Sensors FAQ

From dataTaker Wiki (FAQ)

Sensor Frequently Asked Questions

Contents

Sensor Notes

The following are links to sensor Notes on the dataTaker web page.

DT80 Range

Wiring & Installation issues

Errors and good wiring practice

Many sensor measurement error and reliability issues can be traced directly to the quality of the installation. Using good wiring practices and professional techniques will reduce errors, improve reliability and decrease time spent commissioning your installation. Always use proper grounding, shielding and terminations for your installation.

Wiring low voltage signals

Particular care must be taken with sensors that have low level outputs like strain and thermocouples. Low level signal are very susceptible to noise from other sources such as power cables, radio transitions etc. Do not install signal wiring in close proximity to high power cables or other sources of electro magnetic fields.

Twisted pair cables

Wiring should be good quality, low impedance, twisted pair cabling with shielding.

Shielding

Shielding should be grounded at one point only.

Enclosures & Ambient conditions

Enclosures

To ensure long term reliability, the dataTaker unit should be housed in an enclosure that is appropriate for the installation. This enclosure should protect the logger from exposure to corrosive materials or conditions that exceed the published environmental rating. Typically a dataTaker unit is rated for exposure to a temperature range of –45°C to 70°C and a humidity of 85% RH, non-condensing.

Ambient Conditions

Ambient conditions such as temperature range, humidity, solar radiation and exposure to sea water all impact on sensor selection. Please ensure your sensors are suitable for the environment you are going to use them in.

Temperature Measurement

Thermocouples

Why does my thermocouple temperature reading go down when I heat it up?

You have most likely connected the thermocouple wires around the wrong way.

Note: For Type K thermocouples the red wire should go in the negative terminal.

Why are my thermocouple readings noisy?

Thermocouples can be affected by many different noise sources. This is because the thermocouple output is typically around 0.040 milli Volts per degree C. As with any very low level signal care must be taken to ensure correct shielding from sources of noise and electrical interference.

Some things to consider are;

Temperature stability of the logger

A thermocouple outputs a voltage that is proportional to the temperature difference between the tip of the thermocouple and the end attached to the dataTaker. To calculate the temperature at the thermocouple tip we must also know the temperature of the terminals of the dataTaker so we can apply isothermal compensation. There fore it is very important to keep the dataTaker Isothermal. (Isothermal means at one temperature.)

  • Keep out of any drafts.
  • Keep away from heat sources.
  • Keep out of direct sun light.
  • Allow the dataTaker to temperature stabilize before taking measurements.

This usually means the logger is kept in an insulated box.

Keep your thermocouples as short as possible

The shorter the thermocouple the less chance of picking up noise from other sources. If you have a particularly noisy area and you need to drive long distances you should consider using a 4-20 mA signal conditioner.

Routing of thermocouples

Keep the thermocouples as far away from mains power cabling as possible. AC power produces a changing magnetic field around them that will induce voltages in the thermocouple.

Shielding of thermocouples

In an area where electrical noise could be a problem choose thermocouples that have shielding built in. The shield should be connected at one end only. Either connect the shield at the Ch terminal on the DT800, Ground on the DT50/500/600 range, DGND on the DT80 range or at the contact point being measured but never both.

Conductive surfaces

If using thermocouples that have a bare tip and the tip is in contact with a surface that is electrically conductive, then the thermocouple might be picking up stray currents that can affect the readings. For example; If measuring the temperature of a hot water pipe in an old house it was often the case that the water pipes were used as the grounding. This ground return means that there will be stray currents causing noise.

With conductive surfaces we can either electrically isolate the tip of the thermocouple from the surface with a thermocouple that has shielding or ground the surface to the AC terminal on the DT800 or Ground on the DT50/500/600 range. The DT80 range can float the ground reference so should not have noise issues as long as the point being measured is within 100 VDC of the DT80 ground.

Different metals in moist atmosphere

Because the thermocouple will most likely be of different metals to the surface being measured it is possible to have a galvanitic potential (Basically a small battery) being produced at the point of contact. If this is the case the only option is to electrically isolate the tip of the thermocouple from the surface being measured.

Keep the terminals clean

Make sure there is no oil, grease or corrosion on the dataTaker terminals. Because the low voltage output of a thermocouple can be affected by a change in terminal resistance.

