier. Do you have any idea about what dynamic reserve or gain in lock-in amplifier to give me an idea that how it works? Does lock-in amplifies signal?

[@torontoapl6987]

+Tahhan Sadaf Dynamic reserve parameterizes how well a lock-in can extract a signal from noise. It is normally defined as Dynamic Reserve (in dB) = 2 x 10 x log10(V_noise/V_smallest_signal) where V_smallest_signal is the smallest signal amplitude that can be accurately measured in the presence of noise with amplitude V_noise, (The factor of 2 in the definition is because dynamic range is traditionally defined in terms of power, which is typically proportional to voltage squared.) For example, a lock-in that can accurately measure a 100nV signal in 10mV of noise has a Dynamic Reserve of 100. Caution: I have seen the term “Dynamic Reserve” used to describe what I think of as Dynamic Range, which (if I follow the pattern of the above definition) should be Dynamic Range (in dB) = 2 x 10 x log10(V_max/V_min) where V_max and V_min are the amplitudes of the largest and smallest signals that can be measured by the lock-in. The name "lock-in amplifier" is a bit confusing, since it is actually an AC meter, not an amplifier. Its output is not an amplified signal, but a meter reading or DC level proportional to the amplitude of the signal input. When you change the Sensitivity scale on an analog lock-in, you are likely changing the internal gain of the input amplifier, but a digital lock-in may not need to change any internal gain if its analog-to-digital converter (ADC) has sufficient bits. You may have control over the DC Output Gain, which changes the proportionality of the DC output level to the amplitude of the input signal. I hope this helps.

[@wagnergarcia3285]

HELLO I'm with the problem of Overload ! You know how to fix this problem ? I have reseted the lock-in , but it worked . thank you

[@torontoapl6987]

+Wagner Garcia I am not sure I understand the question, but the usual reason for an overload is that the input signal is too large for the current sensitivity scale. Depending on the lock-in, the overload lights might also flash to show it can't find a reference signal.

[@baxbaz]

Hi...I have an old Ortec 9501E Lock in Amplifier....tUnfortunately don't have a manual .....could you further clarify what the read-out on the needle gauge indicates ? Thanks

[@torontoapl6987]

To understand your particular lock-in, you might try contacting the University of Twente EEMCS Historic Instrument Collection (studieverzameling.utwente.nl, click on the British Flag for the English version). They apparently have a paper manual for the 9501E (studieverzameling.utwente.nl/display_item.php?table=manual&id=386). More generally, I think of the meter as essentially showing a running average of the input amplitude times the cosine of the phase difference between the input and the reference. The time-constant essentially determines the time period over which the “average” is made. To find the actual input amplitude, you need to adjust the phase knob until the meter reaches its maximum amplitude, at which point you can read the amplitude off the meter (scaled by the sensitivity setting), and read the phase difference from the knob position. If the amplitude exceeds the +/-1 range, you need to adjust the sensitivity to bring the meter reading back in range. If your meter goes from -1 to 1 and there are no +/- lights, then if input and reference are more than more than 90 degrees out of phase it will show as negative on the meter. If you have a 180 degree button, pressing it should flip the meter reading from - to + or vice versa. As you can tell from the video, having a couple of frequency generators and an oscilloscope makes it much easier to figure out what a lock-in is doing. I hope this helps a bit.

[@baxbaz]

Thanks for the answer! I have the output going into an oscilloscope to graphically see what is going on .... My needle just sticks around the center value and seems to only go to maximum when the sensitivity max's out

[@va4409]

@@torontoapl6987 Thank you so much for the nice two-part experimentation explanation. This helps a lot to understand the functionality of the instrument and the initial basic setup. I too have ORTEC Brookdeal 9501E lock-in amplifier and not able to understand the output in magnitude(R) and phase (theta) terms. The needle gauge in the lock-in display shows the DC signal but changing the value (due to the phase). I have two questions: 1. How to measure the magnitude of the fluctuating DC due to phase? 2. I am trying to change the phase knob but can not make the fluctuating needle stable at one value. Is there any other way to do that?

