Go Down

Topic: solved: "invert" thermistor/Hacking Buck Converter for temp dependent output? (Read 2236 times) previous topic - next topic

wothke

Hi

I am trying to add a cooler (simple two wire 12cm 12V ventilator) in front of my beamer.

I want to control its speed based on the same kind of thermistor that is already used within the beamer itself (I recovered the respective thermistor from a defective beamer that I am using for spare parts: it is a 50kOhm resistor while cold and the resistance falls towards zero when the thermistor is heated).

To keep the ventilator in a "quiet" range I want to power it with between 3V and 6V (i.e. 3V while the beamer is cold and up to 6V while the beamer gets hotter).

My original idea was to use a simple buck converter (like this one: https://www.aliexpress.com/item/Free-Shipping-1-pcs-lot-MP1584EN-ultra-small-DC-DC-3A-power-step-down-adjustable-module/32639186437.html ) and just replace the potentiometer with something based on the thermistor - so that I could vary the output voltage based on the thermistor.

The problem now seems to be that the buck converter always associates lower voltages with lower resistance, i.e. the ventilator is getting slower with rising temperatures (instead of getting faster).
And so far I have failed to find a thermistor/resistors combination that would allow to invert that behavior.

From what I see the buck converter is not using all 3 pins of the potentiometer (see product photo on the above link): Of the two contacts that face to the left, the top contact (pin 1) is actually shorted with the contact that faces to the right (pin3) - that's the one that actually also produces the + output of the buck converter.

Is there some circuit that allows me to "invert" the thermistor's resistance? Or is there a way to patch the buck converter accordingly (mb cut the pin1 - pin3 connection)? What are the currents that a thermistor would typically survive - is it likely to be damaged by the current drawn by the ventilator?

raschemmel

#1
Oct 18, 2016, 05:53 pm Last Edit: Oct 18, 2016, 06:04 pm by raschemmel
Quote
I want to control its speed based on the same kind of termistor that is already used within the beamer itself (I recovered the respective termistor from a defective beamer that I am using for spare parts: it is a 50kOhm resistor while cold and the resistance falls towards zero when the termistor is heated).
  
Quote
beamer  
What's a "beamer" ? (BMW ?)

schematic ?

Unless I have misunderstood you , the explanation for what you describe is simple. The pot on the buck converter that sets the output voltage is meant to be set at a SPECIFIC INPUT voltage WHICH IS NOT TO BE CHANGED. Every time you change the input voltage , you have to recalibrate the converter for the new input voltage. You cannot simply set it at one input voltage and then change the input voltage and expect it to maintain some relationship of input to output voltage.

Quote
To keep the ventilator in a "quiet" range I want to power it with between 3V and 6V (i.e. 3V while the beamer is cold and up to 6V while the beamer gets hotter). 
Did you READJUST (recalibrate)  the converter output voltage AFTER you CHANGED the INPUT voltage ?

wothke

#2
Oct 18, 2016, 06:32 pm Last Edit: Oct 18, 2016, 06:45 pm by wothke
What's a "beamer" ? (BMW ?)
beamer = home video projector


Unless I have misunderstood you , the explanation for what you describe is simple. The pot on the buck converter that sets the output voltage is meant to be set at a SPECIFIC INPUT voltage WHICH IS NOT TO BE CHANGED. Every time you change the input voltage , you have to recalibrate the converter for the new input voltage. You cannot simply set it at one input voltage and then change the input voltage and expect it to maintain some relationship of input to output voltage.

Did you READJUST (recalibrate)  the converter output voltage AFTER you CHANGED the INPUT voltage ?

you  misunderstood -- I do not change the input voltage of the buck converter AT ALL - it is statically connected to some power supply... the idea is to "readjust" the OUTPUT voltage at runtime depending on the current temperature of the "video projector". technically a respective runtime readjustment seems to work fine but unfortunately the temperature/resistance correlation of the thermistor that I have at my hands is the inverse of what I'd need to readjust the converter.. (higher temperature results in LOWER resistance but lower resistance makes the buck converter output LOWER voltage ... but I need HIGHER voltage)

raschemmel

#3
Oct 18, 2016, 06:36 pm Last Edit: Oct 18, 2016, 06:41 pm by raschemmel
I don't understand the correlation between a thermister and the output voltage trim pot on the buck converter which has no thermister. Are you saying you are "HACKING" the converter to change the output voltage ? (by adding a thermister ?)

