SET DC Curve Tracer - Initial Test Results -------------------------------------------- Initial Rev. 12-Aug-2020 Current Rev. 31-Aug-2020 The intent of these initial tests of the SET DC Curve Tracer is to get an idea of the performance that we can expect of the final system running in the SET research cryostat. Some pictures of this initial test setup and of the various components in the SET DC Curve Tracer are available in the web directory: https://web.pa.msu.edu/people/edmunds/SET_Curve_Tracer/Pictures/ For an overall view of the initial test setup please see the pictures: set_dc_curve_tracer_b112_test_1.jpg set_dc_curve_tracer_b112_test_2.jpg From left to right the various boxes are: - The 25 kHz Power Source - Two low power stable quiet power supplies to provide the SET Bias: S-D Bias and Offset from Ground Bias - A small die cast room temperature metal box to hold: the Instrumentation Amplifiers, the 1000 to 1 SET S-D Bias attenuator, the 100k Ohm current measurement shunt resistor, and a 50k Ohm resistor to stand in for the SET transistor. All of these components will be inside the outer layer of the SET research cryostat in the real system. - The box that hold the SET Bias Filter and the Instrumentation Amplifier Power Supply (below this is called Bias Fltr and IAPS for short). - The box that holds the Signal Filters for the Voltage and Current channels. - Two old HP model 3490A Digital Voltmeters (now with USB readout interfaces) and a Linux laptop computer to collect the Voltage and Current channel data. Internal and external pictures of these various boxes are available in the web pictures directory noted above. Detailed drawings of the SET DC Curve Tracer are on the web in the directory: https://web.pa.msu.edu/people/edmunds/SET_Curve_Tracer/Drawings/ Please see drawing: 24_initial_test_conceptual_view.pdf The actual wiring of the two Inst Amps in their die cast box that implements this conceptual design is shown in the drawing: 19_inst_amp_internal_wiring.pdf 2. Nominal Target Values: -------------------------- The design of this SET DC Curve Tracer is based on the following nominal target values: - Voltage drop across the SET about 1 mVolt - Current through the SET about 1 nAmp - Resistance of the SET about 50k Ohms These nominal values are needed later to estimate how will this DC Curve Tracer electronics will perform in the real SET research cryostat. 3. Values of 1000: ------------------- The nominal value of 1000 appears in two places in this system so it is useful to remind ourselves where this number appears and to look at the actual circuit values. SET Bias Attenuator The system has a nominal 1000 to 1 attenuation of the SET S-D Bias Voltage from the value that is provided by the external SET S-D Bias Power Supply to the value that appears across the series combination of the SET Transistor and the 100k Ohm current measurement shunt resistor. Note that there is no attenuation of the SET Bias Offset from Ground between the power supply that provides this voltage and the voltage actually seen by the SET. There are the following differences in this 1000 to 1 attenuator between the "Conceptual View" drawing and the actual circuit: - The Attenuator Series Resistor is not 100k Ohm but rather: 2x 1k Ohm + 2x 24.9k Ohm + 49.9k Ohm = 101.7k Ohms. The 1k Ohm and 24.9k Ohm resistors are in the Bias Filter circuit and the 49.9k Ohm resistor is in the die cast box with the Inst Amps. Why these funny values ? 