# ptc heater control inrush limiter

Discussion in 'General Electronics Discussion' started by mpx96, Mar 25, 2010.

1. ### mpx96

8
0
Mar 25, 2010
I'm working with a PTC heater circuit. I've been stuck with something anyone who worked with a PTC heaters will understand. The PTC heater was chosen due to its self limiting characteristics. I'm running a small PTC heater element off of 24vdc. The problem is the inrush current on the heater. If the heater is run at 100% duty cycle, the current quickly drops to 3amps (which is acceptable). Here is the problem, the heater will quickly jump to 9amps when initially turned on cold, then over several seconds, drop to a steady state of 3amps. I've been working on a PWM control that limits the current by adjusting the duty cycle to 4amps, but the heater never makes it over the hump and runs crippled. If I allow the duty cycle to climb to 9amps, then the heater will run all day at 3amps after going over the 9amp peak.

I've been working on the idea that the voltage should be controlled using a PWM chopper circuit, but it looks like I may need to add some sort of pass transistor limit that allows the heater to be at 24v constantly. Originally, we tried a BJT shunt limiter but it generated as much heat as the heater itself!

Any clever people out there who solved this problem or should I punt and find a linear resistance heater?

2. ### 55pilot

434
3
Feb 23, 2010
I am not sure you can do what you are trying to do. I certainly do not see why you would want to!

The whole idea of a PTC heater is that you let it get hot enough to limit its current. The temperature at the core of the heater is usually well over 100C. If you start limiting the current by PWM, you will never let the core get hot enough to limit the current on its own.

If you are going to go through the trouble of adding a PWM to limit the current, then just use a regular heater or resistor (depending upon what exactly you are doing), put a small RC or LC filter at the input and PWM a lower resistance load to get whatever current you want.

---55p

3. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,412
2,780
Jan 21, 2010
The best solution is to limit your current to a higher value. The datasheet should tell you what this current it.

edit: I'm sorry, I read it as you were using PWM to control the heater inrush current, not to adjust the temperature.

Last edited: Mar 25, 2010
4. ### mpx96

8
0
Mar 25, 2010
I had thought of adding a rather large inductor to the heater to reduce the inrush but the inductance would have been very large and I am trying to keep things small here.

So, from your replys, it seems that if the current doesn't get high enough, then the core will not heat sufficiently and the PTC will not self regulate.

I think you understand what I was attempting to accomplish. The regulation of temperature part of the PTC device is not the problem, but the initial current inrush is.

Does anyone have any thoughts about adding a resistive heater in parallel with the PTC as sort of a core pre-heater to reduce the inrush in the PTC? I would turn off the resistive pre-heater once the PTC began to self regulate.

5. ### 55pilot

434
3
Feb 23, 2010
Let me repeat my question. If you are going to go through all this trouble, of adding parallel circuits, disconnects and bypass switches, PWMs and so on, why not just use a proper value resistor or a smaller resistor with a PWM to limit the average PWM?

What exactly are you trying to achieve? The big picture please.

---55p

6. ### mpx96

8
0
Mar 25, 2010
Figured out the solution I think

55pilot,

I think I figured out the solution today. As can be seen from the I Vs V plot, there is a sharp peak where the current peaks then drops quickly. This is what you hinted at when you said the core temperature needs to be met first. The I vs Temperature plot shows that the device is self regulating when the PTC temperature is near the set point. For my control to work, I need to add a resistive heater in parallel, turn it on as a preheater to get to the core temp, turn it off, then turn on the PTC and use it. The other method, which is simpler is to connect the resistive heater in series with the PTC, energize both, then allow the PTC to limit the resistive heater to its target temperature.

By using a linear resistive element to contribute the bulk of the heat required, I can simplify the electronics to a simple FET switch and allow the heater to limit its own current.

The big picture here is creating an air heater that doesn't require complex PID loop controls that is cost effective with no risk of runaway when a fan locks up or the air input side of the heater module is blocked (a problem we had in the past).

Regards, MPX96

#### Attached Files:

File size:
24 KB
Views:
795
• ###### PTC_I_VS_V.jpg
File size:
21.9 KB
Views:
671
Last edited: Mar 26, 2010
7. ### Resqueline

2,848
2
Jul 31, 2009
How about adding an NTC in series with the PTC, being in reasonable thermal contact?

8. ### 55pilot

434
3
Feb 23, 2010
You can try it, but it is unlikely to work. The resistor will not be able to heat the core of the PTC to the point where it gets into current limiting. The outside of the PTC will have to be significantly hotter to get the core to be that warm and to get the outside that hot, you will need a lot of energy. And what happens if the power is dropping out? With no temperature feedback, are you going to keep turning on the resistor every time the power cycles? So much for a safe and reliable system. That one is clearly not inherently safe, as you have a failure mode (just as likely as a failed fan) that can leave the resistor on almost all the time.
As I already pointed out, that one is certain to not work. The PTC is not going to get hot enough to go into regulation. As a result, you will end up with basically two resistors in series. You should also consider running that one at a few different temperatures. The setup has several different local minima that it can settle in and which one it settles in will depend upon the starting conditions.

You can try those circuits if you are unable to follow the analysis or unwilling to believe the results. It is easy to prove one way or the other.

Good luck.

---55p

9. ### mpx96

8
0
Mar 25, 2010
55pilot,

After speaking on the phone with an app engineer for about an hour, I now understand the fundamentals for the operation of the PTC heater. He informed me that they can formulate a high resistance PTC element that does not have the high inrush problem and operates similarly to the low resistance PTC. The good news is that they can make these devices in low volumes.

If anyone is interested in PTC heaters, check out the EPCOS website for application notes and a good paper on PTC thermistors.

I expect to see my custom formulated PTC in about 4 weeks.

Thanks. -MPX96

10. ### 55pilot

434
3
Feb 23, 2010
That is news to me. It may come in handy some day. You learn something new every day!

Thanks mpx96.

---55p