Dependable Data Loggers

[blueEnergy]

Hello,

Has anyone ever used the Hobo U12-006 data logger?  We are trying to record voltages and current to create a power curve for our turbines.  We have an NRG wind explorer to log wind speed and direction, and have been using it's extra channel to log an amplified voltage drop across a shunt to get current, but we'd also like to record the battery voltages.  It looks like the logger and software will cost almost $200, so we are wondering if there are better options for that much money.  

Also, it would be nice to get a logger that could record the voltage drop across the shunt (0-.045V) without needing to amplify it, since our OP AMP circuit has some fluctuations in its gain that we haven't figure out.

Thanks!

bE

[commanda]

Have a look in my diary; there's a circuit for an ammeter that may help.

Generally, the op-amp needs to be a rail-to-rail device, unless you can arrange dual supplies. Regular op-amps dislike having inputs at ground potential.

Amanda

[wdyasq]

Might look here:

http://www.dataq.com/products/startkit/di148.htm

8 Channel data logger ~$50 and as dependable as Windows....

Which is "good enough" for most folks.

Ron

[Opera House]

A hall effect current snsor from LEM or BELL might be a better choice than a shunt.  This would eliminate the potential for smoke and eliminate the amplifier build.  Unfortunantly, many of the units still require a +- supply.  I use a LEM 400 amp version and loop more turns when measuring lower currents to increase the output voltage.

[blueEnergy]

Thanks for all of the suggestions...I didn't have much luck locating an adequate product on the LEM website - do you have any more detailed info about the hall effect sensor that you are using?  Thnx

[SamoaPower]

You might consider the Allegro Hall Effect current sensor. It matches up pretty well with the Hobo. Have used Hobos and like them. For battery voltage you may want to put in a series zener before a voltage divider to get an expanded scale - better resolution.

http://www.allegromicro.com/datafile/0754-200.pdf

[Flux]

You can also consider Honeywell hall effect cts using their Lohet II hall sensors.

I agree that hall sensors are much easier to use than shunts. You need to be pretty smart with single rail op amps and differential inputs to make much sense of shunts especially when you have several channels.

I often use the Lem Heme LTS 25-NP from RS components. I assume the LTS 25-NP is the Lem reference number. Not sure if LEM is the same as LEM-HEME

[jimovonz]

I have had no trouble using the UGN3503U hall effect sensor for current monitoring. Costs a few dollars, operates on 5V and is capable of well over +/-100A. Check out this brief discription of how I use it: http://www.fieldlines.com/comments/2006/2/27/0479/09688/20#20. If you have access to a supply of picaxe micros (http:\www.picaxe.co.uk), I suggest you look at using one for your data logger. If you have a PC spare for the actual data storage, you could set up a 4 channel (3x10 bit ADC) logger for around $10. With a bit more complexity/cost you can setup a stand alone logger with its own storage. Many more channels possible using larger/multiple micros.

[blueEnergy]

I looked at the data sheet for the LTS 25 NP.  Its output signal varies around 2.5 V, and most of the loggers take an input signal from 0-2.5V.  What is the easiest way to offset the output by -2.5V?  Also, what do you use for the 5V power supply?  Thnx, bE

[blueEnergy]

Hi SamoaPower,  

Same question I asked Flux - what do you use for the power supply voltage, and how do I scale the output voltage b/w 0 and 2.5V?  Thnx

[SamoaPower]

It depends on what you want to measure. The Allegro ACS754SCB-200 can measure +- 200 amps with zero amps = Vcc/2. It's basic sensitivity with a 5 volt supply is 10 mv/amp. If you are only interested in unipolar current and directly scaled output, then a supply of +- 2.5 volts will give you an ouput of 0-2 volts for currents from 0 to +200 amps. I suggest you download the data sheet to see the possibilities.

[Flux]

If you have to use one supply you will need an op amp to increase the gain to your 5 v and you can use this to offset the output to zero.

If you can use a separate centre tapped 5v supply to the hall sensor you can connect the p/s centre tap to your 0 v of the logger. Trouble with loggers on wind power is the pain of power supplies compared with normal transformer isolated power supplies.