25k Ohm and 50k Ohm are not standard values in the "E192" series of resistors. See Wikipedia "E series of preferred numbers". The tolerance of these resistors is 0.1% - The Attenuator Shunt resistor is not 100 Ohm but rather: 100 Ohm in parallel with the series combination of 100k and 49.9k Ohm. This results in a 99.933 Ohm shunt resistance. The tolerance of these resistors is 0.1%. The overall result is that the "gain" of the attenuator is: 0.0009817 instead of the nominal: 0.001 Inst Amp + Signal Filter Gain The nominal gain of the Inst Amps and Signal Filters working together is 1000. The actual DC gain of these circuits is: 100.50 for the Inst Amps and 10.045 for the Signal Filters. The overall result is that the DC gain of the Inst Amp plus Signal Filter combination is: 1009.57 instead of the nominal gain of: 1000. Again these funny numbers result from resistors being readily available only in the E Series. The Inst Amp gain setting resistors are a combination of 0.1% and 0.01% tolerance. The Signal Filter gain setting resistors are 1% tolerance. There are 4 of these 1% tolerance resistors controlling the DC gain of the Signal Filters so they probably dominate the uncertainty in the overall gain of the system. 4. Range and Resolution of the Voltage & Current Channels: ----------------------------------------------------------- All data collected for this note used the DVMs on their 10 Volt full scale range. On this range the Least Significant Digit of these DVMs is 0.1 mVolt. Voltage Channel Using the nominal gain of 1000 for the Inst Amp plus Signal Filter combination the full scale range of the Voltage Channel becomes 10 mVolts and the Least Significant Digit is 0.1 uVolt. These values should be compared to the 1 mV nominal voltage drop across the SET transistor. Current Channel Using the nominal gain of 1000 for the Inst Amp plus Signal Filter combination and assuming the use of a 100k Ohm current measurement shunt resistor the full scale range of the Current Channel becomes 100 nAmps and the Least Significant Digit is 1 pAmp. These values should be compared to the 1 nAmp nominal current through the SET transistor. 5. Raw Data Measurement Method: -------------------------------- Measurements of Voltage and Current were collected in the following way for each data point: - The output voltage of the SET S-D Bias supply and of the SET Bias Offset from Ground supply were set by hand. The output voltage of these supplies is probably accurate to 0.1% and is much more stable than that. - Data for the Voltage and Current channels was collected simultaneously. 100 readings were collected by the computer from each of the two HP model 3490A Digital Voltmeters. The Voltage and Current channels each had its own Digital Voltmeter. The HP DVMs read the output of the 1 Hz Signal Filters. The 100 readings were spaced 1 second apart. The 10% to 90% rise time of the 1 Hz Signal Filters is about 0.4 seconds. The intent is to insure that these 100 readings are independent samples. For each channel the average value of the 100 samples was calculated and this is called the "Average DC" value. With appropriate scaling and removal of any zero offset this Average DC value becomes the Voltage or Current data point for the SET Bias values that are currently in effect. The Average DC value is then subtracted from each sample to make 100 signed residuals. The RMS value of these residuals was calculated and is called the "AC RMS" value. This AC RMS value is used as a measurement of the "noise" in the Average DC value described in the paragraph above. Note that the AC RMS noise measurement does include any thermal drift that occurs during the 100 second time period while the DVM readings are collected. Thermal drift of the Inst Amplifiers is a potential issue as they are true DC amplifiers working down in the sub uVolt region. - To help verify that there are no technical problems with the system, the computer software collects, calculates, and displays additional information during each set of 100 DVM readings. This additional information includes: A summary warning message if there were any USB errors detected while reading the DVM - this includes an audio alert tone. A display showing the Highest and Lowest DVM reading of the 100 samples and which sample numbers these are. Optionally you can display a small histogram of the 100 samples to see the shape of the noise distribution. The length of time that it took to collect the 100 samples is displayed as another check that the equipment is working correctly. All data