Lecture B3 - Strain Gauges, Loughborough University

Dr Gianfranco Claudio & Dr Tim Harrison WSC353 INTERFACING FOR MECHATRONIC SYSTEMS Lecture plan Sensors (1) 31st Oct 13:00 Sensors (2) 7th Nov 13:00 Strain Gauges, Charge Amplifiers, 14th Nov 13:00 Sensor Networks & Relays Actuators (1)...

Dr Gianfranco Claudio & Dr Tim Harrison WSC353 INTERFACING FOR MECHATRONIC SYSTEMS Lecture plan Sensors (1) 31st Oct 13:00 Sensors (2) 7th Nov 13:00 Strain Gauges, Charge Amplifiers, 14th Nov 13:00 Sensor Networks & Relays Actuators (1) 21st Nov 13:00 Actuators (2) 28th Nov 13:00 Thermal and Noise considerations 5th Dec 13:00 Some practical examples 12th Dec 13:00 Part B: Practical Considerations Lecture B3a: Strain Gauges Learning Outcomes Understand what a strain gauge measures Understand how a strain gauge works Understand how to connect a strain gauge to a practical circuit  Environment Overview: Measuring Device Analogue Electronics Transduction Data Acquisition Microcontroller What is Strain? A term from mechanical engineering It is used to describe deformation of a solid: L1 This is called ‘engineering strain’ L2 ‘True strain’ is given by: 𝐿𝐿2 − 𝐿𝐿1 Δ𝐿𝐿 𝐿𝐿2 𝑑𝑑𝑑𝑑 𝐿𝐿2 𝜖𝜖 = = 𝜖𝜖𝑡𝑡 = = ln 𝐿𝐿1 𝐿𝐿 𝐿𝐿1 𝐿𝐿 𝐿𝐿1 What is Stress? You will often hear mechanical engineers talk about stress and strain together Stress (σ) is basically the force divided by the area: F 𝜎𝜎 = 𝐴𝐴 By Jorge Stolfi - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=24499456 Stress-Strain Mechanical and materials engineers often want to know what a material does under loading The way we find this out is using a stress-strain curve: We are often interested in measuring stress: Measure strain Use the linear Note the relationship (Young’s linear region modulus) to calculate stress Stress can be related to force and in turn mass https://en.wikipedia.org/wiki/Stress%E2%80%93strain_curve Back to Strain Gauges Used to measure the ‘stretch’ of a material Not just an electrical device, low tech ones also available: By The original uploader was RoySmith at English Wikipedia - Transferred from en.wikipedia to Commons., CC BY-SA 2.5, https://commons.wikimedia.org/w/index.php?curid=7198770 Strain gauge uses Load cells Platform load cell In-line load cell Pressure sensing Vibration analysis Fatigue calculations …  How do we Measure the Stretch? https://commons.wikimedia.org/w/index.php?curid=61871859 By Pleriche - Own work, CC BY-SA 4.0, By measuring a change in resistance A strain gauge is effectively just a piece of plastic with a long flexible resistor on it: Sensitivity direction Alignment markings Resistive element Connection pads Lots of different shapes for different applications CC BY-SA 2.5, https://commons.wikimedia.org/w/index.php?curid=893221 Resistance & strain Resistance of a wire is proportional to Length ρ𝐿𝐿 ρ𝐿𝐿 1 / Cross sectional area 𝑅𝑅 = = 𝐴𝐴 𝜋𝜋𝑟𝑟 2 Resistivity of material (ρ) Stretching a wire Volume remains constant r L, A (& r) change A Hence R changes A r L L Strain Gauge in Practice Wikipedia page They’re tiny! https://youtu.be/lWFiKMSB_4M?t=1m13s Measuring from an Actual Strain Gauge Change in resistance is very small How do we normally measure small (changes in) resistances? Wheatstone bridge But temperature changes the resistance!  Wheatstone Bridge Good for first approximation, but susceptible to temperature variation Often use rheostats (variable resistors) to match the resistances to the load cell Solution: Half bridge  Ultimate sensitivity This can be used to compensate for strain in other directions (i.e. the ones you’re not interested in measuring)  Half-bridge We normally use a half-bridge for interfacing to strain gauges This one is in compression F This one is in tension  Half-bridge - electronics 𝑉𝑉𝐷𝐷 = 𝑉𝑉𝐴𝐴 − 𝑉𝑉𝐵𝐵 𝑅𝑅3 𝑉𝑉𝐴𝐴 = 𝑉𝑉𝑆𝑆 R1 R2 𝑅𝑅1 + 𝑅𝑅3 VS VD ≡ VA VB 𝑅𝑅4 R3 R4 𝑉𝑉𝐵𝐵 = 𝑉𝑉𝑆𝑆 𝑅𝑅2 + 𝑅𝑅4 0V GND 𝑅𝑅3 𝑅𝑅4 𝑉𝑉𝐷𝐷 = 𝑉𝑉𝑆𝑆 − 𝑉𝑉𝑆𝑆 Deriving the bridge equation: 𝑅𝑅1 + 𝑅𝑅3 𝑅𝑅2 + 𝑅𝑅4 Wheatstone bridge can be simplified as a resistor network 𝑉𝑉𝐷𝐷 𝑅𝑅3 𝑅𝑅4 Calculate VD as a function