heater control

[Taylor]

I have a compact DAQ chassis that I'd like to use to control a strip heater.  The heater I am using works off of AC.  I was thinking about controlling it using one of the analog output modules but I'm not sure what other hardware I need.  I'd like to control it to a certain amperage (ie maintain a constant heat load).  Does anyone have any suggestions as to what hardware I would need?  I'm guessing I need a relay and power supply or amplifier of some sort, I'm just not sure wha will work best with the hardware I have.

Thanks,

Taylor

[Henrik_Volkers]

There are different ways to control a heater. Each has his own advantages and drawbacks.

What power range are we talking about?  (There are regulations for higher rated heaters not to disturb other powerline users ;->

Time constant of the heater (seconds or hours?)

How much overshoot allowed? 

Temperature how stable, what range?

PRICE?

......  

Short answer (right out of the crystal ball): use a SSR (solid state relay) rated for the inrush current of the cold heater and a temperature sensor and build up a simple 2 point regulation with 1 degree hysteresis.

Easy (pricy) analog: Use the 0-10V output to control a phase cutting dimmer with AC filters if you need. 

....

Message Edited by Henrik Volkers on 10-29-2009 03:47 PM

[Taylor]

The wattage will be less than 200 Watts.  I will be monitoring temperature but if possible, I would like to just control amperage to the heater.  In other words, I'd like to use a vi to control the heater wattage and monitor the temperature response of various heat loads and durations.

Thanks.

[ping2002]

FYR: I use DC power supply from KIKUSUI for a heater application  

http://www.kikusuiamerica.com/product/index.html

PAN70-2.5A, (120V input voltage for US)

~$1000 each

then use NI 9264 output 0-10V to KIKUSUI remote control panel 2 & 3 terminal to control its output voltage, and use NI 9485 to its control panel 10 & 11 terminal to control the ON/OFF of DC ouput.  and use NI 9205 to monitor the heater voltage drop and current (via resistor in series and in parallel) to calculate the exact heater power, and feedback to NI 9264 to fine tune the DC power supply output to exactly meet the heater load specification.

[AnalogKid2DigitalMan]

A resistive heater should work off of DC too. Instead of using a power supply in constant-voltage mode, an option would be to find one that can also perform constant-current control.

Note that the resistance of the heater will change depending on the temperature of the elements, so output power will change even when supplied a constant current.

-AK2DM

Message Edited by AnalogKid2DigitalMan on 10-29-2009 03:06 PM

[Taylor]

Thanks for the info!  Both of these sound like great options.  I will do a little more research into both scenarios.

Thanks again.

[ping2002]

yes, heater resistance does have significant drift when its temeprature changes, so it is necessary to monitor the heater current and voltage drop in real time and dynamicly feedback the measurement from NI 9205 to NI 9264 to fine adjust the output voltage of your DC power supply, if you do need a very precise heater load. in my application, the heater power can be maintained within 0.05% of the specified load (0.1W-120W) when the heater temperature change from 2K-300K.

the resistors I use are very stable and precise wire-wound resistor, they are not expensive, less than $10 each.

http://www.precisionresistor.com/plv-series-4-terminal-rect-axial-shunts-c-32.html

in order to get the best accuracy from NI 9205 measurement, you need to careful choose the resistor value to have the voltage input falling in the most sensitive measurement range of NI 9205 (some compromise needed), and use higher sampling rate multiples of 50 and 60 power line frequency, such 108000HZ, then average them (such as 0.1 second sampled data) to reduce the noise.  I customize the DAQ assistant VI to have more flexibility to dynamicly change the measurement range of NI 9205 to make sure it is always in the best DAQ range,  and create and close a DAQ task each time of NI 9205 measurement to improve the accuracy, and use two 9205 to measure the current and voltage at the same time (perform two channel acquisition in one DAQ task), and element by element divide the voltage array by current array, then average the divided arrary to calculate the average resistance. this is more accurate than you average the voltage and current array first, then divide the average of voltage and current. the later method is actually a weighted averaging, which gives a lower resistance, because lower resitance value is more weighted in the later method.

in the hardware wiring, I use individually shielded twisted pairs cable to reduce noise, and ground the shield. the cable I use:

http://www.newark.com/87K2827/cable-wire-accessories/product.us0?sku=belden-8723-060100

[Taylor]

I am building a thermal test setup where I need to be able to control heater power.  After some research (and the above posts) and looking at the hardware I have currently available in my lab, here is what I've decided to do.....

I have a power supply with remote control capability via 0-5V input signal.  I am going to use an NI9219 to control that and use a single NI9263 to measure voltage and current at the heater.  I will use a VI to calculate power based on the voltage and current measurements and feedback to the NI9219 to control voltage.  Does this seem like a reasonable approach?  My main question is if I need additional circuitry to make the voltage and current measurements.  I am not looking for extremely high precision control of heater power.  I'd like to keep it within about 5% or so.

Thanks,

Taylor

[EricS.]

Hi Taylor,

I think you're backwards, in that the 9219 will be reading the voltage and current, and then the 9263 will be used as the analog output to your heater.

A five percent error should be achievable.  The 9263 will have a typical accuracy of ±12mV, which is 0.012/5V = 0.2% error.

What is the expected current output of your power supply?  I just want you to be aware of the ±25mA limit of the 9219.

The formula for calculating accuracies is

 %gain error * reading  + %offset error * range.

the %gain and %offset can be found in the detailed specs for each of the devices. (sometimes it's given in ppm instead of %, convert it to a decimal in either case for the equation)

[Taylor]

Thanks for the info. 

Yes I had those backwards.  I will be using an external power supply with a remote control function.  The 9263 will just be supplying a 0-5V control signal to the power supply which will then adjust its power output.  It is a 600W power supply with adjustable voltage and current.

Thanks,
Taylor