vfd speed controlled by a thermistor

So does it not make an issue if the signal from the thermistor is not linear, when the RTD would be linear?
Neither are linear, but an RTD is more linear than a thermistor. It also depends a LOT on the thermistor you chose, some have a VERY steep curve and are essentially useless as analog signal sources. Plus when you try to USE the thermistor as a potentiometer, the signal current flowing THROUGH the thermistor heats it, affecting the resistance curve and making the non-linearity worse. That's why a converter is used, it normalizes the non-linearity to be a linear analog value for the VFD input.


Here's an example of a particularly problematic PTC thermistor for your application, the type usually used for thermal PROTECTION of a device, looking for a threshold being crossed, not a linear analog value.
PTC%20Thermistor%20RT%20Curve.gif


You can get ones that are not this bad, but if you don't have control of that, i.e. your customer is dictating it for you, you can get in trouble really fast.
 
Neither are linear, but an RTD is more linear than a thermistor. It also depends a LOT on the thermistor you chose, some have a VERY steep curve and are essentially useless as analog signal sources. Plus when you try to USE the thermistor as a potentiometer, the signal current flowing THROUGH the thermistor heats it, affecting the resistance curve and making the non-linearity worse. That's why a converter is used, it normalizes the non-linearity to be a linear analog value for the VFD input.


Here's an example of a particularly problematic PTC thermistor for your application, the type usually used for thermal PROTECTION of a device, looking for a threshold being crossed, not a linear analog value.
PTC%20Thermistor%20RT%20Curve.gif


You can get ones that are not this bad, but if you don't have control of that, i.e. your customer is dictating it for you, you can get in trouble really fast.

So through your post and other reading, I have learned that thermistors are far from a standard. To really know how to make this work right, we need to have all the details specs on the exact thermistor.

The customer is trying to replace/upgrade an existing system and he has this thermistor already.

Copeland p/n: 985-0318-00, but I have found no information about that online. I know it is a 10K, but that is it.

So if we can figure no more information about this thermistor, our better bet would be to swap it out with something that we know what it is?
 
What are you trying to do?

I am guessing that is a starndard 10K Type II or Type III thermistor that would be used in refrigeration/etc system. The curve for this is very non-linear and you can't just use analog input or RTD transmitter on it.


Are you simply trying to get a signal for temperature into the VFD which does not have a thermistor input?

How about this:
I would very much guess that the VFD has a 4-20MA input.

So, get a small PLC system to read the Thermistor and output 4-20MA scaled to temperature.

You can use Automation Direct Productivity P1K for this.
CPU: $171
https://www.automationdirect.com/ad...ctivity1000_(stackable_micro_plc)/cpus/p1-540

Thermistor Input Module: $121
https://www.automationdirect.com/ad...0_(stackable_micro_plc)/analog_i-z-o/p1-04ntc

Analog Output 4-20MA: $96
https://www.automationdirect.com/ad...(stackable_micro_plc)/analog_i-z-o/p1-04dal-1

If you need power supplies add that on.

The P1K thermistor input can work with different types of thermistors. The programming would be 1 or 2 lines only.
 
What are you trying to do?

I am guessing that is a starndard 10K Type II or Type III thermistor that would be used in refrigeration/etc system. The curve for this is very non-linear and you can't just use analog input or RTD transmitter on it.

I don't know about the Type, but from my knowledge of the project, I think your guess could well be accurate.

I had thought of a plc, and if that is the simplest way, I can do that without a problem, but I appreciate all the feedback on the best way to do this.

Thanks again!
 
So does it not make an issue if the signal from the thermistor is not linear, when the RTD would be linear?
A thermistor is NOT an RTD. An RTD implies a specific element and nearly linear performance.


RTD's are widely used for process temperature measurement, as eveidenced by the wide spread availability of RTD input

temperature transmitters (4-20mA or 1-5Vdc output)


Thermistors are seldom used for process temperature measurement, hence the relative rarity of a thermistor input transmitter.


I'd suggest not using a thermistor, but using a Pt100 RTD (widely available, very common) with an RTD temperature transmitter input, ranged for the mA or Voltage input range needed at the drive input.


All this is assuming that the drive handles a temperature input as a PV to PID loop. A temperature input will be a demand signal to the drive.
 
A thermistor is NOT an RTD. An RTD implies a specific element and nearly linear performance.