Can I extend my thermocouples?

Yes thermocouples can be extended but you must use the correct type of thermocouple extension wire to suit the thermocouple type in use (example Type K). These are made of the same material as your thermocouples but of a lower quality and so are cheaper in price. Using copper wire to extend the thermocouple will add error to the readings. Thermocouple extension wire can be purchased from your thermocouple supplier.

Note: DO NOT use thermocouple extension wire as thermocouples. They will not produce a correct temperature reading.

I have variations between thermocouples purchased at different times or from different suppliers

Thermocouples have manufacturing tolerances. For example a standard grade K type thermocouple has a stated accuracy of 2.2Deg C or 0.75% which ever is greater. This means that thermocouples manufactured at different times or from different material batches can have up to 2.2 degrees difference in the reading and still be within tolerance.

To reduce this type of error it is recommended that all thermocouples used on an installation are made from the same roll of thermocouple wire. You can specify this when you order your thermocouples.

Can I connect a single thermocouple to two measuring instruments?

No, this is not recommended as you are creating additional junctions and will not have accurate readings. Refer to thermocouple theory if you would like more information on why this happens. You can purchase duo sensors having two independant thermocouples in a single probe assembly to assist when you require a common measurement point for two instruments.

Can I calibrate my thermocouples?

Yes. Please refer to Technical note TN-0024-A0 Calibration of thermocouples for details.

What thermocouple temperature standard does Datataker use?

Datatakers comply to the international standard ITS-90 for thermocouple temperature measurment.

How do I use an external temperature reference when using statistical channel options?

When measuring thermocouples using statistical channel options you must use a statistical option in the external temperature reference channel specification as well. This is required so that the external reference is sampled at the same rate as the statistical samples. If the statistical channel option is not specified for the external temperature reference then the logger will silently use the internal temperature reference. The following example shows how to correctly use an external temperature reference with statistical channels.

RS1S RA10S 10PT385(TR,AV) 1TT(AV) 2TT(AV)

Note that the actual statistical function used with the external temperature reference does not matter. Often the working channel option (W) would be applied anyway as the cold junction is not often returned or logged.

How do thermocouples work and which type should I use?

The wikipedia has a good article that answers this and other questions about thermocouples generally.

RTDs

Does my dataTaker support PT1000 RTD's?

A PT1000 uses a small Platinum resistor that is sensitive to temperature change. It has a resistance of 1000 Ohms at 0 deg C. The default settings in the dataTaker assumes 100 Ohms at 0 Deg C. This can be changed with the resistance at 0 Deg C being entered as a channel option.

Another issue to consider is the excitation current used to measure the RTD. On the DT80 and DT5/6xx range of data loggers, the excitation current used is selected using a channel option (either "I" for 250/213uA or "II" for 2.5mA). The default current source (II - 2.5mA) is good for PT100 devices, but too large for PT1000 devices. For PT1000 devices the smaller (I - 250/213uA) current source should be used.

For example to measure a PT1000 on a DT80 or DT5/6xx range of loggers the follow channel specification should be used;

1PT385(1000,I)

Where "1000" is the resistance at 0 Deg C and "I" specifies the use of the smaller excitation current.

The DT800 data logger does not require the specification of the size of the excitation current, as it automatically adjusts, so only the resistance at 0 Deg C needs to be specified.

Does my dataTaker support PT100 RTD's?

Yes it does. There are two different types of PT100 RTD's. Some have an Alpha correction factor of 0.00385 and others have an Alpha correction factor of 0.00392. The dataTaker uses the last three numbers of the Alpha factor to distinguish between the two different types

For example;

1PT385 - for a PT100 RTD with 0.00385 Alpha

or

1PT392 - for a PT100 RTD with 0.00392 Alpha

Thermistors

Non Standard Linear Thermistors (YSI 44212)

This thermistor does not obey the standard thermistor 'rules'. It is a combination of thermistors such that the resultant output is linearised, the standard thermistor needs a polynomial applied. These thermistors are 4 wire and are consequently wired as a 4 wire resistance measurement. The temperature can be calculated from the resistance using the following equations.