Flux

[chuckh]

Samoapower

Regarding the Allegro Hall Effect sensor.  I'm just a dabbler in electronics and often get confused by data sheets.  The data sheet says:

"Noise (VNOISE): The product of the linear IC amplifier gain (mV/G) and the noise floor for the Allegro Hall effect linear IC (≈1 G). The noise floor is derived from the thermal and shot noise observed in Hall elements. Dividing the noise (mV) by the sensitivity (mV/A) provides the smallest current that the device is able to resolve."

Does this mean the smallest current which can be reliably measured is 35(mV noise)/10( mV per Amp sensitivity)= 3.5 Amps smallest reliable current measurement?  Or does it mean 35/10=3.5mV which  at 10mV per Amp would be 0.350 Amp smallest reliable current measurement?

Chuck H

[Opera House]

There is an easy low tech method to cause an offset.  Just use a magnet near the coil!  Far too much effort is wasted trying to make the bits be even values and to have the zero at some precise value.  That is what math packages are for.

[blueEnergy]

Hmmm, I am a novice to all of this electronics and don't quite follow your explanation.  At first it seemed like the utility of the hall effect transducers was that they have internal amplification so that you don't need to worry about OP AMPS and dual power supplies, but now it seems like I'll have to figure out a good solution for the dual power supply to run the op amp to amplify the output from the transducer!? (right now I am using two 9V batteries as my center tapped supply voltage for the non rail-to-rail op amp and shunt, but the batteries don't last that long...)

[Flux]

If you are using dry batteries don't use the hall sensors, they all take about 20mA.

I assumed you at least had a car battery available.

Flux

[willib]

jim i found that post very interesting and innovative

[jimovonz]

UGN3503U ~ 9ma

[domwild]

Jim,

Thanks for that, very informative. Have hit Allegro for a sample but if they do not come good, then I try the UGN from Jaycar Electronics and use your trick. Glenn from the backshed is using a similar system with his Picaxe controller.

Any chance of seeing your Picaxe code?? I know this is a tall order.

Thanks.

[jimovonz]

Hey Dominic, any code in particular? I presume you are refering to reading current using the UGN3503? Reading the current once the sensor output is connected to one of the picaxe's ADC inputs is as simple as:

readadc10 inPinNo,w0

where inPinNo is the number of the pin the sensor is connected to and w0 is the register recieving the data. This will put a 10 bit number into w0. With no current flowing you should get a reading around 512 but this will vary somewhat from chip to chip and with the voltage supplied to the sensor. Current in one direction will cause the read value to increase and viceversa. The magnitude of the increase relative to the current increase will also vary from setup to setup depending on the core you use, wether you shape the ends to focus the flux, the position of the sensor, how many turns you use etc. It will however be linear. Often you don't actually require an absolute current measurement and you can leave it at that. If you do require an absolute measure (say to display a current reading) you can run a known current through the sensor and take a reading (use the debug command to display the register contents after performing the readadc10). Say a +10A current gives you a reading of 600, and your no current reading in 512. 600-512=88. 88/10=8.8 counts per Amp. Because the picaxe only supports integer math, you can scale the numbers up to achieve the accuracy you desire. To calculate absolute values you would do something like this:

readadc0 1,w0 'take raw reading

w0=w0-noCurrentValue 'subtract previously determined zero offset (constant)

w0=w0*10 'scale to achieve desired accuracy

w0=w0/scaleValue 'scaleValue is constant previously set to 10x the figure calculated as above

w0 now contains the 10x the Amps measured. You can divide by 10 to get the actual amps but be aware that the divide function only returns the quotient. You would also want to perform some over/underflow checks.

Simple!

[SamoaPower]

Chuck,

The noise spec is for no filtering of the output which allows it's rated bandwidth of 35 kHz. If we're concerned primarily with DC or low frequency current the ouput can be filtered (usually just a single capacitor) to reduce this noise to a low level and reduce the minimum detectable current at DC.

[chuckh]

Thanks Samoapower.  I think I understand better now.   What do you guess would be the minimum detectable DC current for this device with a noise filter on the output?

Chuck H.

[domwild]

Jim,

Thanks for that. Read the Allegro data sheet and saw, as you mentioned, that 0 Volts is equal to 0.5Vcc and half of 1024 is 512. The Hall Effect Sensor runs at 5VDC. You confirmed my suspicions!

Thanks for the Picaxe code.