Data Acquisition. NHMFL Summer School 2015 Scott Hannahs

Transcription

1 Data Acquisition NHMFL Summer School 2015 Scott Hannahs

2 Keithley Low Level Measurements Handbook LowLevMsHandbk.pdf Stanford Research About Lock- In Amplifiers

3 Generic Data Acquisition Experiment Analysis Publication Physical Measurement Digital Representation Raw Data Files Analysis Processed Data & Graphs Presentation

4 Digitizing Errors

5 Digitizing Errors 40 x

6 Digitizing Errors 40 x

7 More Bits = Better

8 Speed vs. Accuracy But remember 1/f noise!!!

9 Add Noise - Better Measurements!!

10 Not This Way!

11 Shielding 100 mm 10 mm 1 mm 300K 4K 300 λ (Vacuum) 100 Mm 10 Mm 1 Mm Skin Depth 100 µm 100 km 10 km 1 km Wavelength (in Vac) 10 µm 100 m 1 µm 10 m 60 Hz 720 Hz 1270 khz (AM radio) 1 m 100 nm 1 Hz 10 Hz 100 Hz 1 khz 10 khz 100 khz 1 MHz 10 MHz 100 MHz 1 GHz Frequency

12 Not This Way!

13 Sub-harmonic Sampling Hz, S/N = S S

14 Sampling Speed Nyquist Frequency

15 0.2 Average 100 pts (0.01 sec, 100 Hz) S

16 0.2 Average 100 pts (0.01 sec, 100 Hz) S Average 167 pts (0.017 sec, 60 Hz) S

17 0.2 Average 100 pts (0.01 sec, 100 Hz) S Average 167 pts (0.017 sec, 60 Hz) S Average 200 pts (0.02 sec, 50 Hz)

18 Transferring Data (or catching the right bus) Protocol MBytes/sec Mbits/sec Width (bits) Length (m) # Nodes Thunderbolt / Light Peak 2,500 20, USB 3.1 Superspeed+ 1,000 10, USB 3.0 only 500 5, Firewire 1394b 400 3, USB 2.0 only Firewire 1394a Base-T Enet USB Base-T GPIB/ (961) Serial 230kbaud RS

19 Latency vs. Speed

20 Serial (RS232 / RS422 / RS485) Slow Cheap RS == Recommended Standard Connectors Vary Wiring Varies No handshaking No data integrity No addressing (except 485) Can be Isolated

21 GPIB (General Purpose Instrument Bus) Standard Connectors - Daisy Chain! Handshaking Moderate Speed Instrument Language (488.2 / SCPI) 15 Devices Bulky Connectors and Cable Expensive

22 USB (1.0 & 2.0) Moderately Fast to Fast Cheap - Need Hubs Can be Isolated Need Driver for Each Device Not Addressable Instrument Bus No Handshaking / Data Integrity Cheap Small Cables

23 Ethernet Native Isolation High Speed / Cheap Small Cables Need Hubs Well Developed Protocols / Handshaking / Data Integrity Network Bandwidth Problems Can Have High Latency (indeterminate) Non-local (Where is my voltmeter?) Web Server in Every Voltmeter?

24 Firewire / IEEE 1394a &b High Speed (Smart Topology) Data Integrity / Handshaking Reserved Bandwidth for Instruments - Low Latency Can Daisy-chain Instruments Cheap Small Cables Isolation is Easy No Instrumentation Protocols

25 Thunderbolt / PCIe / Light Peak Daisy Chain 7 devices Data Integrity / Handshaking Designed for optical transport layer Small Cables / Expensive lots of wires (20) Isolation is Easy??? No Instrumentation Protocols

26 Programming in LabVIEW

27 Producer/Consumer

28 Producer/Consumer

29 Producer/Consumer

30 Producer/Consumer

31 Producer/Consumer

32

33

34

35

36

37

38 Plug-in Architecture

39 Multiple InstrumentTypes

40 Depth!

41 Depth!

42 Re-use code

43 Re-use code

44 All Writes and Reads use this (easy debug) In parallel for efficiency Resource locking for traffic control

45 All Writes and Reads use this (easy debug) In parallel for efficiency Resource locking for traffic control

46 All Writes and Reads use this (easy debug) In parallel for efficiency Resource locking for traffic control

47 All Writes and Reads use this (easy debug) In parallel for efficiency Resource locking for traffic control

48 All Writes and Reads use this (easy debug) In parallel for efficiency Resource locking for traffic control

49 All Writes and Reads use this (easy debug) In parallel for efficiency Resource locking for traffic control

50 Version control (snapshots) Comments Modular (lots of subroutines, don t let diagram get too complicated!) Debugging Reuse Readability Clean formatting (automatic) Self Documenting UI hide unnecessary items

51 Version control (snapshots) Comments (do as I say, not as I do) Modular (lots of subroutines, don t let diagram get too complicated!) Debugging Reuse Readability Clean formatting (automatic) Self Documenting UI hide unnecessary items

52 Version control (snapshots) Comments (do as I say, not as I do) Modular (lots of subroutines, don t let diagram get too complicated!) Debugging Reuse Readability Clean formatting (automatic) Self Documenting UI hide unnecessary items (Think Mac, Windows)

53

54 Organize Your Data for Analysis Label Files with general description Do NOT try to put entire description in the file name. Do NOT use cryptic numbers or dates Keep log of data for lab book

55 Example Notebook File Files Clark_CuPzClO4.### # Type B-supply T[K] Iph[mA] Rate Comments 001 T >base 0 Cooldown 002 t 0 step steps Current step 003 T > >5 12 hr sweep (linear sweep) min 004 bad 005 T > >5 12 hr sweep (log! sweep) 006 B -0.1->18-> field sweep 007 t > Temp stabilize 008 B -0.1-> T/min to 0.1,.1T/min to 0.5,.2 T/min to 1, then 0.4T/min 009 B 18-> same 010 t B -0.1-> T/min to 0.1,.1T/min to 0.5,.2 T/min to 1, then 0.4T/min 012 B 18-> same 021 T > >1 1 hr sweep 022 B 0.03(0) Sweep to nominal "0" field 023 T 0.03(0) > >1 1 hr sweep 024 t 0.03(0) Temp Stabilize 025 bad 026 T > >1 2 hr log sweep of Iph 027 t 0.03(0) Temp Stabilize 035 t Temp Stabilize 036 T > >1 2 hr log sweep of Iph 037 t Temp Stablize

56 Analysis Keep Notebook File, show intermediate results (easy to back out of wrong track) Keep records of original data files Write functions so all data is treated exactly the same Do NOT ever ever use Excel!

57 Cardinal Rule! Signal MUST be Apparent in Raw Data!!

58 Presentation PowerPoint makes you stupid... Slideware often reduces the analytical quality of presentations. In particular, the popular PowerPoint templates (readymade designs) usually weaken verbal and spatial reasoning, and almost always corrupt statistical analysis. Visual Display of Quantitative Information Envisioning Information Visual Explanations Beautiful Evidence Visual and Statistical Thinking Artful Sentences: Syntax as Style (Virginia Tufte)