[@torontoapl6987]

@@va4409 If the needle is jumping around quasi-randomly, it usually means there is no lock. Have you checked that your reference signal is large enough? If the needle is oscillating smoothly, that might be beats if the input frequency is slightly different from the reference frequency. I am not familiar with the Ortec (or EG&G) 9501E. Looking at online photos I don't see an Auto button so I assume it cannot automatically set the phase to maximize the reading, but the reading should not be jumping around. If, using an oscilloscope as in the video, you are sure you have a nice reference and in-phase test input signal, then there may be something wrong with the lock-in or its settings. I notice the 9501E has several knobs and switches on its back. Have you played around with the switches to the left of the Reference input? i.e. the "External Reference" and Ground switches.

[@va4409]

@@torontoapl6987 Thank you so much again for the resources like Uni Twente link and advice on the instrument usage. I was using two different sources (one for ref in and another for signal in). There is a slight difference in frequency hence giving me the fluctuating needle.

[@ruzmetovkhujabek]

Thank you for the video but why the speed is soo fast

[@va4409]

Hello, I have an old Ortec 9501E Lock-in Amplifier and I am getting the variation (noise) in the output signal in the range of 50 mV but I was expecting it to be in microVolt range. Is it normal for this instrument? Isnt this instrument should be used to remove the noises. p.s.: I am recording the output using NI DAQ usb-6009 (kept in a faraday cage to isolate it from noise as well). Please help me, I am trying to see the signal in the 50 mV range as well.

[@torontoapl6987]

No, 50 mV output noise is not what I'd expect for a lock-in, but your instrument is old. I would expect a few mV, maybe 10mV max. If the output noise with no input or shorted input is the same on all sensitivity scales and ground setting , that does suggest there is a problem with the lock-in, not the input signal. I believe the maximum output voltage on the 9501E is 1V, so depending on what you are doing, you might still be able to do 5-10% measurements. (Or perhaps even better if you can average the output on a scope and the noise is random over the timescale of the averaging.)

[@va4409]

@@torontoapl6987 I have tried by wrapping the instrument in the aluminium foil (Faraday cage) and earthen it. Still with the short circuit noise is minimum 35 mV. Maximum output voltage on 9501E is -10V to +10V. And if 5-10% means 100 mV to 200 mV. Averaging is not an option for me as I wanted to take time dependent measurement (Impedance flow cytometry). Now I am running short of ideas to remove any noises.

[@torontoapl6987]

@@va4409 If the maximum output voltage is ±10V, then 35mV is only 0.35% of full scale, so it may still be possible to get measurements better than 1% as long as the noise is random. Have you looked at the spectrum of the noise, it case that provides any hints about its origin?

[@va4409]

@@torontoapl6987 In the frequency spectrum, noises is less than 5Hz. Random peak at 1Hz, 2Hz and 4Hz. And I am not able to understand the source of noise. I can only think about the mechanical vibration of table.

[@torontoapl6987]

@@va4409 That is very low frequency. Since the output noise level is independent of the input, it appears to be something internal. If mechanical vibration is causing the problem, increasing the vibration - e.g. whacking the table, shaking cables - should produce a noise spike. If it is a flaky component, e.g. something thermally or charge cycling, there is not much to do unless you want to start replacing internal components. I probably can't provide much more in the way of help. If you have not already done so, you could try asking on forums such as electronics.stackexchange.com.

[@jan861]

Why does the signal voltage need to be sinusoidal??

[@torontoapl6987]

It doesn't have to be sinusoidal. The math is simpler if it is, but the lock-in output should be proportional to the Fourier component of the input signal that matches the frequency of the Reference. If either the signal or reference is not sinusoidal, you also have to worry about any filtering inside the lock-in. Take a look at kzbin.info/www/bejne/kHrNc4V3Z7hjhLM.

