Inserting a variable resistor into a DC circuit

[gerbold]

Hi,
I've built a hot-wire device for cutting polystyrene. I used a computer power module (ATX-350) drawing off from the 12v rail/line. I've use 28 gauge Nichrome wire. The arrangement works insofar as the wire heats up to orange glow. However I feel I need a fine control. I bought a Lutron dimmer switch but have since discovered this is appropriate to the AC side and would not have any affect on DC regulated voltage. I did actually insert it but, given the disparity in function, teh obvious happened; nothing. It appears that, on the DC side, I need to control amperage rather than voltage. Can any member suggest a device which I could insert into the circuit? A LED dimmer switch has been suggested (12volt DC 8amp) but the potential amperage on the ATX may be as high as 17amp max. I sense using any device rated less could lead to an overload situation. Perhaps too a switch, fuse and indicator would be sensible insertions into such a circuit. Suggestions please. Thanks in anticipation of your assistance.

 

[ard]

Hi, using a resistor of any sort in series with your nichrome wire would be a highly wasteful process, and for your current requirement at 12v, the wasted power would be very significant.

What you need is a pwm circuit, driving a proper mosfet capable of handling the required current.

Regards,

Anand

 

[gerbold]

Hi Anand - Thank you so much for your reply. I am unfamiliar with the terms MOSFET and PWM - although I've already encountered them in internet searches. I will investigate them more closely. If you can specify a particular variant and/or explain the theory further given the parameters I gave then that would be very helpful and appreciated.

Hi, using a resistor of any sort in series with your nichrome wire would be a highly wasteful process, and for your current requirement at 12v, the wasted power would be very significant.

What you need is a pwm circuit, driving a proper mosfet capable of handling the required current.

Regards,

Anand

 

[ard]

Hi ,

PWM is pulse-width-modulation, by way of which you can control the total power delivered to the load. Multiple ways to achieve this, one such design is at **broken link removed**

For the parameters you describe, this circuit should be perfect. You might need to mount the IRFZ Mosfet on a heat sink.

Regards,

Anand

 

[gerbold]

Hi ,

PWM is pulse-width-modulation, by way of which you can control the total power delivered to the load. Multiple ways to achieve this, one such design is at **broken link removed**

For the parameters you describe, this circuit should be perfect. You might need to mount the IRFZ Mosfet on a heat sink.

Regards,

Anand

Hi again, Anand, thanks for giving me of your time and expertise and the reference.

I looked up MOSFET. I think I understand how they work. A variable resistor would have current continually applied to it whereas a MOSFET switches off and on continously as it senses a given power level being reached - the heat sink is for coping with the heat energy produced while the MOSFET is in its 'on' phase - is that right? And pulse-width-modulation is the method for which the MOSFET capabilities need to be specified in relation to the given parameters i.e Current type; amperage; voltage; resistance. Is that a fair interpretation - in layman's parlance?

Learning is so much fun! Thanks again, Anand.



Anand - one factor eluded me in my euphoria - how do I use the MOSFET to manually control the wire temp.?

 

[betwixt]

Superfast tutorial:

There are two ways to control your wire temperature:

1. Linear control using a variable resistor or similar. The power dissipated in the wire determines it's temperature and can easily be calcuated by dividing the voltage across the wire squared (V * V) by the wires resistance (R). Watts = (V*V)/R. Nichrome (Nickel & Chrome alloy) has quite a high resistance which makes it suitable for heating purposes. The trouble with linear control is it wastes power in the control circuit. Over the usable range you are interested in, as you reduce the power to the wire you increase the power lost in the control circuit.

2. PWM control which works in a far more efficient way. With PWM you don't limit the voltage to the wire at all, instead you control how long it is applied for. It works by turning the voltage on and off cyclically. This makes the wire go through a heating period followed by a cooling period so although power is full when heating, the AVERAGE temperature will be lower. Obviously you don't want the temperature to pulsate like that so the on/off period is made very fast, often as fast as hundreds of thousands of times every second, so the wires own mass evens out the temperature 'bumps'. You vary the temperature by adjusting the length of 'on' time compared to 'off' time, for exampe if you made them equal the heating and cooling periods would be the same and the result would be half the heating power. This is what the PW in PWM stands for, "Pulse Width" Modulation. An electronic timer circuit allowing you to control the ratio of on and off time within each cycle would control the power delivered to the wire and hence it's temperature. 25% on and 75% off would give on quarter power, 33.3% on and 66.6% off would give one third power and so on.

A MOSFET can be considered as a nearly perfect voltage operated switch. They can operate at very high speed which is why they are useful as PWM on/off switches and they require very little power to operate them so they can be driven from simple timer circuits. The beauty of PWM is that the power lost in the controller is very small. Like in the wire, the power dissipated in the MOSFET is V*V/R although in this case the 'R' isn't the resistance of the Nichrome wire but the 'on resistance' of the MOSFET which can be as low as hundredths of one Ohm. When the MOSFET is turned off (wire cooling period) the voltage is he 12V from your PSU and no current is flowing so no power is dissipated at all. when the MOSFET is turned on, the voltage across it drops to almost zero (it's all across the wire) so again the power is very low. The only significant power loss is during the transition from on to off and back again but this is an extremely short time so the average power loss is very small.

What you need is a circuit to produce an adjustable PWM ratio, driving a MOSFET as an efficient switch in series with the wire.

Brian.

 

[walt]

3. You can modify your computer power module to reduce its voltage with external resistor and add voltmeter to measure it.