LaTeX: \normalsize T_{DegC}=\frac{R-13698.3}{-129.163}

LaTeX: \normalsize T_{DegF}=\frac{R-154994.5}{-71.757}

Therefore the datataker 80/800 code

1R(4W,=1CV) 2CV("Temp~DegC")=(1CV-13698.3)/(-129.163)
1R(4W,=1CV) 2CV("Temp~DegC")=(1CV-154994.5)/(-71.757)

Where the resistance is measured on channel one (Code untested)

How can I extend the temperature range of a thermistor to read lower temperatures?

Most thermistors increase in resistance as the measured temperature decreases (i.e. they have a negative temperature coefficient). They tend to be quite non linear and increase in resistance rapidly as the temperature falls into the the lower part of the sensing range. The datataker DT80 range of loggers has an upper limit of 10K Ohms for resistance measurement which can limit the usable sensing range of thermistors. To overcome this you can use a parallel resistor to scale the output of the thermistor back into the measurement range of the logger. To determine the size of the resistor required you need to determine the maximum resistance that is required to be measured. This can be determined by using the following formula.

LaTeX: \normalsize R_p=\frac{10000 \times R_{max}}{R_{max} - 10000}

Where Rmax is the maximum value of the thermistor's resistance at the lowest expected temperature. If using one of the supported YSI thermistor types then the value of Rp can be placed in the channel options for the channel directly as these types support the use of a parallel resistor.

Here is an example

1YS05(11111) ' Read YSI440005 thermistor that has a 11111 ohm parallel resistor.

If using the standard resistance type and thermistor scaling (Tn) option then you will need to use a channel variable to scale the resistance read using the following formula.

LaTeX: \normalsize R_{actual}=\frac{R_p \times R_{read}}{R_p-R_{read}}

Here is an example (for the DT80/800 range)

BEGIN
 T1=1.129241E-3,2.341077E-4,8.775468E-8 ' Thermistor A,B,C terms definition
 1CV=11111 ' Parallel resistor value to scale thermistor to measurement range of logger

 RA10S
  1R(4W,=2CV,W) ' read thermistor resistance
  1CV("Temperature",T1)=(1CV*2CV)/(1CV-2CV) ' scale back to correct resistance value then apply thermistor linearisation

END

The parallel resistor can be located either near the sensor or near the logger. The best location is next to the sensor so that the lead compensation circuitry works at its best. Often, however, it is impractical to locate the resistor at the sensor end as the sensor may already be installed or may be manufactured with a special cable such that it is difficult to include a resistor at the sensor end. If the resistor is located near the logger then you should use a 2 wire measurement and the results will include an error due to lead length, which is generally small and can often be ignored.

Use the following wiring if the resistor is located near the sensor

4W-High-Resistance.jpg

Use the following wiring if the resistor is located near the logger.

2W-High-Resistance.jpg

  • Rp is the parallel resistor.
  • Rcable is the resistance of the cable.

Note: You should not use a 4 wire configuration if the resistor is located near the logger as this will lead to large measurement errors. i.e. omit the "4W" option from the resistance measurement if the resistor is located near the logger.

There is no upper limit on the resistance that can be read by using a parallel resistance, but the more you scale the resistance the less resolution you will have for lower resistance (i.e. higher temperature) measurements.

Resistance devices

Connecting potentiometers

Potentiometer devices, such as those below, can be connected to all models of dataTakers. The wiring can be read as a dataTaker resistance channel type with the channel options for three or four wire as required.

The three wire configuration will not provide lead length compensation. Three wire connection

The four wire configuration will provide for lead length compensation. Three wire connection

Draw wire transducers

Draw wire transducers have a wire attached to a spring tensioned spool. When the wire is extended the drum turns a potentiometer. The potentiometer provides the position feed back.

These devices can be read either as a resistance type of or dataTaker half bridge.

For the DT80 series or DT500 series 3 the maximum resistance reading is 10 kOhms. When read as a half bridge the maximum resistance is 5 kOhms For DT500 Series 1 and 2 the maximum resistance is 7 kOhms for resistance readings and 3.5 kOhms for half bridge configurations.


Linear potentiometers

Linear potentiometers are often used as position sensors and consist of a wiper that is drawn across a resistance element.

