Want an efficient way to control heat in DIY heated clothing

[seanspotatobusiness]

I would like to make an item of electrically-heated clothing, using nichrome wire. I would like a way to control the heat and a PWM circuit has been suggested by others. In particular, I've seen 555-based circuits suggested but I've noticed people frowning on them. I'd like to know what's wrong with 555-based PWM circuits - are they inefficient?

I will most likely power my project with 4.8V NiMh.

Thanks in advance for any insight!

[AndyC_772]

I've done this with a simple R-C oscillator and a comparator. Rather than using the square wave output from the oscillator, take the sloping voltage across the capacitor and compare it against a variable threshold (eg. from a potentiometer). The output of the comparator becomes a rectangular wave with variable duty cycle from 0 to 100%, which can drive a power MOSFET.

Another good option is to use a small microcontroller - I've used a PIC very successfully to do the same job.

[Jimmy]

What you are trying to do is PID
http://en.wikipedia.org/wiki/PID_controller

[seanspotatobusiness]

I've done this with a simple R-C oscillator and a comparator. Rather than using the square wave output from the oscillator, take the sloping voltage across the capacitor and compare it against a variable threshold (eg. from a potentiometer). The output of the comparator becomes a rectangular wave with variable duty cycle from 0 to 100%, which can drive a power MOSFET.

Another good option is to use a small microcontroller - I've used a PIC very successfully to do the same job.

Hi. Are these options more energy-efficient? I didn't know what a comparator was, so I looked them up. I don't think I need one with hysteresis - it doesn't matter if the circuit is noisy. How can I find the simplest comparator? These ICs all have so many legs.

Do I need to know the difference between MOSFET enh and dep, p-channel and n-channel?

I've never used PIC before. I'm out of my depth.

Here's something probably-stupid that I came up with (don't laugh; I've spent all morning on this!):

[Rerouter]

now with a better understanding of your level of experience,

first off a comparitor is just a op amp with a fast slew rate (it can increase or decrease its output fast) and a high gain (so small changes in input cause it to swing between its supply rails (VCC and VDD) for the frequencies you would be working any old op amp should do the trick for you, even a basic 741,

andys design has a resistor and capacitor giving off a trangle waveform, this is fed into one of the input terminals of the op amp, on the other channel would be a potentiometer setting a voltage level for the other input, as soon as one voltage exceeds the other the output of the op amp will swing to the other rail, (it does depend on which is higher for what direction but i'll get to that)

so say your potentiometer is set at 3V and your triangle wave is swinging between 1 and 4V, it means 25% of the time the output will be at one rail and 75% of the time at the other rail, for better reference here is a better image of whats happening with input vs output,

now this method will give you something known as "constant wattage" as there is no temperature measuring device on it, that is what jimmy was err-ing to even if through a very unnecesarily complex manner, now if you want something you set at 32 degrees vs "warm" you need to use another op amp as an "error amplifier" now while that may sound technical, it just means it amplifies the difference between where you are and where you want to be, used to get your temperature to your "set point" with the normal method for this using the inverting input, e.g. too low a temperature results in a higher output wattage and thus a warmer wearer,

this is also done with a generic op amp, and seeing as you may want to learn a bit about them, here is the basics http://www.radio-electronics.com/info/circuits/opamp_basics/operational-amplifier-basics-tutorial.php

as for your "pass element" (what you use to control most of your current with your weak op amp output) you would probably want to look at using a p-channel mosfet, as this means you only need to pull its "gate" down to turn it on rather than needing a raised voltage for an n-channel, equally you would be looking for an "enhancement" type, firstly as you only want it to pass current as you pull it down, and they are more common,

now while there are near endless ways written on how to "drive a mosfets gate" in this case one of the best ways would be to have your op amp output go into an npn transistor, whose collector is connected to the mosfets gate, working as a switch to pull it down, equally you will need a lift up resistor, (say 1K) between the gate and the source, this will pull the gate back up when the transistor is not pulling it down, making it work like an on off switch, or "switchmode"

remeber when i mentioned error amplifiers? well to use on in this circuit, you would have lets say a simple thermistor in a voltage divider to give you something to measure, this goes into the inverting terminal of your op amp, and your "set point" voltage would go into the non-inverting input, and to make things a whole lot easier on yourself you would have an upper and lower "bias" resistor on your "set point" potentiometer to limit its voltage to ones the termistors divider would give out, in this circuit whenever the termistor voltage is lower than the "set point" voltage the op amps output will be raised, if its greater it will fall, and generally it will balance somewhere in the middle after a while and control your temperature,

this voltage would then be fed into the original op amp with the triangle wave, and would form a "closed" control loop,

i hope this has somewhat helped you

[T4P]

A comparator is a high slew versions of an opamp, they have been optimized to switch high between low instantly much faster than say a similar opamp
A comparator only exists to do two things : High Or Low. A square wave, but also basic digital circuits look like this on a scope (Basic! 1 and 0 only! Not states 2 and 3 and so on and so forth)

TI has a app-note explaining why you shouldn't really use a opamp like a comparator for higher speed circuits (As like TI says, can blow the opamp up  )
But for such a circuit it doesn't matter at all, after all.