of Vs and the = − 𝑉𝑉𝑆𝑆 𝑅𝑅1 + 𝑅𝑅3 𝑅𝑅2 + 𝑅𝑅4 various resistances  Maths For a strain gauge under load: Strain gauge data sheet 𝑅𝑅 = 𝑅𝑅0 + Δ𝑅𝑅 R2 +Vs Total resistance Change due to loading Unloaded resistance And we have a ‘gauge factor’: VD Δ𝑅𝑅 𝐺𝐺𝐹𝐹 ≡ 𝑅𝑅0 ϵ Gauge factor strain GND R4  Maths II Normal half-bridge equation: 𝑉𝑉𝐷𝐷 𝑅𝑅3 (𝑅𝑅40 − Δ𝑅𝑅) = − 0 𝑉𝑉𝑆𝑆 𝑅𝑅1 + 𝑅𝑅3 𝑅𝑅4 − Δ𝑅𝑅 + (𝑅𝑅20 + Δ𝑅𝑅) R2 +Vs Replace with gauge factor: Δ𝑅𝑅 = 𝑅𝑅0𝐺𝐺𝐹𝐹 𝜖𝜖 & 𝑅𝑅1 = 𝑅𝑅3 = 𝑅𝑅20 = 𝑅𝑅40 = 𝑅𝑅0 To get: VD 𝑉𝑉𝐷𝐷 𝑅𝑅0 (𝑅𝑅0 −𝑅𝑅0 𝐺𝐺𝐹𝐹 𝜖𝜖) = − 𝑉𝑉𝑆𝑆 𝑅𝑅0 + 𝑅𝑅0 (𝑅𝑅0 −𝑅𝑅0 𝐺𝐺𝐹𝐹 𝜖𝜖) + (𝑅𝑅0 +𝑅𝑅0 𝐺𝐺𝐹𝐹 𝜖𝜖) 𝑉𝑉𝐷𝐷 𝑅𝑅0 𝑅𝑅0 (1 − 𝐺𝐺𝐹𝐹 𝜖𝜖) = − 0 GND R4 𝑉𝑉𝑆𝑆 2𝑅𝑅0 𝑅𝑅 1 − 𝐺𝐺𝐹𝐹 𝜖𝜖 + 𝑅𝑅0 (1 + 𝐺𝐺𝐹𝐹 𝜖𝜖)  Maths III 𝑉𝑉𝐷𝐷 1 (1 − 𝐺𝐺𝐹𝐹 𝜖𝜖) = − 𝑉𝑉𝑆𝑆 2 1 − 𝐺𝐺𝐹𝐹 𝜖𝜖 + (1 + 𝐺𝐺𝐹𝐹 𝜖𝜖) R2 +Vs 𝑉𝑉𝐷𝐷 1 1 − 𝐺𝐺𝐹𝐹 𝜖𝜖 = − 𝑉𝑉𝑆𝑆 2 2 𝑉𝑉𝐷𝐷 𝐺𝐺𝐹𝐹 𝜖𝜖 VD = 𝑉𝑉𝑆𝑆 2 GND R4  Maths IV 𝑉𝑉𝐷𝐷 𝐺𝐺𝐹𝐹 𝜖𝜖 = 𝑉𝑉𝑆𝑆 2 R2 +Vs Key points: The output voltage is linear in strain It only depends on the gauge factor, assuming your bridge was balanced VD to start with Be careful with wire lengths: if they are long then they will put the bridge out of balance GND R4  Example: Two strain gauges are attached to a beam: With a supply voltage of 5V, a gauge factor of 2 (common for strain gauges), and a strain of 0.5% (0.005) F 𝑉𝑉𝐷𝐷 𝐺𝐺𝐹𝐹 𝜖𝜖 2 × 0.005 = = = 0.005 𝑉𝑉𝑆𝑆 2 2 𝑉𝑉𝐷𝐷 = 5 × 0.005 = 0.025V This is small!  What kind of considerations must we make? Practical Considerations Ensure the bridge is balanced before we begin: ’10-turn trimmers’ (rheostat) Is there resistance in the wires to the gauge? We need to compensate for this Strains are typically very small! The changes in resistance are small The changes in voltage are small The circuit needs to be robust to noise And have sufficient quantisation to measure the voltage accurately! Difficult to install: Surface prep, soldering and protecting can take a long time! How do we get the wires to-from the strain gauges?! Aside: Alternative Strain Gauge Polarised light can show strain in transparent plastics link: Mechanical strain gauges, as mentioned earlier: Additional Resources YouTube tutorial on strain gauges and load cells Wikipedia: Strain Strain gauge Stress-strain curve Gauge factor Omega sensors overview National Instruments Strain gauges at an online store TechniMeasure – a strain gauge manufacturer Part B: Practical Considerations Lecture B3b: Charge amplifiers Learning Outcomes Understand the need for charge amplifiers Understand the operating principles of charge amplifiers Derive the equations used for an ideal op-amp implementation of a charge amplifier  Environment Overview: Measuring Device Analogue Electronics Transduction Data Acquisition Microcontroller Why do we need them? Remember piezoelectric sensors? These generate a very small charge when strained But can generate large voltages [with low current] if strained enough In the order of ~10pC/G (for accelerometers) Coulombs, remember them? Unit for electric charge, 1C = 1A x 1s What is it? “Current integrator” Takes the charge on one side and integrates it into a voltage Made using an operational amplifier (op-amp) and a capacitor What is it? “Current integrator” Takes the charge on one side and integrates it into a voltage Made using an operational amplifier (op-amp) and a capacitor How does it work? If input bias current IB is zero, then iF = i1 𝑄𝑄 𝑉𝑉0(𝑡𝑡) = − 𝐶𝐶𝐹𝐹 V0 is proportional to Q, this is why is called a charge amplifier Choose a small capacitor to amplify a small charge A practical circuit This circuit works for an ideal op-amp, but: They can’t supply infinite gain If the signal is DC, then the capacitor is effectively an open circuit Input bias current is not always zero The assumptions we made about IB might not be true A practical circuit To allow current to flow when input voltage is DC To offset any bias current  Capacitive sensors Charge amplifiers are also used to Sensor measure the response of capacitive sensors (distance, level etc.) This requires an AC voltage source Still has a resistor to deal with DC Things to keep