These devices can be read either as a resistance type of or dataTaker half bridge.

For the DT80 series or DT500 series 3 the maximum resistance reading is 10 kOhms. When read as a half bridge the maximum resistance is 5 kOhms

For DT500 Series 1 and 2 the maximum resistance is 7 kOhms for resistance readings and 3.5 kOhms for half bridge configurations.

Current and 4-20mA Loops

Current

How do I measure a Current?

When a current flows through a resistor there is a voltage drop across the resistor. By measuring difference in voltage at either end of the resistor and knowing the value of the resistor we can calculate the current flow. LaTeX: \normalsize I=\frac{V}{R}

The dataTaker measures the voltage and if no resistance value is entered as the channel option will use 100 Ohms as default resistance. The dataTaker will display the calculated Current.

Code example:

  • 1I will read a current on channel 1 where the shunt resistor used has a value of 100 Ohms.
  • 1I(0.01) will read a current on channel 1 where the shunt resistor has a value of 0.01 Ohms.

Can I measure large currents?

Yes, If you want to measure large currents then you need to use an external shunt resistor To calculate the size of the resistor needed Select the maximum Voltage drop (This is the input range of the dataTaker) and the maximum current to be measured. Then use the formula LaTeX: \normalsize R=\frac{V}{I}

For example; If you want to measure a Current of 1000 Amps and the maximum input is 2.5 VDC then the Value of the shunt resistor is; As LaTeX: \normalsize R=\frac{V}{I} then LaTeX: \normalsize R=\frac{2.5}{1000} = 0.0025 Ohms

Why are my Current measurements drifting?

There could be a number of reasons for current measurements drifting.

Some of the more common reasons are;

  • Temperature sensitive shunt resistor.
  • Power rating of the shunt resistor causing self heating.

If you are not using a shunt resistor with very good temperature stability then self heating can cause the shunt value to change

How do I calculate the power rating of a shunt resistor?

To calculate the power rating required use the formula;

LaTeX: \normalsize P=I^2 \times R

Where

LaTeX: \normalsize P is the power in Watts

LaTeX: \normalsize I is the Current in Amps

LaTeX: \normalsize R is the shunt resistor value in Ohms

Why aren't my Current measurements accurate?

All resistors have a tolerance. If your resistor has a tolerance of +/- 5% then the best accuracy you can achieve can’t be better than the tolerance of the shunt resistor. The better the quality and accuracy of the shunt resistor the more accurate the current readings.

4 to 20 mA

The calculations and comments about current measurements apply to 4 to 20 mA measurements as well.

Do I use the internal shunt resistor or an external one?

The DT80 range is designed to be a low power device. To reduce the power consumption of 4 to 20 mA devices used in remote locations the internal shunt resistor is switched into the measurement circuit when the sensor is being read and switched out when the reading is finished.

This switching has two side effects.

  • Sensors may need to have a warm up period to ensure a stable reading before the measurement is taken.
  • If the sensor is part of a current loop that other devices are reading then the loop is only active when the DT80 is reading the channel.

We recommend using an external 100 Ohm shunt resistor when:

  • Where mains power is readily available.
  • When other devices such as PLC's and HMI's or displays need to read the same sensor.

Use the internal resistor when:

  • Saving power is important and the DT80 is the only device reading the sensor.

How do I allow warm up time for my 4 to 20 mA sensor?

To allow warm up time for 4 to 20 mA sensors us the Measurement Delay (MD) channel option. The MD channel option specifies the amount of time in milli-seconds between the shunt resistor being switched in to the measurement circuit and when the measurement is taken.

Your sensor specification should state the minimum warm up period. You should use the warm up period as the minimum MD setting.

Load Cells, Bridges and Strain Gauges

Load Cells and other Strain gauge Sensors.

Why don't you use a 10 VDC power supply?

Load cells and other sensors based on strain gauge technology output a voltage that is proportional to the input voltage, that is they are a ratiometric device. This means the actual excitation voltage level is not important as long as we know exactly what the voltage is when we calculate the output. Traditionaly the excitation voltage has been a highly stable supply (Very expensive) of a known value, usually around 5 to 10 VDC. The power supply voltage is assumed to be stable and assumed to be exactly at 10 VDC. If the power supply drifts then so will the results.