IF this is the case, then I SERIOUSLY recommend you CHANGE your post title in your ORIGINAL post to:

"Hacking Buck Converter for temp dependent output"

wothke

I don't understand the correlation between a thermister and the output voltage trim pot on the buck converter which has no thermister. Are you saying you are "HACKING" the converter to change the output voltage ? (by adding a thermister ?)
yes, that's what I explained in my original post: I remove the trim pot and replace it with a thermister (and whatever additional resistors or other components might be needed..) - so that the thermister will will perform dynamic trimming at runtime..

IF this is the case, then I SERIOUSLY recommend you CHANGE your post title in your ORIGINAL post to:

"Hacking Buck Converter for temp dependent output"

done.. I had chosen the original title since my problem would also be solved if a circuit exists that allows to construct "a resistor R1" such that R1= 50kOhm - resistance of thermistor


MrAl

yes, that's what I explained in my original post: I remove the trim pot and replace it with a thermister (and whatever additional resistors or other components might be needed..) - so that the thermister will will perform dynamic trimming at runtime..

done.. I had chosen the original title since my problem would also be solved if a circuit exists that allows to construct "a resistor R1" such that R1= 50kOhm - resistance of thermistor


Hi,

See if you can find a positive temperature coefficient thermistor.  If not, then you'd have to modify the feedback circuit to maybe put the thermistor in the top of the voltage divider instead of the bottom, or vice versa depending on what buck you are using.

Many bucks work by a feedback from a voltage divider.  When the bottom resistor value gets smaller the output voltage increases in order to try to equalize the feedback voltage and the reference voltage.

One thing to keep in mind here though is that there is no guarantee that the thermistor resistor value will change in a manner that provides the right feedback level anyway, which controls the fan speed.
In the old days it took a lot of trial and error with series and/or parallel resistors to get a thermistor to properly compensate a device that was considered too temperature dependent.

A better idea, and since we are in the Aruduino forum, would be to use an Arduino to measure the fan speed or temperature and provide the exact programmed speed profile you are after.  A temperature sensor would probably be all that you need, besides a driver of some kind.

raschemmel

Quote
A better idea, and since we are in the Aruduino forum, would be to use an Arduino to measure the fan speed or temperature and provide the exact programmed speed profile you are after.  A temperature sensor would probably be all that you need, besides a driver of some kind.
In my mind, that's the only sensible approach.

Jiggy-Ninja

Any reason you can't PWM the fan with an ATtiny85 reading temperature from the NTC? With a micro in the path you can put whatever temperature-fan speed mapping you want, rather than more complicated analog signal conditioning.
Hackaday: https://hackaday.io/MarkRD
Advanced C++ Techniques: https://forum.arduino.cc/index.php?topic=493075.0

Jiggy-Ninja

Is there some circuit that allows me to "invert" the thermistor's resistance? Or is there a way to patch the buck converter accordingly (mb cut the pin1 - pin3 connection)? What are the currents that a thermistor would typically survive - is it likely to be damaged by the current drawn by the ventilator?
Answering these questions about the NTC:

No, you cannot "invert" its temperature characteristic, or in fact change it in any way. You can only change the circuit it is attached to.

You could achieve what you want by putting the NTC on the bottom of the feedback voltage divider with a calculated fixed resistance replacing the pot, but then you are very constrained in how the two variables are mapped unless you use a more advanced signal conditioning method (likely needing op amps).

The thermistor is not in the direct path of the load power. The feedback circuit is in parallel with the load, so whatever the load is doing will have no effect on the current in the thermistor.
Hackaday: https://hackaday.io/MarkRD
Advanced C++ Techniques: https://forum.arduino.cc/index.php?topic=493075.0

allanhurst

Can I suggest you stick with the original (NTC) thermistor. BMW obviously designed the fan to be adequate to cool the engine sufficienly under all conditions - and if you restrict the fan voltage it may overheat..

regards

Allan.