reported in this note is based on measurements taken with 100 DVM samples. The software allows you to collect any number of samples and process them as described above so for example you could take 3600 samples over an hour and look at the AC RMS noise to study longer term thermal drift. All data reported in this note is based on measurements taken with adjacent DVM samples spaced 1 second apart. The time between readings can be varied over a wide range but that has not been investigated. In any case samples taken to close together from the output of the 1 Hz Signal Filters will not be independent. The software also calculates and displays the "DC RMS" value, i.e. the RMS value of the 100 samples without the Average DC value subtracted from each sample. This is useful if you need to know the "power" in a noisy DC signal. 6. Initial Tests of the SET DC Curve Tracer Electronics: --------------------------------------------------------- Study of: Current and Voltage Channel Response vs SET S-D Bias --------------------------------------------------------------- Raw Data Measurements: ---------------------- The set of data shown below was all collected with the SET Bias Offset from Ground set to 0.0 Volts. The SET S-D Bias voltage was adjusted over the range of 0.00V to +10.00V. The following raw data measurements are all in Volts read from the HP Digital Voltmeters, i.e. the actual signals are about 1000 times smaller. These measurements are based on 100 samples from the HP DVMs and have been displayed with one more Least Significant Digit than the DVMs actually provide. With the nominal gain of 1000 the Least Significant Digit in these Average DC values represent 10 nVolts at the input to the Instrumentation Amplifiers. I and V Channel Measured Values as a Function of SET S-D Bias Current Channel Voltage Channel SET Bias ----------------------- ----------------------- S-D Volts Avg DC AC RMS Avg DC AC RMS --------- ---------- --------- ---------- --------- 0.00 -0.06672 0.00015 -0.00914 0.00016 0.00 -0.06691 0.00014 -0.00927 0.00016 0.00 -0.06681 0.00016 -0.00929 0.00017 0.10 -0.00109 0.00016 0.02385 0.00018 0.10 -0.00113 0.00016 0.02383 0.00015 0.20 0.06462 0.00016 0.05663 0.00018 0.20 0.06452 0.00014 0.05669 0.00015 0.50 0.26158 0.00015 0.15492 0.00019 0.50 0.26172 0.00014 0.15503 0.00014 1.00 0.59026 0.00018 0.31881 0.00013 1.00 0.59021 0.00015 0.31875 0.00017 2.00 1.24762 0.00016 0.64665 0.00018 2.00 1.24763 0.00015 0.64658 0.00017 5.00 3.21978 0.00016 1.62984 0.00019 5.00 3.21991 0.00016 1.62986 0.00017 10.00 6.50544 0.00015 3.26825 0.00018 10.00 6.50528 0.00014 3.26810 0.00017 0.00 -0.06703 0.00011 -0.00930 0.00014 0.00 -0.06713 0.00012 -0.00932 0.00016 When the system was turned ON to take the above raw data measurements the Inst Amp temperatures were: Current Channel = 10.23k Ohm, Voltage Channel = 10.29k Ohm. After the above data was collected the Inst Amp temperatures were: Current = 7.11k Ohm and Voltage = 7.02k Ohm Analysis of the I and V Channel Values as a Function of SET S-D Bias with SET Bias Offset from Ground = 0.0 Volts --------------------------------------------------------- Response to S-D Bias = 0.0 V ---------------------------- I Channel V Channel --------- --------- Average of the 3x Measurements with S-D Bias = 0.0 Volts taken Before the S-D Bias Scan: -0.06681 -0.00923 Average of the 2x Measurements with S-D Bias = 0.0 Volts taken After the S-D Bias Scan: -0.06708 -0.00931 Shift in the Zero Response During the S-D Bias Scan: 0.27 uV 0.08 uV Average of all 5x Measurements with S-D Bias = 0.0 Volts: -0.06692 -0.00926 Now for Each Value of SET S-D Bias Show: ---------------------------------------- - the Current and Voltage Channel Response with their Zero Responses Subtracted - the and Ratio of the Theoretical Response (assuming an Attenuator "gain" of 0.0009817 and a division of the Attenuator Output into bins 100k & 49.9k and an Inst Amp + Signal Filter Gain of 1009.57) to the Measured Response - the Ratio of Current Channel Response to