The dataTaker measures both the input voltage and the output voltage and expresses the output as a ration of the two and multiples the output by 1,000,000 so the putput is in Parts Per Milliom PPM.

LaTeX: \normalsize PPM=\frac{V_{out}\times 10^6}{V_{ex}}

The advantage is that even if the power supply drifts slightly the results will still be accurate because we know what the excitation voltage actually was.

Why use Current excitation?

The bridges can also be powered using a precise current. Current excitation is less susceptible to electrical noise and doesn’t suffer from lead length losses. With a 4-wire load cell that is Voltage powered, the longer the lead length the lower the excitation Voltage at the resistance bridge and therefore the output is less.

With current excitation the current is the same when measured anywhere in the circuit. By supplying a precise current to excite the resistance bridge and knowing the resistance of the bridge the excitation Voltage can be calculated.

How do I convert PPM to load?

The calibration sheet for the load cell supplied from the manufacturer will have a calibration factor that is often stated as so many mV per Volt at the full load of the load cell. (E.g. 2.000 mV/V @ 100 N) Stated in this format the calibration factor is actual in parts per thousand. (1mV = 0.001V) to convert the calibration factor to PPM is a simple matter of multiplying the calibration factor by 1000.

So for a calibration factor of 2.000 mV/V @ 100 N the dataTaker will output 2000 PPM @ 100N. There for 1 PPM = 0.05 N

Strain Gauges

How do I connect a 1/2 or 1/4 bridge?

There are two basic ways to connect a bridge circuit to a dataTaker data logger. You can use a traditional bridge circuit where all four of the resistive elements are external to the logger. Alternately you can connect the bridge circuit using a "simulated" bridge where only two resistive elements are connected. The simulated method is only suitable for 1/4 and 1/2 bridge circuits (ie either one or two active gauges) and is only available on DT5/6xx and DT80 ranges.

Simulated bridge (DT80 range / DT5/6xx series only)

For a 1/4 bridge, Ra can be a single strain gauge and Rc can be either a precision bridge completion resistor or a temperature compensation gauge.

For a 1/2 bridge both Ra and Rc are strain gauges that "see" opposite strain. That is one senses tensile load and the other compression. Dt80 ainput faq11.jpg

Note 1: The a temperature compensation gauge is not considered to be an active gauge for the purpose of calculating the strain because it only removes the temperature induced strain.

Note 2: The 1/2 bridge is temperature compensated.

Code example 1 (DT80)

Begin"Strain"

RA1S 'Scan every 1 second
  1BGI   'Read a 350 Ohm strain gauge dataTaker half bridge current excitation
  2BGI(120)   'Read a 120 Ohm strain gauge dataTaker half bridge current excitation
end

Traditional bridge

For a 1/4 bridge, G1 can be a single strain gauge and R3 can be either a precision bridge completion resistor or a temperature compensation gauge.

For a 1/2 bridge both G1 and R3 are strain gauges the "see" opposite strain. That is one senses tensile load and the other compression.

DT800 Single Gauge.JPG

Note 1: The a temperature compensation gauge is not considered to be an active gauge for the purpose of calculating the strain because it only removes the temperature induced strain.

Note 2: The 1/2 bridge is temperature compensated.

Code example 2 (DT80)

Begin"Bridge"

RA1S 'Scan every 1 second
  3BGI(120,4W) 'Read 1 120 Ohm full bridge - 4 wire and current excitation
  4+V(BR,2,V) 'Read the bridge excitation Voltage on channel 4
  4BGV(4W) 'Read a full bridge 4 wire Voltage excitation
  5*V(BR) 'Read the bridge excitation Voltage on channel 5
  5BGV(6W) 'Read a full bridge 4 wire Voltage excitation
end

How do I convert PPM to micro-strain?

Traditional bridge circuit

For a traditional bridge configuration (i.e. all four resistive elements external to the logger)the micro-strain can be calculated from;

LaTeX: ue=\frac{PPM \times 4}{G_f \times N_g}

Where;

LaTeX: ue = Microstrain.

LaTeX: G_f = Gauge factor.

LaTeX: N_g = Number of active gauges.

