I made a scan of the bulk reference sample for many positions of the detection unit as I did in this post for other samples.
The position closest to the sample is 34mm, which is the usual position. Other positions are gradually further from the sample.
Modulation reference is from the chopper at 430Hz. Pump power is 30mW before the sample.
Plot1 shows the scan of the sample for different positions of the detection unit. AC signal
Plot2 shows the scan of the sample for different positions of the detection unit. AC signal / DC
For each scan, I took the point at which the calibration value is taken (3rd mm of the scan) and I took 10 minutes of data to check how the noise looks like with a large signal.
Plot3 shows the calibration signal and noise (Y/DC vs X/DC) for each detection unit position.
Plot4 is a zoom of Plot3
The following table is a summary of the values.
| position | AC | DC | Phase | R |
| (mm) | (V) | (V) | (°) | (W-1) |
| 34 | 0.12368±0.00087 | 4.87±0.02 | 115.70±0.09 | 0.728±0.006 |
| 30 | 0.1124±0.0015 | 4.24±0.02 | 113.35±0.11 | 0.76±0.01 |
| 25 | 0.095±0.001 | 3.78±0.02 | 112.21±0.12 | 0.725±0.009 |
| 20 | 0.089±0.001 | 3.31±0.02 | 109.76±0.12 | 0.77±0.01 |
| 15 | 0.0983±0.0008 | 3.56±0.02 | 109.1±0.1 | 0.792±0.008 |
| 10 | 0.0538±0.0006 | 2.05±0.02 | 109.01±0.12 | 0.75±0.01 |
Error values are the size of the clouds of points on the XY plane (standard deviations)
I notice that
- the calibration factor (R=AC/DC/abs/Power) doesn't change more than 10% when moving the detection unit and doesn't show a clear trend
- the noise on the XY plane is more on the AC value rather than on the Phase value (in other words the cloud is squeezed )
- both the AC and DC get smaller when putting the detection unit further.
- the measurement at 10mm is not reliable because the reflection of the probe on the prism after the sample was on the boundary of the prism.