Voltage Channel Response which should remain constant across the various values of S-D Bias - Measured Values are shown in Volts read by the HP DVMs Current Channel Voltage Channel Ratio SET Bias ------------------- ------------------- Measured S-D Volts Measured Theo/Mes Measured Theo/Mes Cur/Volt --------- -------- -------- -------- -------- -------- 0.10 0.06581 1.00467 0.03310 0.99675 1.9882 0.20 0.13149 1.00566 0.06592 1.00098 1.9947 0.50 0.32857 1.00613 0.16424 1.00440 2.0005 1.00 0.65716 1.00610 0.32804 1.00574 2.0033 2.00 1.31455 1.00593 0.65588 1.00605 2.0043 5.00 3.28677 1.00581 1.63911 1.00641 2.0052 10.00 6.57228 1.00600 3.27744 1.00665 2.0053 Plots of the values of Theory/Measurement and Current/Voltage vs SET S-D Bias are presented in the web directory: https://web.pa.msu.edu/people/edmunds/SET_Curve_Tracer/Plots/ files: current_theory_over_measured_vs_S-D_bias.pdf voltage_theory_over_measured_vs_S-D_bias.pdf current_over_voltage_vs_S-D_bias.pdf In the above table the values of the Voltage Channel appear to have up to a 1% error at low values of SET S-D Bias, but note that a +- 0.2 uVolt shift in the Zero Response for the Voltage Channel can fully account for this. Study of: Common Mode Rejection and Zero Drift Over Time --------------------------------------------------------- Raw Data Measurements: ---------------------- The set of data shown below was all collected with the SET S-D Bias set to 0.00 Volts and the SET Bias Offset from Ground set to either: -1.00 or 0.00 or +1.00 Volts. The following raw data measurements are all in Volts read from the HP Digital Voltmeters, i.e. the actual signals are about 1000 times smaller. These measurements are based on 100 samples from the HP DVMs and have been displayed with one more Least Significant Digit than the DVMs actually provide. With the nominal gain of 1000 the Least Significant Digit in these Average DC values represent 10 nVolts at the input to the Instrumentation Amplifiers. The SET S-D Bias = 0.00 Volts for all measurements shown below. SET Bias Current Channel Voltage Channel Offset ----------------------- ----------------------- from Gnd Avg DC AC RMS Avg DC AC RMS --------- ---------- --------- ---------- --------- 0.00 -0.06768 0.00023 -0.00945 0.00023 0.00 -0.06723 0.00015 -0.00937 0.00014 0.00 -0.06713 0.00014 -0.00920 0.00016 3.5 Hours Later 0.00 -0.06708 0.00015 -0.00919 0.00018 0.00 -0.06715 0.00016 -0.00924 0.00016 0.00 -0.06719 0.00014 -0.00921 0.00014 +1.00 -0.06650 0.00013 -0.00919 0.00016 +1.00 -0.06655 0.00012 -0.00923 0.00013 -1.00 -0.06787 0.00015 -0.00932 0.00015 -1.00 -0.06778 0.00018 -0.00933 0.00016 1.5 Hours Later 0.00 -0.06724 0.00011 -0.00916 0.00010 0.00 -0.06747 0.00011 -0.00909 0.00011 Analysis of the Common Mode Rejection and Zero Drift Over Time Raw Data: -------------------------------------- Initial Zero Measurement I Channel V Channel --------------------------- --------- --------- Average of the second set of 3x Zero Measurements S-D Bias =0 Offset = 0.0 V -0.06714 -0.00921 Positive Common Mode Measure I Channel V Channel ---------------------------- --------- --------- Average of the 2x Pos CM S-D Bias =0 Offset = +1.0 V -0.06653 -0.00921 Negative Common Mode Measure I Channel V Channel ---------------------------- --------- --------- Average of the 2x Neg CM S-D Bias =0 Offset = -1.0 V -0.06783 -0.00933 Final Zero Measurement I Channel V Channel --------------------------- --------- --------- Average of the Final set of 2x Zero Measurements S-D Bias =0 Offset = 0.0 V -0.06736 -0.00913 Results: -------- Shift with +1.0 V Common Mode +0.61 uVolt 0.00 uVolt Shift with -1.0 V Common Mode -0.69 uVolt -0.12 uVolt Aprox Zero Shift in 1.5 Hours: +0.22 uVolt -0.08 uVolt Study of: Long Term Stability at a Nominal 1 mVolt Voltage Operating Point: -----------------------------=======----------------- Raw Data Measurements: ---------------------- I have now placed a 5 sided steel box over the Inst Amp Die Cast Box to act as either an electrical shield and/or to