Note; A gauge is active if it is being stressed by loads applied to the structure. A temperature compensation gauge is not considered as a active gauge.

dataTaker Simulated Bridge

The dataTaker simulated bridge (i.e. only two resistive elements external to the logger) has a slightly different formula for calculating the strain.

LaTeX: ue=\frac{PPM \times 2}{G_f \times N_g}

Where;

LaTeX: ue = Microstrain.

LaTeX: G_f = = Gauge factor.

LaTeX: N_g= = Number of active gauges.

Note; A gauge is active if it is being stressed by loads applied to the structure. A temperature compensation gauge is not considered as a active gauge.

How can I check my calibration is correct?

The calibration can easily be checked by shunting a known resistor across the strain gauge. The change of strain for a known resistor can be calculated form the formula;

LaTeX: ue= \frac{-G_r \times 10^6}{(S_r+G_r) \times G_f}

Where;

LaTeX: ue = Micro Strain

LaTeX: G_r = Gauge resistance

LaTeX: S_r = Shunt resistance

LaTeX: G_f = Gauge factor

Or to calculate the shunt resistance for a known microstrain change

LaTeX: S_r = \frac{G_r \times 10^6}{|ue| \times G_f}

As a rule of thumb a shunt resistor of 1000 times the value of the resistance of the gauge will give an offset of 500 micro-strain compression.

Vibrating Wire Strain Gauges

How to determine the ideal command setting for a Vibrating Sensor

The default measurement command (eg. 1FW) is equivalent to 1FW(MD350,200), where the channel option MD350 is known as the measurement delay (in this case 350 milliseconds) and the 200 is the sample period (again in milliseconds). At times it may be necessary to change these channel options for a particular sensor/task.

The measurement timing is explained below:

  1. The pluck circuit begins to store energy and is fully charged after approx 100ms
  2. This stored energy is applied the Vibrating Wire sensor over a period of approx 200us
  3. The sensor is disconnected from the pluck circuit and connected to the measurement circuit
  4. Wait for the remainder of the MD period to expire. The starting point of the MD period is the instant the pluck circuit begins to charge
  5. Measure the Vibrating Wire sensor frequency over the period of time refered to by the channel factor

Using the default settings {ie: 1FW(MD350,200)} we see that after 100.2ms the sensor has been plucked. We then wait a further 250ms for the sensor to settle before we begin measuring the fundamental frequency over a period of 200ms.

What do we do if a sensor appears to have noisy readings (> +/- 10Hz) using the default command?

The optimum time to measure the fundamental frequency is when the signal is strong and not affected by pluck or system noise. The DT80G & DT85G cannot do this automatically, but we can do it manually.

First connect a set of headphones to the logger so that you can listen to the sound coming back from the sensor. Send the following commands:

P21=1	
P62=1
RA5S 1+fw(MD350,200)

You will hear a pulse noise then the fundamental ringing sound of the sensor. The ringing sound should be quite loud. This fundamental will decay so that you will only hear system noise. This sequence will repeat every 5 seconds.

  • If the ringing sound is quiet then the cable length is probably long. We suggest reducing MD350 to 250. Don’t reduce to less than 150 as you will be trying to measure an invalid signal.
  • If the ringing sound decays quickly then reducing the sample period from 200ms to 50 or 100ms.
  • If the ringing sound is strong, without noise and decays slowly over a second or more, then increase MD350 to MD550 and increase the sample period from 200 to 400ms.

What if I don’t have a set of headphones?

Without headphones one can run a program on each sensor that responds with readings that will help you determine the ideal settings for that each sensor. This involves running the following program:

