# Issue controlling PWM with thermistor

#### TheMasterXXXXX

Joined Aug 23, 2022
29
I am attempting to use a thermistor to control the PWM of a 555 output. Ideally, I want the duty cycle to be 100% when hot (~90°C) and 0% when not hot (~25°C). I am doing so by placing the thermistor as R2 (RB) in astable mode with a forward biased diode in parallel to R2 (allowing a duty cycle under 50%). Frequency is not important.

However, with the most suitable thermistor I have found, I can only achieve ~18% at 25°C and ~82% at 100°C.
I would like to even more lower the duty cycle when cold, and raise it when hot. How can I achieve this?

The thermistor's values are:
150kΩ at 25°C and 7.2kΩ at 100°C
B = 4500K

In principle a higher B value should provide me with the desired result, however it seems as though thermistors with much higher B values do not exist.

Here you can see a graph I made showing duty cycle as a function of temperature: https://www.desmos.com/calculator/yixsrolmwm. My desire is to increase the slope, or at least cause the duty cycle to jump to ~100% once a certain temperature is reached.

Any help is greatly appreciated

#### Papabravo

Joined Feb 24, 2006
21,254
I am attempting to use a thermistor to control the PWM of a 555 output. Ideally, I want the duty cycle to be 100% when hot (~90°C) and 0% when not hot (~25°C). I am doing so by placing the thermistor as R2 (RB) in astable mode with a forward biased diode in parallel to R2 (allowing a duty cycle under 50%). Frequency is not important.

View attachment 274559

However, with the most suitable thermistor I have found, I can only achieve ~18% at 25°C and ~82% at 100°C.
I would like to even more lower the duty cycle when cold, and raise it when hot. How can I achieve this?

The thermistor's values are:
150kΩ at 25°C and 7.2kΩ at 100°C
B = 4500K

In principle a higher B value should provide me with the desired result, however it seems as though thermistors with much higher B values do not exist.

Here you can see a graph I made showing duty cycle as a function of temperature: https://www.desmos.com/calculator/yixsrolmwm. My desire is to increase the slope, or at least cause the duty cycle to jump to ~100% once a certain temperature is reached.

Any help is greatly appreciated
I don't think you can achieve your goals if RB is the only resistance you can change. If RA and C remain fixed, then you are stuck with the limits on the value of RB

#### TheMasterXXXXX

Joined Aug 23, 2022
29
C has no effect on duty cycle. When RB is 0 duty cycle is 100%, and RB is high it approaches 0%, so I can obtain the desired value by only changing the effective resistance of RB. Any ideas?

#### crutschow

Joined Mar 14, 2008
34,673
Don't offhand know how to do what you with a 555, but below is the LTspice simulation of a PWM circuit using a LM339 or LM393 comparator chip that will go from 0% to 100% duty-cycle.
It goes to 100% when the Therm voltage is above the peak of the triangle-wave, and 0% when Therm is below the minimum triangle-wave voltage.
I was lazy so, instead of calculating the values, I just tweaked R7 and R8 to give a near 100% duty cycle with a thermistor resistance (Rt) of 7.2kΩ (yellow trace) and near 0% at 150kΩ (red trace).

#### TheMasterXXXXX

Joined Aug 23, 2022
29
Don't offhand know how to do what you with a 555, but below is the LTspice simulation of a PWM circuit using a LM339 or LM393 comparator chip that will go from 0% to 100% duty-cycle.
It goes to 100% when the Therm voltage is above the peak of the triangle-wave, and 0% when Therm is below the minimum triangle-wave voltage.
I was lazy so, instead of calculating the values, I just tweaked R7 and R8 to give a near 100% duty cycle with a thermistor resistance (Rt) of 7.2kΩ (yellow trace) and near 0% at 150kΩ (red trace).

View attachment 274563
Thanks, although It would really be ideal for me to to implement this with the 555 if possible. Would there be a way to short around RB when its resistance falls below a threshold at higher temperatures?

#### crutschow

Joined Mar 14, 2008
34,673
Would there be a way to short around RB when its resistance falls below a threshold at higher temperatures?
Possibly, but that sounds like a somewhat complicated and kludgy circuit.

555's are not well suited to go from 0% to 100% duty cycle by changing a single resistance.

Good luck with trying to press a square-peg into a round hole.

Last edited:

#### Papabravo

Joined Feb 24, 2006
21,254
C has no effect on duty cycle. When RB is 0 duty cycle is 100%, and RB is high it approaches 0%, so I can obtain the desired value by only changing the effective resistance of RB. Any ideas?
You already pointed out that thermistor values don't go to 0 or "high" so you can't get the duty cycles you want. Feel free to experiment and maybe you'll come up with "new" electronics.

#### TheMasterXXXXX

Joined Aug 23, 2022
29
You already pointed out that thermistor values don't go to 0 or "high" so you can't get the duty cycles you want. Feel free to experiment and maybe you'll come up with "new" electronics.
Yes, but I can also modify any part of the circuit as I so choose. I was hoping there was some not overly complicated method to accomplishing this.

#### crutschow

Joined Mar 14, 2008
34,673
Is your reason for trying to do this with a 555 is because you happen to have those in stock?

Last edited:

#### Papabravo

Joined Feb 24, 2006
21,254
Yes, but I can also modify any part of the circuit as I so choose. I was hoping there was some not overly complicated method to accomplishing this.
Of course, you can, and I invite to come up with some truly revolutionary insight.

#### crutschow

Joined Mar 14, 2008
34,673
It should be apparent by now, that if you really want to go from 0% to 100% duty-cycle using that thermistor as a control signal, you will need to use something other than a 555.
The typical 555 circuit that does that, uses a 3-terminal potentiometer for control of the duty-cycle, which the thermistor cannot readily emulate.

#### Alec_t

Joined Sep 17, 2013
14,370
Even if you could get a 0%-100% duty cycle, what do you intend doing with the duty cycle?

#### TheMasterXXXXX

Joined Aug 23, 2022
29
Even if you could get a 0%-100% duty cycle, what do you intend doing with the duty cycle?
Drive a fan using a switching transistor. I know there are easier ways, but I already have the 555 on my board and was thinking I could multitask with it. One point to note is that, in the circuit above, as the temperature increase the frequency also increases. Could this increase in frequency also be used to somehow further increase the duty cycle (by further lowering the effective impedance of RB)?

#### Ya’akov

Joined Jan 27, 2019
9,228
It should be apparent by now, that if you really want to go from 0% to 100% duty-cycle using that thermistor as a control signal, you will need to use something other than a 555.
The typical 555 circuit that does that, uses a 3-terminal potentiometer for control of the duty-cycle, which the thermistor cannot readily emulate.
Hmm… a voltage divider featuring “matched“ PTC and NTC thermistors?

#### crutschow

Joined Mar 14, 2008
34,673
I already have the 555 on my board and was thinking I could multitask with it
What other task is it doing?

#### TheMasterXXXXX

Joined Aug 23, 2022
29
Hmm… a voltage divider featuring “matched“ PTC and NTC thermistors?
NTC and PTC thermistors have drastically different graphs, so that would likely not be possible. Might the fact that frequency also increases with temperature be useful?

#### TheMasterXXXXX

Joined Aug 23, 2022
29
What other task is it doing?
Being used as Schmidt trigger. The board is powering a fan. Essentially I want to have one mode which toggles the fan on once a certain temperature is reached (Schmidt trigger/comparator), and another mode with PWM according to temperature.

#### crutschow

Joined Mar 14, 2008
34,673