in mind Make sure you understand how much you are amplifying the signal Get the right ADC Make full use of the analogue/digital range The output is being filtered by the op-amp If you get strange results, this might be why Additional Resources Measurement and Fundamentals of Piezoelectric Instrumentation – Morris and sensors (Texas Instruments) Langari p 380: Measuring vibration with Wikipedia: accelerometers (National Charge Amplifier Instruments) Piezoelectric sensor Piezoelectric effect Charge mode vs voltage mode Videos pros and cons Integrator circuit (I) Charge mode sensor datasheet Integrator circuit (II) Voltage mode sensor datasheet Piezoelectric sensor & circuit Part B: Practical Considerations Lecture B3c: Sensor Networks Learning Outcomes Understand the different types of sensor networks Understand the difference between 4-20mA and 0-5V sensor networks Choose the best network for a given situation Understand the pitfalls of each of the sensor network technologies  Environment Overview: Measuring Device Analogue Electronics Transduction Data Acquisition Microcontroller Routing Signals Sensor Microcontroller Actuator Main signal routing methods: 1. Optical fibre 2. Radio 3. Pneumatic – old method for process control (more later) 4. Electric – two different methods (more later)  Routing Signals – Fibre optics Why might we use fibre optics? Fast, safe, no noise corruption Optical transmission w/out fibre also possible – rare  Routing Signals – Fibre optics How do they work? Refractive indices of the cable ensure internal reflection of light Can use more than one frequency of laser to do multi-channel Theoretically possible to change: intensity, frequency, phase, or polarization (intensity most common) Possible to multiplex, using different frequencies - multimode Measurement and Instrumentation – A. Morris, p225 Wiki page  Routing Signals – Radio Useful for data transfer where object is moving (i.e. high-speed rotor) Traditionally uses FM format: Possible to transmit digital data over radio signals as well Standards such as Bluetooth will work for digital signals/data Routing Signals – Radio Rotor testing Difficult to connect directly Slip rings would wear or have contact issues Radio interference should be minimised by the very short range Long S, Edney S, Reiger P, Elliott M, Knabe F, Bernhard D. Telemetry System Integrated in a Small Gas Turbine Engine. ASME. J. Eng. Gas Turbines Power. 2012;134(4):044501-044501-5. doi:10.1115/1.4004260.  Routing Signals – Pneumatic Now generally seen as a hangover from the beginnings of industrial automation (before dependable electronics) Transmit analogue signals as varying pressure (3-15psi) Simple control is possible using air operated valves  Routing Signals – Pneumatic Advantages: ‘intrinsically safe’ – no current/voltage to cause fires etc Immunity to electrical noise Disadvantages: Relies on air pressure waves  slow What happens if there is a temperature gradient along the tube? Routing Signals – Electric Sensor Microcontroller Actuator There are two ways this is done on most systems: 1. Sending the value as a low voltage signal (e.g. 0 to 5V, 0 to 24V) 2. Sending the value as a current signal (e.g. 4-20mA) Voltage Signal Sensors are powered individually They are connected to a voltage-in port on the microcontroller The microcontroller samples this voltage and uses that to determine signal value Output of sensor is proportional to measurand (ratiometric) Simple! Voltage Signal (II) What kind of problems might this system have? 1. Does 0V mean the measurand is zero (or bottom of the range) ? Offset zero? 2. What happens over long cable runs Cable resistance More susceptible to interference Different Connection Methods Power Power supply supply Sensor ADC Sensor ADC Two-wire Three-wire Mainly used for on/off switches Used for analogue signals (e.g. proximity) Signal voltage varies in the range Requires a load resistor inline between +’ve and –’ve supply Voltage Signal - Summary Advantages: Simple