P21=1	
P62=1
/k

BEGIN
RA1M
 1+fw(MD100,50,=1cv,"MD100")
 1+fw(MD200,50,=2cv,"MD200")
 1+fw(MD300,50,=3cv,"MD300")
 1+fw(MD400,50,=4cv,"MD400")
 1+fw(MD500,50,=5cv,"MD500")
 1+fw(MD600,50,=6cv,"MD600")
 1+fw(MD700,50,=71cv,"M7100")
 1+fw(MD800,50,=8cv,"MD800")
 1+fw(MD900,50,=9cv,"MD900")
 1+fw(MD1000,50,=10cv,"MD1000")
 1+fw(MD1100,50,=11cv,"MD1100")
 1+fw(MD1200,50,=12cv,"MD1200")
 1+fw(MD1300,50,=13cv,"MD1300")
 1+fw(MD1400,50,=14cv,"MD1400")
 1+fw(MD1500,50,=15cv,"MD1500")
 1+fw(MD1600,50,=16cv,"MD1600")
 1+fw(MD1700,50,=17cv,"MD1700")
 1+fw(MD1800,50,=18cv,"MD1800")
 1+fw(MD1900,50,=19cv,"MD1900")
 1+fw(MD2000,50,=20cv,"MD2000")
 1+fw(MD2100,50,=21cv,"MD2100")
 1+fw(MD2200,50,=22cv,"MD2200")
 1+fw(MD2300,50,=23cv,"MD2300")
 1+fw(MD2400,50,=24cv,"MD2400")
 1+fw(MD2500,50,=25cv,"MD2500")
 1+fw(MD2600,50,=26cv,"MD2600")
 1+fw(MD2700,50,=27cv,"MD2700")
 1+fw(MD2800,50,=28cv,"MD2800")
 1+fw(MD2900,50,=29cv,"MD2900")
 1+fw(MD3000,50,=30cv,"MD3000")
 1+fw(MD3100,50,=31cv,"MD3100")
 1+fw(MD3200,50,=32cv,"MD3200")
 1+fw(MD3300,50,=33cv,"MD3300")
 1+fw(MD3400,50,=34cv,"MD3400")
 1+fw(MD3500,50,=35cv,"MD3500")
 1+fw(MD3600,50,=36cv,"MD3600")
 1+fw(MD3700,50,=37cv,"MD3700")
 1+fw(MD3800,50,=38cv,"MD3800")
 1+fw(MD3900,50,=39cv,"MD3900")
 1+fw(MD4000,50,=40cv,"MD4000")
END

This program will take 40 measurements using 40 different Measurement Delays. You can then compare the results and look for the readings that are the most stable. Then choose an appropriate MD value. Choose a sample that is within the stable measurement band. Don’t choose a sample period greater than 1000ms.

Below is the results from a typical sensor

MD100 493.0 Hz
MD200 3017.2 Hz
MD300 3029.3 Hz
MD400 3029.9 Hz
MD500 3031.7 Hz
MD600 3030.9 Hz
M7100 3028.6 Hz
MD800 3031.7 Hz
MD900 3030.7 Hz
MD1000 3031.9 Hz
MD1100 3024.8 Hz
MD1200 3024.5 Hz
MD1300 3024.1 Hz
MD1400 3027.4 Hz
MD1500 3034.4 Hz
MD1600 2981.3 Hz
MD1700 3028.3 Hz
MD1800 3030.2 Hz
MD1900 3021.0 Hz
MD2000 2828.8 Hz
MD2100 2878.8 Hz
MD2200 2695.3 Hz
MD2300 2396.8 Hz
MD2400 2276.9 Hz
MD2500 2212.9 Hz
MD2600 2254.6 Hz
MD2700 2263.9 Hz
MD2800 2355.8 Hz
MD2900 2199.1 Hz
MD3000 2244.9 Hz
MD3100 2275.9 Hz
MD3200 2029.3 Hz
MD3300 2099.6 Hz
MD3400 2281.1 Hz
MD3500 2232.8 Hz
MD3600 2098.8 Hz
MD3700 2219.6 Hz
MD3800 2087.4 Hz
MD3900 2182.0 Hz
MD4000 2189.2 Hz

You can see that we have stable readings from about MD300 to MD1500. In this case we would recommend that you choose 1+FW(MD500,400) , which gives a good margin for error.

How do I read a Vibrating Wire Strain Gauge?

The dataTaker DT515 and DT615 have Vibrating Wire Strain Gauge (VWSG) support and the frequency of the VWSG is read with the FW channel type. The following program will read A single VWSG connected to channel 1 as a differential input.

BEGIN

RA1M
 1FW
END

How do I convert Frequency to Strain?

You will need to record the frequency of the VWSG when installed as this is needed to zero the reading. You also need the Gauge factor (The gauge factor relates the frequency to the strain in the wire)

Generally the formula for converting the frequency to strain is

LaTeX: ue = G_f \times (f1^2-f2^2)

where

LaTeX: ue = The strain reading.