provide thermal lagging. As for the tests described above, the following raw data measurements are all in Volts read from the HP Digital Voltmeters, i.e. the actual signals are about 1000 times smaller. These measurements are based on 100 samples from the HP DVMs and have been displayed with one more Least Significant Digit than the DVMs actually provide. With the nominal gain of 1000 the Least Significant Digit in these Average DC values represent 10 nVolts at the input to the Instrumentation Amplifiers. Test started 18-Aug-2020 about 14:00 Pre Power On check the Inst Amp Temperatures: Current Amp = 10.79k Volt Amp = 10.85k Ohm Power On for 2 minutes Inst Amp Power Supply Check: VCC = 9.579 V VEE = -9.361 V Current Channel Voltage Channel Condition ----------------------- ----------------------- & Time Avg DC AC RMS Avg DC AC RMS --------- ---------- --------- ---------- --------- ---> The SET S-D Bias = 0.00 Volts Bias Offset = 0.00 Volts Power On -0.06766 0.00008 -0.00945 0.00010 10 min. -0.06756 0.00009 -0.00937 0.00008 Power On Temp Check: I Amp = 7.33k V Amp = 7.24k Ohm 20 min. Power Check: VCC = 9.619 V VEE = -9.403 V Power On Temp Check: I Amp = 7.17k V Amp = 7.07k Ohm 95 min. Power Check: VCC = 9.627 V VEE = -9.412 V Power On -0.06766 0.00010 -0.00942 0.00009 95 min. -0.06764 0.00008 -0.00929 0.00008 -0.06755 0.00010 -0.00921 0.00010 ---> The SET S-D Bias = 3.20 Volts Bias Offset = +1.00 Volts Power On +2.03762 0.00009 +1.04044 0.00007 110 min. 2.03765 0.00009 1.04042 0.00009 2.03767 0.00008 1.04046 0.00010 Power On +2.03760 0.00009 +1.04051 0.00009 190 min. 2.03758 0.00009 1.04049 0.00008 2.03760 0.00009 1.04050 0.00010 Power On +2.03691 0.00009 +1.04054 0.00011 1600 min. 2.03685 0.00009 1.04056 0.00009 2.03692 0.00009 1.04054 0.00009 Power On +2.03695 0.00009 +1.04049 0.00010 2740 min. 2.03697 0.00012 1.04049 0.00008 2.03712 0.00009 1.04048 0.00009 Power On +2.03694 0.00012 +1.04062 0.00010 3085 min. 2.03705 0.00009 1.04065 0.00011 2.03695 0.00009 1.04073 0.00010 Power On +2.03727 0.00009 +1.04075 0.00007 4285 min. 2.03711 0.00012 1.04070 0.00009 2.03718 0.00010 1.04082 0.00011 Power On +2.03688 0.00009 +1.04076 0.00010 4540 min. 2.03688 0.00009 1.04069 0.00008 2.03695 0.00010 1.04072 0.00008 Power On +2.03734 0.00008 +1.04090 0.00010 5755 min. 2.03733 0.00011 1.04091 0.00009 2.03725 0.00008 1.04089 0.00010 ---> The SET S-D Bias = 0.00 Volts Bias Offset = 0.00 Volts Power On -0.06810 0.00008 -0.00891 0.00008 5765 min. -0.06816 0.00010 -0.00886 0.00010 -0.06814 0.00008 -0.00893 0.00009 Power On Temp Check: I Amp = 7.05k V Amp = 6.96k Ohm 5770 min. Power Check: VCC = 9.632 V VEE = -9.416 V End this test run about 14:00 Saturday 22-Aug-2020 Analysis of the Long Term Stability at a Nominal 1 mVolt Voltage Operating Point Raw Data: ------------------=======--------------------------- Current Measurement The average value of the 24 Current measurements over the 4 day period was: 2.03719 Volts across 100k Ohm ---> 20.3719 nAmps The lowest and highest measurements were: 2.03685 Volts across 100k Ohm ---> 20.3685 nAmps 2.03767 Volts across 100k Ohm ---> 20.3767 nAmps So the lowest measurement was below average by: 3.4 pAmps and the highest measurement was above average by: 4.8 pAmps Voltage Measurement The average value of the 24 Voltage measurements over the 4 day period was: 1.04063 Volts ---> 1.04063 mVolts The lowest and highest measurements were: 1.04042 Volts ---> 1.04042 mVolts 1.04091 Volts ---> 1.04091 mVolts So the lowest measurement was below average by: 0.21 uVolts and the highest measurement was above average by: 0.28 uVolts Results: -------- - Assuming a nominal SET Current of 1 nAmp the worst measurement over 4 days was off by about 0.5% - The typical RMS noise on a given measurement was about 1 pAmp or about 0.1% of a nominal 1 nAmp SET Current. - Assuming a nominal SET Voltage of 1 mVolt the worst measurement over 4 days was off by about 0.03 % - The typical RMS noise on a given measurement was about 0.1 uVolt or about 0.01% of a nominal 1 mV