Jiggy-Ninja

Can I suggest you stick with the original (NTC) thermistor. BMW obviously designed the fan to be adequate to cool the engine sufficienly under all conditions - and if you restrict the fan voltage it may overheat..

regards

Allan.
I think you missed something.
Quote
beamer = home video projector
Not an automobile.
Hackaday: https://hackaday.io/MarkRD
Advanced C++ Techniques: https://forum.arduino.cc/index.php?topic=493075.0

raschemmel

Quote
No, you cannot "invert" its temperature characteristic  
Yes it is a shame we can't just "invert" Physics.

Quote
I think you missed something.
Quote
beamer = home video projector 
;D

wothke

Thanks for the various replies. I meanwhile found out that the buck converter is based on this: http://www.monolithicpower.com/Products/Product-Detail/101/Step-down(Buck)/SwitchingRegulators/DC-DCPowerConversion/Converters/MP1584
and I located the resistor that is used for the other half of the voltage divider. I just replaced it with my thermistor and with a bit of tuning on the original trim pot side now everything works as desired.


To answer the question why I am not using an Arduino/PWM approach: If I had started from scratch I would probably have bought some PWM enabled fan and controlled it using some small Arduino. But since I already had some old fan lying around I did not want to waste money for an additional fan nor wait for its delivery.  Directly using the buck converter seems to be a simple but good enough solution. Also the patched buck converter is compact enough to glue it into the fan's frame - which makes for a tidy design.

MrAl

Thanks for the various replies. I meanwhile found out that the buck converter is based on this: http://www.monolithicpower.com/Products/Product-Detail/101/Step-down(Buck)/SwitchingRegulators/DC-DCPowerConversion/Converters/MP1584
and I located the resistor that is used for the other half of the voltage divider. I just replaced it with my thermistor and with a bit of tuning on the original trim pot side now everything works as desired.


To answer the question why I am not using an Arduino/PWM approach: If I had started from scratch I would probably have bought some PWM enabled fan and controlled it using some small Arduino. But since I already had some old fan lying around I did not want to waste money for an additional fan nor wait for its delivery.  Directly using the buck converter seems to be a simple but good enough solution. Also the patched buck converter is compact enough to glue it into the fan's frame - which makes for a tidy design.
Hi again,


I am happy that method worked for you.  It's really an old method but hey if it works, why not use it right?

If you dont get the right profile if it is too extreme then you can try adding a parallel resistor and change the upper series resistor to compensate.   The parallel resistor decreases the resistance swing of the thermistor so there is less resistance change over the full temperature range.  There is some relatively simple mathematics to it i am not sure if you are interested or not, but the simplest part is to calculate the parallel resistance for both temperature extremes and see if that matches your desired end point settings.  You dont get much control over the center range without adding a second thermistor, so the end point settings become a sort of compromise between getting the right end point settings and the desired mid range operation.  I am also not sure how accurate you really need this to be anyway though :-)


wothke

Hi again,

I am happy that method worked for you.  It's really an old method but hey if it works, why not use it right?
exacty my line of thought :-)


If you dont get the right profile if it is too extreme then you can try adding a parallel resistor and change the upper series resistor to compensate.   The parallel resistor decreases the resistance swing of the thermistor so there is less resistance change over the full temperature range.  There is some relatively simple mathematics to it i am not sure if you are interested or not, but the simplest part is to calculate the parallel resistance for both temperature extremes and see if that matches your desired end point settings.  You dont get much control over the center range without adding a second thermistor, so the end point settings become a sort of compromise between getting the right end point settings and the desired mid range operation.  I am also not sure how accurate you really need this to be anyway though :-)
that's what I'll do in case I need to make improvements..

In fact the fan is a non-critical backup for the 4 regular fans that are already built into the projector. Theoretically it should not be necessary at all. But I had just repaired my 8 year old device using a 2nd defective projector for spare parts. And I noticed that there seem to be design flaws regarding the upside-down mounting of this particular model, which had caused the plastic case to age as badly that it was near impossible to (un)fasten any screws without having the brittle plastic break. The extra fan will hopefully provice some extra cooling and extend the life expectancy of my "new" replacement housing.. Also the big external fan is rather quiet (as compared to the small internal ones - which can get rather noisy if the device decides to put them into overdrive..). I just put the stuff together and from what I see the design seems to work nicely: the temperature inside the projector actually stays a notch cooler than before (eventough the extra fan is not even going fast yet)..


Go Up