[@jan861]

@@torontoapl6987 Thanks a lot! Is this why choppers are being used to modulate the input signal?

[@torontoapl6987]

@@jan861 Yes, one of the most common uses of lock-ins is with optical choppers to help detect small optical signals.

[@cessromer7078]

Here's an example of how you can implement the interface connection to the SE1022D lock-in amplifier using classes and objects in VBA: ```vba ' Add a reference to the VISA COM 3.0 Type Library ' Go to Tools > References > Browse > select 'National Instruments VISA COM 3.0 Type Library' ' Define a class for the SE1022D lock-in amplifier Class SE1022DLockInAmplifier Private rm As New ResourceManager Private instr As Visa Private timeout As Integer ' Constructor Public Sub Initialize() Set instr = rm.Open("GPIB0::10::INSTR") timeout = 1000 End Sub ' Method to configure the lock-in amplifier Public Sub ConfigureLockInAmplifier() instr.Write "*RST" ' Reset the device to default settings instr.Write "FREQ 1000" ' Set the frequency to 1000 Hz instr.Write "AMPL 1" ' Set the amplitude to 1 V End Sub ' Method to query and print the current settings Public Sub PrintCurrentSettings() instr.Timeout = timeout ' Set the timeout value for the query instr.Write "FREQ?" ' Send the query command Dim freq As String freq = instr.ReadString ' Read the response instr.Write "AMPL?" ' Send the query command Dim ampl As String ampl = instr.ReadString ' Read the response ' Print the current settings Debug.Print "Frequency: " & freq & " Hz" Debug.Print "Amplitude: " & ampl & " V" End Sub ' Destructor Public Sub Terminate() instr.Close rm.Close End Sub End Class ' Main subroutine Sub Main() Dim lockInAmp As SE1022DLockInAmplifier ' Declare a variable for the SE1022D lock-in amplifier ' Create an instance of the SE1022DLockInAmplifier class Set lockInAmp = New SE1022DLockInAmplifier ' Initialize and configure the lock-in amplifier lockInAmp.Initialize lockInAmp.ConfigureLockInAmplifier ' Query and print the current settings lockInAmp.PrintCurrentSettings ' Clean up and release the lock-in amplifier object lockInAmp.Terminate Set lockInAmp = Nothing End Sub ``` In this example, we define a class called `SE1022DLockInAmplifier` to encapsulate the functionality related to the SE1022D lock-in amplifier. The class has methods to initialize and configure the lock-in amplifier, query and print the current settings, and terminate the connection. The main subroutine creates an instance of the `SE1022DLockInAmplifier` class, initializes it, configures the lock-in amplifier, queries and prints the current settings, and then cleans up the resources. Note that you'll need to replace the `"GPIB0::10::INSTR"` with the actual VISA resource string for your SE1022D lock-in amplifier. Ensure that you have the correct VISA address for your specific device. Enjoy

[@trollmaiden2112]

I though it would be a tutorial how to use a guitar amp :(

[@torontoapl6987]

Yes, there are lots of different kinds of amplifiers and in this case the name is a bit confusing. It might be better called a "lock-in voltmeter", but these devices have called "lock-in amplifiers" since at least 1941 (aip.scitation.org/doi/10.1063/1.1769919), so it is probably too late to change now. I hope you found a good guitar amp tutorial.

[@vs24bv]

@@torontoapl6987 you should record this like 10 times more slowly and not by using a tape recorder, throwing in the bathtub, filling it with acid, and then playing it into a talkboy

[@torontoapl6987]

​@@vs24bv Please see the recommendations under "Show More" if you find the sound difficult to understand.

[@rafiqulalam5348]

Please solve those slowly

[@adrianllanos8562]

you can slow the video down. OR pause it haha

[@lineakristensen1821]

I think I have paused it 100 times. Not even a professional could follow that in real time.

[@harbselectronicslab3551]

if its possible, can you please talk faster