Ways to enhance PTC heater accuracy ?

Go To Last Post
14 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I want to use PTC components of size 0805 as mini-heaters in an environment which has a temperature changing from 25 to 90°C.

I want to keep the PTC temperature as stable as possible, goal is a maximum deviation of 3°C.

If I use the standard way to connect the PTC (Directly on 5V) I still get a PTC teperature difference of about 20°C when varying the environment on the given range.

Is there a way to get better results ? I was thinking of switching the PTC OFF when >500ohm and ON when <500ohm but before building a schematic to achieve this I would like your opinion.

Some remarks :
- Application : Keeping a sensor on temperature using a PTC with minimum board dimensions in an temperature-changing environment
- Because of board dimensions I dont want a seperate heater/sensor
- Yep, I tested and the 0805 size gives enough power

Thanks

Patrick

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Are these PTCs just regular PTC thermistors or are they specifically meant for heating purposes?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Are these PTCs just regular PTC thermistors or are they specifically meant for heating purposes?

I'm testing with several Epcos SMD PTC Thermistors, several applications mentioned (mainly over-temp protection). Didnt find 0805 PTCs yet that mention my specific application but I think it wont be a problem mainly because in my application the heating is only activated for 30 minutes and will be used only 2/3 times a week. Not continuously heating 24x7

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Not sure what you are doing. It might be possible to use PWM to energize the PTC when on, then with delay, measure when off. The measurement current should be low to avoid sensor heating during measurement. Thermal mass of the system will help. Also the thermal resistance between the PTC, and heated device should be low.

It all starts with a mental vision.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Not sure what you are doing

In the past I made several different heating systems including seperate heating elements, sensors(NTC/Thermocouple/Chip).
But for this application I only have a very limited space and a very small object to heat. Thats why using a heater and separate sensor takes too much space, the space is limited to about 2x2mm.

I am now preparing a proof-of-concept test setup using a ptc as heater AND sensor. My goal is to make a setup where the PTC resistance is measured (using serial resistor) and than make a control loop that 'controls the PTC resistance to be 500 ohm'.

Just want to be sure if someone of you out there did the same and allready knows the results.

I'll keep you informed of the results.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

There are special heating PTC elements that need no control at all. They are used in car mirror heaters for example, but there are numerous other applications.

Unfortunately, these are very hard to get hold off, it's a typical OEM product you have to buy by the millions at a time.

I wonder if a regular PTC tempsensor used a heater lives that long.

2x2mm is indeed quite small. What is it what you are heating?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
2x2mm is indeed quite small. What is it what you are heating?

Thanks for you reply. I would love to share every detail of this project but we are very anxious the competitor is listening... At this moment lets just stick to 'the sensor'. The 2x2mm dimensions is INCLUDING the heater.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Paddy wrote:
... we are very anxious the competitor is listening...
No I'm not ... :lol:

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hahaha Ross McKenzie, I'm lucky you're on the other side of the world. That way you must read all messages upside down, isnt it ?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well, besides the obvious question of "What are you working on?", the two follow up questions are:

What is the desired system temeprature, and HOW are you presently measuring the temperature?

If the ambient temp ranges from 25-90'C, then presumably your set point is > 90'C, otherwise your system would also have to actively cool the system.

The question of measuring how the system is currently performing is of interest. You need an accurate method against which to measure the system's performance. With such a small "object" the thermal load of the temperature sensor may be an issue, particularly if one is cycling the temperature, and not just reaching steady-state and staying there.

You may want to look at how 3-D plastic printers regulate the print head extrusion nozzle temperature. You might find they have similar size and characteristics.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

@DocJC

I planned the setpoint indeed above 90, probably 95 or 100 degrees. Instead of the 'Sensor X' I now placed an NTC resistor which has about same dimensions and I can read that one out with a system accuracy of about 0.5 degree (calibrated), enough to do proof-of-concept measurements. This way I have similar thermal load/transfers as in the definitive setup.

The changing ambient temperature is done by using a themal cycler I have. It can be programmed to be e.g. 5 minutes 90°, 10 minutes 50° etcetera, no problem.

It could be that I use 2 PTCs, on both sides of sensor X to get a better heating uniformity. Or maybe one PTC below sensor X. All components are placed on a PCB, tests have to be done to decide which PCB material (metal ?) and thickness (0.5 ?) but bottleneck is now the heater element itself.

Quote:
You may want to look at how 3-D plastic printers regulate the print head extrusion nozzle temperature. You might find they have similar size and characteristics.

Thanks for the hint, I will have a look into that.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Since you can't say directly what it is , you have to live with some stupid questions :)
Do you have space outside the "heater", if yes how hot?
I would give the PTC a "variable" but stable DC, and use a two ADC's (tiny5 could do the job), one for Voltages over and one for current thru the PTC, and from that calculate the temperature.
I know that some old transmitters warm the xtal to have consistent temperature. perhaps there is some ideas.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
...transmitters warm the xtal...
Thanks for the idea, I'll look into that

Quote:
Since you can't say directly what it is , you have to live with some stupid questions

Yep, I know thats why I tried to define my question as clear as possible :

Quote:
Is there a way to get better results ? I was thinking of switching the PTC OFF when >500ohm and ON when <500ohm but before building a schematic to achieve this I would like your opinion.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you measure the temperature of the PTC, you end up controlling the temperature of the PTC and not the device being heated.

So you need to characterise the whole system to see what the actual temperature of the heated sensor is and hope the 'transfer curve' does not change. I guess there will be a few degrees of difference you might statically compensate for.