LaTeX: G_f = Gauge factor.

LaTeX: f1^2 = The current frequency reading squared.

LaTeX: f2^2 = The zero reading squared.

Example code; If the installed zero reading frequency is 1000 Hz and a gauge factor of 5 then

BEGIN

RA1M
 1FW(F6,=1CV,W)    'read the frequency, square it and save in 1CV
 2CV=5*(1CV-1000^2)  'Calculate the strain.
END

How do I compensate the strain reading for temperature?

If the VWSG has an internal tempertature sensor then connect it to the dataTaker with the correct wiring for the sensor type. Then read the temperature and save it to a Channel Variable (CV). The gauge supplies will have the temperature correction factor for the type of gauge. You will also need the temperature at the time the gauge was installed. We can the combine the strain and temperature reading to give teh corrected strain.


LaTeX: ue = G_f \times (f1^2-f2^2) + LaTeX: T_c \times (T-T_0)

where

LaTeX: ue = The strain reading.

LaTeX: G_f = Gauge factor.

LaTeX: f1^2 = The current frequency reading squared.

LaTeX: f2^2 = The zero reading squared.

LaTeX: T_c = Temperature correction factor.

LaTeX: T = Current temperature reading.

LaTeX: T_0 = Zero reading temperature.


A typical program would be;

BEGIN

RA1M
 1FW(F6,=1CV,W)    'Read the frequency, square it and save in 1CV
 2YS04(=2CV)       'Read the temperature
 'Calculate the strain.
 2CV=5*(1CV-1000^2)+18*(2CV-20)  'Where 5 is the gauge factor, 18 is the Tc and 20 the zero temperature.
END

Transistor output sensors

Some sensors have a transistor output, which acts similarly to a switch. There are two types of transistor outputs, PNP or NPN. These sensors will also have a default state, normally open (NO) or normally closed (NC), which are synonymous to the state of a regular switch.

Transistor output sensors typically have three wires, POWER, GND and OUTPUT. The output can generally be hooked up to the digital channels on the DT80 Series data loggers, but the compatible inputs differ depending on the logger model and output type of the sensor.

PNP sensors

PNP output sensors, such as the NI8-S18-AP6X Inductive sensor from Turck should be connected to the digital channels that have an internal pull-down resistor. This is because a PNP transistor is typically used to switch the 'high' side (voltage to the load). In this case the dataTaker input is the load. Compatible inputs on DT80(G)/DT85(G) loggers include D5-D8 whereas on the DT81/DT82E this is D4 only.

NPN sensors

NPN output sensors, such as the KWHR50 Capacitive sensor from Welotec should be connected to the digital channels that have an internal pull-up resistor. This is because a PNP transistor is typically used to switch the 'low' side (GND side of the load). In this case the dataTaker input is the load. Compatible inputs on DT80(G)/DT85(G)/DT82I loggers include D1-D4 whereas on the DT81/DT82E these are D1-D3.

NPN output sensors can also be attached directly to the high-speed counter inputs (1C-4C).


Wireless sensors

Nokeval Wireless Measuring System (FTR Series)

The Nokeval system comprises of a base station and wireless transmitters. The base station has a serial Modbus connection and the transmitters are designed to be used with thermocouples. We have successfully tested the FTR970B PRO system with our DT80 Series loggers over the RS485 Modbus connection. Particular notes for this system include:

  • The system uses reverse endian data with an address offset of +1 (the manual indicates register zero, which translates to a Modbus address of 1)
  • Use INPUT registers 0-179 for channel readings, these inputs are marked as "significant word first, inside word most significant byte first"
  • The base station must first be configured for Modbus RTU using the Mekuwin software. Verify this configuration by re-connecting in Mekuwin using Modbus RTU as the connection method.
  • When configuring the device, ensure jumper J5 is not present otherwise settings will not take.
  • Set jumper J11 for RS485
  • Set jumper J7 for Terminated with 3-wire connection
  • Connect D0 on the base station to (TxZ) on the dataTaker
  • Connect D1 on the base station to (RTSY) on the dataTaker
Views
Personal tools