 

[betwixt]

Walt, 3. is the same as 1. in my posting. The external resistor may heve to be extremely large and might get very hot.

Brian.

 

[gerbold]

Superfast tutorial:

What you need is a circuit to produce an adjustable PWM ratio, driving a MOSFET as an efficient switch in series with the wire.

Brian.

Thank you for your input Brian. It confirmed what I voiced in my reply to Anand. The critical factor in the circuit is the wire. I would need to adjust the current across the wire to an optimum temp. high enough to cut cleanly and low enough to avoid vapourisation of the wire. This is why I asked Anand for info on a device for setting the MOSFET 'pulse-width. Or can that be calibrated on the MOSFET itself? I have very little knowledge of the type of device. Also with a control I could shorten or lengthen the wire in my jig for different widths of material.
thanks

 

[betwixt]

Great - get the parts for the design posted by Anand, it isn't the most advanced of designs but it should suit your purpose. The parts should be very inexpensive. If you can't get those exact diodes, the more common 1N914 will work just as well.

Connect your cuting wire where the motor is in the schematic and join the points marked "+3-18V" and "+V MOTOR" to each other and to the +12V from your PSU.

If you can't find the parts, tell us where in the World you are and we will advise on a local supplier.

Brian.

 

[walt]

I suggested to modify computer power module. There's a lot of schematics in the web.You replace resistor with simple low power potentiometer and can regulate output voltage and output current on constant load as well.

 

[gerbold]

Hi Brian, I'm over the border in England - South-West Leicestershire. I would think that Maplins would be the obvious choice - although they're relatively expensive. No matter.

I found this page in a search for "variable power control in a MOSFET pwm" -
http://weblog.sirajs.com/article/555-timer-based-pwm-power-control-circuit
- gives all the theory and is specific to my requirement. And this site gives the practical, theory, pictures and parts list -http://www.instructables.com/id/Very-simple-PWM-with-555Modulate-every-thing/step5/Building-it-/.

In this one the author (15yo 'Shams') lists the following
1.NE555,LM555 or 7555(cmos)
2.two diodes 1n4148 is recomended but you can also use 1n40xx series diodes
3.100k pot(volume control pots are good for this circuit)
4.100nf green cap
5.220pf ceramic cap
6.breadbord
7.power transistor

The 100k pot (eg volume control pot) is where I can physically adjust the power - is it not?

He also provides a schematic which is alike to Anand's. I think the job is very close to being sorted. Any observations still welcome though. I suspect there already is a PWM circuit in the power unit? Maybe instead of calling myself Gerbold it should have been 'Ug' with a caveman avatar. .

- - - Updated - - -

I suggested to modify computer power module. There's a lot of schematics in the web.You replace resistor with simple low power potentiometer and can regulate output voltage and output current on constant load as well.

Hi Walt, thanks you for your input. A guy in my local electronic shop in England suggested a potentiometer but I needed much more background info before buying. I've found more and more practical suggestion today for a PWM circuit. But I'd welcome your ideas or even spec. for a suitable potentiometer circuit. Thanks

- - - Updated - - -

Sorry Walt - I just relaised 'pot' = potentiometer. Are you suggesting I could put a, say 100k volume potentiometer, into my circuit by itself (or am I suggesting that?). thanks

 

[D.A.(Tony)Stewart]

Depending on the duty cycle of operation and variable speed of cutting thru thick blocks, your control may need to be adjusted continuously.
There is a delay in temperature change after changing the setting due the change in resistance of the filament wire. The resistance can be measured to regulate the temperature regardless of cutting fast or just open air. But this is more complex.

What you need to do is measure the current and voltage when cold , too hot and just right. Then a regulator ideal for your application can be designed.

 

[betwixt]

I think the idea is to make it adjustable so it can be used with different 'hot wires' rather than to find the optimum temperature for a single wire so there is no 'ideal' setting for a regulator.

Yes, "pot" is engineering slang for "potentiometer" but I wouldn't advise using a volume control type. To explain, there are two types of potentiometer in common use called 'linear' and 'logarithmic'. The difference is in how the resistance from each end to the wiper (arrowed connection) varies as it is turned. In a linear potentiometer the value is proportional to the amount of rotation so for example a 10K value set half way would have 5K to each end. A logarithmic one has its resistive track constructed so the value increases logarithmically rather than in equal increments per degree of rotation. Our hearing works logarithmically, we are more sensitive to changes in sound at low level than when its loud so volume controls are usually logarithmic to compensate for that. What you need is a linear type which gives 'smooth' control over it's rotation. If you use a volume control you might find almost no change in temperature over most of the rotation and all the variation you want in just a few degrees at one end.

Maplin should be able to supply all the parts. If not, the source I use is Farnell in Leeds who are masive in comparison with Maplin but they do have a minimum order charge of I think 20.00

Brian.

 

[D.A.(Tony)Stewart]

Does PS smoke initially then cool down to cut smoothly smoke free ?

If so then temperature sensing Constant current control is required.

PS has a melting point of 260C and can start burning at 360C but not cut well below 260.

Nichrome of one type changes resistance by about 100 ppm/ deg C.

Temperature of free air straight AWG 28 target should be around 316'C (600F) or 1.55Amp

does it smoke first then cool down OK?

 

[walt]

You will have to dismantle your PC power module and do some soldering. You can use this link http://danyk.cz/at_atx_en.html or some similar for guidance. Pay attention to all warnings!