SET Voltage. - Plots of the 24 Current and Voltage measurements are available in the web directory: https://web.pa.msu.edu/people/edmunds/SET_Curve_Tracer/Plots/ files: current_stability_at_volt_operating_point.pdf voltage_stability_at_volt_operating_point.pdf Study of: Long Term Stability at a Nominal 1 nAmp Current Operating Point: -------------------=======----------------- Test started 25-Aug-2020 about 15:50 Power has been On many days. Check Inst Amp Temps and supply voltages: Current Amp = 7.11k Volt Amp = 7.02k Ohm VCC = 9.630 V VEE = -9.412 V Current Channel Voltage Channel Condition ----------------------- ----------------------- & Time Avg DC AC RMS Avg DC AC RMS --------- ---------- --------- ---------- --------- ---> The SET S-D Bias = 0.000 Volts Bias Offset = -3.00 Volts Aug 25 -0.07012 0.00011 -0.00899 0.00011 Test Time -0.07007 0.00009 -0.00891 0.00007 0 min. -0.07014 0.00010 -0.00897 0.00008 ---> The SET S-D Bias = 0.200 Volts Bias Offset = -3.00 Volts Aug 25 0.06146 0.00011 0.05693 0.00010 Test Time 0.06138 0.00010 0.05697 0.00008 10 min. 0.06140 0.00009 0.05698 0.00007 Aug 25 0.06147 0.00010 0.05686 0.00012 Test Time 0.06154 0.00005 0.05700 0.00008 160 min. 0.06152 0.00008 0.05685 0.00012 Aug 26 0.06131 0.00008 0.05680 0.00008 Test Time 0.06119 * 0.00005 0.05698 0.00006 1120 min. 0.06121 0.00010 0.05686 0.00009 Aug 26 0.06173 0.00010 0.05698 0.00011 Test Time 0.06155 0.00010 0.05704 0.00010 1540 min. 0.06162 0.00008 0.05695 0.00012 Aug 27 0.06192 0.00009 0.05707 * 0.00008 Test Time 0.06196 * 0.00008 0.05706 0.00010 2590 min. 0.06191 0.00010 0.05702 0.00009 Aug 27 0.06160 0.00008 0.05678 0.00008 Test Time 0.06163 0.00009 0.05677 0.00008 2950 min. 0.06172 0.00009 0.05662 0.00007 Aug 28 0.06193 0.00008 0.05685 0.00010 Test Time 0.06196 * 0.00010 0.05693 0.00009 4030 min. 0.06193 0.00008 0.05681 0.00008 Aug 28 0.06188 0.00010 0.05675 0.00010 Test Time 0.06187 0.00010 0.05671 0.00010 4420 min. 0.06178 0.00010 0.05666 0.00009 Aug 29 0.06151 0.00009 0.05687 0.00014 Test Time 0.06158 0.00010 0.05678 0.00010 5620 min. 0.06158 0.00009 0.05675 0.00014 Aug 29 0.06154 0.00009 0.05651 * 0.00012 Test Time 0.06161 0.00011 0.05659 0.00012 5830 min. 0.06157 0.00010 0.05667 0.00012 ---> The SET S-D Bias = 0.200 Volts Bias Offset = 0.00 Volts Aug 29 0.06389 0.00009 0.05695 0.00011 Test Time 0.06391 0.00008 0.05696 0.00010 5840 min. 0.06390 0.00009 0.05699 0.00010 ---> The SET S-D Bias = 0.00 Volts Bias Offset = 0.00 Volts Aug 29 -0.06771 0.00011 -0.00895 0.00011 Test Time -0.06774 0.00009 -0.00902 0.00009 5850 min. -0.06764 0.00009 -0.00905 0.00012 End of run check of the Inst Amp Temps and supply voltages: Current Amp = 7.09k Volt Amp = 6.99k Ohm VCC = 9.631 V VEE = -9.414 V End this test run about 17:20 Saturday 29-Aug-2020 Analysis of the Long Term Stability at a Nominal 1 nAmp Current Operating Point Raw Data: ---------------=======--------------------------- Current Measurement Including the zero offset based on the 3 measurements taken at the beginning of the run, the average value of the 30 Current measurements over the 4 day period was: 0.13174 Volts across 100k Ohm ---> 1.3174 nAmps The lowest and highest measurements were: 0.13130 Volts across 100k Ohm ---> 1.3130 nAmps 0.13207 Volts across 100k Ohm ---> 1.3207 nAmps So the lowest measurement was below average by: 4.4 pAmps and the highest measurement was above average by: 3.3 pAmps Voltage Measurement Including the zero offset based on the 3 measurements taken at the beginning of the run, the average value of the 30 Voltage measurements over the 4 day period was: 0.06581 Volts ---> 0.06581 mVolts The lowest and highest measurements were: 0.06547 Volts ---> 0.06547 mVolts 0.06603 Volts ---> 0.06603 mVolts So the lowest measurement was below average by: 0.34 uVolts and the highest measurement was above average by: 0.22 uVolts Results: -------- - Assuming a nominal SET Current of 1 nAmp the worst measurement over 4 days was off by about 0.44% - The typical RMS noise on a given measurement was about 