to set up Low power usage Disadvantages: Difficult to differentiate between zero value or a problem with the cable run Voltage drops can be significant over long cable runs Interference can cause drop in signal fidelity over long runs Current Loop Why? Earlier automation systems (pre-digital, pre-electronics) used 3-15psi pneumatic control/sensor signal Once reliable electronics came around  move to electrical signals Still some pneumatic systems around (apparently) These require current to pressure (I to P) and pressure to current (P to I) converters Getting rarer  Current Loop – What? By Dougsim - Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index. php?curid=52560219 Current Loop Advantages: Common interference sources have no effect (current not voltage) Minimal signal drop over cable length ‘Live’ zero signal In-loop indicators can be installed Disadvantages: Require current to voltage (V-I) transformer at the microcontroller (depends on type) Higher power usage Summary Cable Noise Complexity Power Range losses dissipation Flexibility Voltage Bad Bad Low Low Yes Current Good Good Med Med No Other Methods Digital over other systems: Voltage Current FM or AM modulation Pulse width modulation Mains voltage Summary ‘Safe’? Speed Interference? Pneumatic Y Low No (exc. thermal) Current N High No Voltage N High Yes Optical Y (but electronics) V high No Radio Y (but electronics) V high Yes Additional Resources Measurement and Instrumentation – A. Morris (Course text) Tutorial about current loop Twisted pair transmission YouTube video about internal refraction and optical fibre Part B: Practical Considerations Lecture B3d: Relays and Opto-Isolators Learning Outcomes Understand the need for relays and opto-isolators Understand what a relay is, and what it is used for Understand what an opto-isolator is, and what it is used for  Overview Microcontroller Analogue Conversion Filter / De- Glitcher Analogue Electronics Amplifier Actuation Actuator Gearing / Linkage Environment Background – Relays Used for ‘galvanic isolation’ between two circuits Different voltages Power surges, lighting strikes Removal of noise, interference and AC Uses physical switching They allow us to open/close high current/voltage/power circuits using a much more reasonable voltage  Theory – Relays Running a small voltage through the coil produces a magnetic field This magnetic field moves an armature Which moves a contact Which opens/closes the circuit NO vs NC Data sheet A Simpler Relay  Pole and Throw B A B Pole (P): How many circuits can we control at a time? Theoretically infinite, in practice up to ~10 A C Throw (T): SPST SPDT How many positions does the switch have? 1 or 2, rarely more B₁ B₂ A₁ B₁ A₂ B₂ E.g. SPDT: single-pole double-throw A₁ A₂ C₁ C₂ One circuit can be controlled at a time DPST DPDT There are two possible states for the pole Solid-state Relays Use semiconductor devices instead of coils Transistor, thyristor, triode, triac etc. Advantages: No moving parts Faster switching (because no mechanical parts) No bounce No sparking Disadvantages Reverse leakage current  heat needs dissipating, particularly for high power devices Not purely resistive so not linear between current/voltage Tend to fail-closed Caveats – Relays They are mechanical, so they wear out over time They have a large coil (i.e. inductor). Turning the current on/off can induce current/voltage spikes in the driving circuit Flyback/freewheel/snubber diode can help to overcome this If the voltage is too low, then they can become unreliable Contact sequence Sparking between contacts can lead to wear Magnetic effects on nearby kit Other Relay Types Magnetic latching: Pulse one way to turn on, the other way to turn off Maintains position if the power is interrupted Other Relay Types II Mercury-wetted relays (mercury switch) Low-voltage signals Not susceptible to contact wear but mercury Time-delay relays Switch only after a certain delay Use magnetic properties or physical dashpots (dampers) Contactors – basically just high-current relays Overload relay: Temperature moves bimetallic strip which turns current off Additional Resources – Relays Solid-state relays Relays and Optocouplers Background – Opto-isolators (or Optocouplers) Coupling circuits without electrically coupling them ‘Galvanically isolated’ (control and device ground can be different) Instead they are coupled using light Why might we want to do this? To prevent surges from one part of a circuit reaching another Lightning, ESD, RF, How would we have done this power supply spikes … before opto-isolation? (ca 1960s) Function – Opto-isolators Basically just an LED (near-IR) and a photo-sensitive sensor (photoresistor, photodiode, phototransistor …) The LED is driven and generates light The light is sensed by the light-sensor and then turns on the RHS of the circuit 1 3 2 4  Function – Opto-isolators Two main ways of packaging them: They transmit signals, not power! Normally unidirectional so DC not AC on the output side High frequency (no mechanical movement) Useful for breaking ground loops Good for DC or transients: ‘Home-made’ opto-isolator Data sheet By East of Borschov - Own drawing based on concepts patented back in the 1970s and (presumably) not protected now., CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=11958327 Caveats – Opto-isolators Remember: LEDs and phototransistor are current devices – require resistors for voltages They have high breakdown voltages (kV), but given enough voltage, there will be arcs! Current transfer ratio Similar to amplification factor Drawbacks – Opto-isolators Can be slow when compared to modern-day electronics They can’t deal with very large loads They can degrade over time the brightness of LEDs can decrease over time: Future for opto-isolators (& relays) Still used in industry, but There are now other devices that can outperform traditional opto-isolators: TRIACS ‘digital isolators’ Capacitive coupling … These tend to be faster (up to 150Mbit/s) and more easily manufactured than traditional opto-isolators Additional resources – Opto-isolators Wikipedia Undergrad YouTube video Great Scott video Galvanic isolation Arduino video Electronics tutorial Current transfer ratio Lecture plan Sensors (1) 31st Oct 13:00 Sensors (2) 7th Nov 13:00 Strain Gauges, Charge Amplifiers, 14th Nov 13:00 Sensor Networks & Relays Actuators (1) 21st Nov 13:00 Actuators (2) 28th Nov 13:00 Thermal and Noise considerations 5th Dec 13:00 Some practical examples 12th Dec 13:00 Exam format Q1 - short answer questions 25 marks total 5 or 6 unrelated short questions, 4 or 5 marks each 2 or 3 each from Tim & Gianfranco Questions 2 to 4 – long answer questions 25 marks for each Whole question on a related theme, several sub-parts May include calculation or proof of formula 2 from Gianfranco, 1 from Tim Short Answer questions These questions are about charge amplifiers. i. Draw a simple charge amplifier. [3 marks] ii. What type of sensors are charge amplifiers commonly used with, and why? [2 marks] Lecture 3 slide 33 Short Answer questions These questions are about charge amplifiers. i. Draw a simple charge amplifier. [3 marks] ii. What type of sensors are charge amplifiers commonly used with, and why? [2 marks] Used with sensors that generate only a small charge Piezo electric type sensors Lecture 3 slide 32 Capacitive sensors Lecture 3 slide 38 Both have very low-level outputs that must be amplified to the level that can be input into other devices Short Answer questions Give the chain of conversions from fluid level to voltage when using a potentiometer type fluid level sensor. Give the constants and conversion factors at each step. [3 marks] Fluid level (m) lifts a float attached to a potentiometer to give a change in resistance (Ω). Resistance Lecture 2 slide 18 proportional to fluid level. Fixed voltage is applied across the potentiometer. Lecture 1 slide 24 Output voltage (V) is proportional to resistance (Ω) Output voltage is therefore proportional to fluid level