RTD's are widely used for process temperature measurement, as eveidenced by the wide spread availability of RTD input

temperature transmitters (4-20mA or 1-5Vdc output)


Thermistors are seldom used for process temperature measurement, hence the relative rarity of a thermistor input transmitter.


I'd suggest not using a thermistor, but using a Pt100 RTD (widely available, very common) with an RTD temperature transmitter input, ranged for the mA or Voltage input range needed at the drive input.


All this is assuming that the drive handles a temperature input as a PV to PID loop. A temperature input will be a demand signal to the drive.
I wholeheartedly agree. If you have the opportunity to suggest a redesign, I would do it like this.
 
Correction: A temperature input will NOT be a demand (speed setpoint) signal to the drive.


Well it could be. Depending on the application, if the drive is to be controlled in a simple proportional band type operation, the 4-20MA from a temperature transmitter could be used to directly control vfd speed.

Example:
Simple ventilation fan control.

4-20MA = 50-100F. Perhaps you want the fan to run slowly at 50f and max speed at 100F.

That would be a pretty simple control loop/system.
 
I think it really depends on the application.

I agree with

Thermistors are seldom used for process temperature measurement, hence the relative rarity of a thermistor input transmitter.

, but with emphasis on "process".

So if it is some kind of process control, just tell them to buy a RTD with build in 4-20mA trnsmiter..

But they (NTCs) are widely used in building automation, or refrigeration etc.. So they aren't non-professional-toys or something..

So if it is just a non critical temperature control, I'd do it..

NTCs have a non-linear characteristic, but the standard ones that are used a for measurement (not the protection PTCs jraef is talking about) can be linearised easily with a combination of resistors, one in parallel and one in series.

Here is the calculation table in attachment you can use to see what can you get this way.

So with 15k both in series and parallel and your 10k type2 NTC, you'd get a linear measurement with 5°C=9,25V => 40°C=5V.
 
Neither are linear, but an RTD is more linear than a thermistor. It also depends a LOT on the thermistor you chose, some have a VERY steep curve and are essentially useless as analog signal sources. Plus when you try to USE the thermistor as a potentiometer, the signal current flowing THROUGH the thermistor heats it, affecting the resistance curve and making the non-linearity worse. That's why a converter is used, it normalizes the non-linearity to be a linear analog value for the VFD input.


Here's an example of a particularly problematic PTC thermistor for your application, the type usually used for thermal PROTECTION of a device, looking for a threshold being crossed, not a linear analog value.
PTC%20Thermistor%20RT%20Curve.gif


You can get ones that are not this bad, but if you don't have control of that, i.e. your customer is dictating it for you, you can get in trouble really fast.

As mentioned previously, PTC's are generally poor and not used for temperature measurement precisely because of the curve you displayed.

NTC's despite their negative exponential curve are a bit more liner (if you can call them that).

The other bit of information we're also missing is the realistic temperature range this thermistor is bound to be subject to... if it's a small range, chances are that the thermistor will be linear enough for the application.
 
Well it could be. Depending on the application, if the drive is to be controlled in a simple proportional band type operation, the 4-20MA from a temperature transmitter could be used to directly control vfd speed.

Example:
Simple ventilation fan control.

4-20MA = 50-100F. Perhaps you want the fan to run slowly at 50f and max speed at 100F.

That would be a pretty simple control loop/system.

This is how we plan to use the drive. The higher the temperature the faster the motor will run. I plan to use the analog signal to directly control the speed of the motor.

My customer has finally gotten back to me with more information on the application. The motor is a fan in a cooling tower. The thermistor will monitor outdoor temperature so basically the range would be 0 -100.

He also given me permission to replace it with an RTD.
 

Similar Topics

Working on 2 VFDs for two motors, they follow any speed command when local and in hand. Once I try to give a Speed reference from my PLC they both...
Replies
10
Views
2,119
Hello, So I was asked to turn up the speed on a VFD to allow a carriage to move a little bit faster and avoid getting stuck. Im not sure the exact...
Replies
2
Views
712
I'm trying to figure out how to read the actual speed reference value from the move blocks, writing to a DB that is being sent to the VFD. For...
Replies
5
Views
2,613
I don't understand how vfd detects speed when we are running in open loop control (sensorless). For me only possible way; during auto tuning it...
Replies
4
Views
2,619
Hey PLC.net, I'm working with a PowerFlex 525. It's communicating within the panel via ethernet but I need to send the speed out 4-20mA via the...
Replies
2
Views
1,857
Back
Top Bottom