1 pAmp or about 0.1% of a nominal 1 nAmp SET Current. - Assuming a nominal SET Voltage of 1 mVolt the worst measurement over 4 days was off by about 0.03 % - The typical RMS noise on a given measurement was about 0.1 uVolt or about 0.01% of a nominal 1 mV SET Voltage. - Plots of the 30 Current and Voltage measurements are available in the web directory: https://web.pa.msu.edu/people/edmunds/SET_Curve_Tracer/Plots/ files: current_stability_at_current_operating_point.pdf voltage_stability_at_current_operating_point.pdf 7. Executive Summary: ---------------------- Offset Voltage of the Inst Amps: At 25 C the Offset Voltage of the OpAmps that make up an Inst Amps should typically be within +- 75 uV and have a maximum of 150 uV. While actually operating at about 35 C both Inst Amps measure within the typical 75 uV specification (and this is for the combination of 3 OpAmps that make up one Inst Amp). Offset Voltage Drift with Temperature of the Inst Amps: This has not yet been accurately measured but in the Rm B112 lab environment this does not appear to be an important source of error. Bias Current at the Input of the Inst Amps: This has not yet been accurately measured but should typically be under 0.3% of the nominal 1 nAmp SET current if we can operate the Inst Amps at 25 C or less. Common Mode Rejection of the Inst Amps: Recall that Common Mode Rejection of the Inst Amp measures how much the output of the Inst Amp changes as the SET Bias Offset from Ground Voltage is changed. The Common Mode Rejection of both Inst Amps with plus or minus 1 Volt of SET Bias Offset from Ground is better than 120 dB (one million to one). But recall that the SET Bias Offset in the real system is about one thousand times bigger than the SET S-D Voltage signal so that in the real measurement of SET S-D Voltage you will see a rejection of a little more than one thousand to one with respect to SET Bias Offset from Ground Voltage. Noise of the Inst Amps: In the measurements described above, the AC RMS noise level at the input to the Inst Amps is about 0.1 uVolt or a little less. This is with an input resistance of 50k or 100k Ohm. This value fits with the noise specification of the OpAmps that are used to make the Inst Amps. Expected drift, noise and resolution vs nominal signal levels: Assuming a nominal SET Current of 1 nAmp the worst drift (including thermal electric drift) over a period of days was measured at about 0.5% The typical RMS noise on a given measurement was about 1 pAmp or about 0.1% of a nominal 1 nAmp SET Current. Assuming a nominal SET Voltage of 1 mVolt the worst drift (including thermal electric drift) over a period of days was measured at about 0.03 % The typical RMS noise on a given measurement was about 0.1 uVolt or about 0.01% of a nominal 1 mV SET Voltage. In general this system involves voltages that are low enough and temperature differences that are big enough so that thermal EMFs could be a problem. 8. Issues and Steps Going Forward: ----------------------------------- Steps that must be done or could be done going forward: - We should discuss the ESD concerns and precautions that are needed with the Inst Amps in this system. - Make the KF Flange Mounting Bracket see drawings: https://web.pa.msu.edu/people/edmunds/SET_Curve_Tracer/Drawings/ Drawings: 22 A & B and 16 A & B - Purchase the Fischer Hermetic Connector for the KF Flange - Make the Additional required cables: Sig Fltr & IAPS Box to cryostat Fischer connector cable Long run SET Bias cables to your power supplies cable - Make the brackets/plates to mount the "boxes" up on top of the Rm B116 SET research cryostat - Run fancier tests in Rm B112, e.g. mounted on an LN2 test cryostat - Run tests in Rm B116: Practice using this SET DC Curve Tracer electronics with your Rm B114 DAQ system and SET Bias power supplies. Make a strictly "engineering test run" in your Rm B116 research cryostat just to check Inst Amp operating temperature and noise level. - Other ideas about what to do next to move forware ?