[seanspotatobusiness]

Hi. I don't think I want to use a thermistor to maintain temperature because it makes the project a bit more complicated and I'm likely to want to keep changing the thermistor setting as much as I'd change the setting of the simpler circuit anyway (the temperature of my entire torso is best integrated by my brain than a single thermistor anyway) but thanks for explaining that suggestion.

Using the other information you provided, I was able to devise the following circuit diagram. Can you confirm that this is what you intended?



Regarding the 1K lift-up resistor: does the value make any difference whatsoever? What if it was a 1M resistor? Would the circuit behave differently?

Edit: I'm thinking of using this wire for the heating element. It's the best value, low-gauge, multistranded wire I can find. I already have some from a previous purchase and have determined that a length of 5m has a resistance of 1.0 ohm. If I use 10m, I can achieve 2 ohm.

@4.8v, the power/current will be maximum 11.5W/2.4A
@3.6v, the power/current will be maximum 6.5W/1.8A

[seanspotatobusiness]

I've made a simulation in Circuit Lab. Click here to use it. The circuit must be wrong because the capacitor charges to 4.8 volts and stays there whilst I thought it was supposed to charge up and then discharge via the LM741. Herp derp.

[AndyC_772]

Have a look at: http://www.allaboutcircuits.com/worksheets/opamp10.html

Questions 5 and 6 show very relevant circuits that you can simulate and adapt for your purposes.

[Rerouter]

here is the likely circuit, http://tinyurl.com/bnwfggk

a relaxation oscialltor set up for a single voltage rail and biased a bit so that it keeps about 1.25V from the rails to work with cheap op amps (5V)

and the set point potentiometer is biased so that your only swinging over the range you care about

the pull up resistor on the mosfet is to charge up its gates capacitance (its effectivly a small capacitor that has to be charged / discharged before its state changes) while the npn dumps it immediately, you want the lift up both small enough that it switches quickly but big enough that your not just wasting power through it when the npn switches, 1K seems like a fair mid point, 10K is also ok if you want to use more of the same value,

[seanspotatobusiness]

Thanks guys. I understand the circuit partially. Maybe I can address the part I don't understand later.

Thanks a lot for simulating that circuit for me, Router. There is a potential problem though, in that the voltage drop over the 2 ohm load is actually very small. Most of the drop occurs over the MOSFET so the power of the heating-element will actually be negligible.

[Rerouter]

the simulator only caters to signal mosfets whereas a normal power fet would have a much lower RSon and thus have no where near the voltage drop

[seanspotatobusiness]

Oh right, thanks! In that case, I think I'm ready to build the controller. Could you confirm that these parts are suitable?

http://www.ebay.co.uk/itm/ws/eBayISAPI.dll?ViewItem&item=390423690017#ht_3579wt_910

http://www.ebay.co.uk/itm/10-x-BC337-25-NPN-Transistor-500mA-/300713324500?pt=UK_Sound_Vision_Amplifier_Parts&hash=item4603e92fd4#ht_1868wt_959

http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=120987330133

[seanspotatobusiness]

Could anyone confirm that the above parts are suitable for this circuit?

[seanspotatobusiness]

C'mon guys, I live in Scotland and winter's coming... so... c-cold

[jeremy]

The 741 is a classic opamp, nothing wrong with that. As for the FET, you have a low Rds(on) so that seems good. BJTs are plain old BJTs.

Honestly, I don't think anyone can tell you if it will work or not without building it themselves (made that mistake before). It's only a few pounds, buy em and try it!

good luck

[seanspotatobusiness]

The 741 is a classic opamp, nothing wrong with that. As for the FET, you have a low Rds(on) so that seems good. BJTs are plain old BJTs.

Honestly, I don't think anyone can tell you if it will work or not without building it themselves (made that mistake before). It's only a few pounds, buy em and try it!

good luck

Thanks Jeremy. I'll report back when I've assembled it.