Variable power supply? Well not yet! Help

I've made several high current supplies using an LM317 & 2N3055 pass transistors. You'll need several heat sinks for the pass transistors & it needs a big thermal runaway resistor off each transistor also. Lots of mech shit on your pristine supply. Google " lm317 high current".

If you're only using it to heat nichrome wire..you might consider just using a lamp dimmer & an isolation transformer.

You don't need much to heat up wire. I think maybe 10W is a good estimate for your application.

If you take about 1 foot of AWG 30 which has a resistance of about 0.10320 ohms then that means you have to run 1A through it to get that.

The issue is that your appling 13V to it and so it will actually draw about

125A. Now if you had a 1V source then you could do it.

(Although I'm neglecting that the resistance will increase when it is heated... maybe by a factor of 10(happens atleast with a lightbulb))

So ultimate what you have to have is some type of conductor that has a large resistance.

if you can get something that has about 10ohm then that means you can draw about 1.3A through it and dissipate 17W.

Before you worry about making it variable you might want to actually get something to work. Adding in a series resistor isn't going to cut it. If you find a 5 ohm pot using the 10 ohm wire above then you have 10 to 15 ohms... this gives you a dissipation from 7.5W to 16.9W but the pot will need to dissipate about 5W(3.76 to be exact).

A light bulb filament is about 10 ohms cold and 100 ohms hot. The above does not take into account this change.

Normal coper wire isn't going to work because your voltage is to high producing extreme currents. So either you need to reduce the voltage(A 1VDC

20A source would work nice) or increase the resistance.

A better way would be to use AC and an SCR. Its very easy to control the power into the load and vary it.

Your best bet is to find some high resistance conducting wire that is made for that sort of stuff. Normal wire isn't going to work. I think what is commonly used for this sorta stuff is nichrome:

Once you find some thing to use, if you haven't already, and are still confused then post what you have.

For some high power pots,

So if you had a nominal Rw ohm hot wire and Rp ohm pot then

When Rp is all the way up,

I = V/(Rw + Rp) Pw = I^2*Rw Pp = I^2*Rp

and when Rp is off,

I = V/Rw Pw = I^2*Rw Pp = 0

You can use these formulas to find out the max current drawn(V/Rw) and the max dissipations in the devices(I^2*Rw for the hot wire and I^2*Rp for the pot). These all should be less than the max ratings. You can also use them to solve for Rw and Rp given whatever power you you want to dissipate in the wire.

Note that when Rw goes up because it gets hot it will decrease all the values.

Incase you are confused then what I would do is get the nichrome AWG 31 which has a resistance of about 8.523 and one of those 10ohm 12.5W pots.

This means you have approximately what I gave above. If you want to have more power then all you have to do is use a larger wire size like AWG 29... but this will increase the current and might burn up the pot(shouldn't as its about a factor of 2 which I think gives a pot dissipation of about 10W).

In any case that one link seems to give some examples. I'm kinda rambling now ;/

Jon

If all you need to do is cut foam, try a transformer fed by a Variac. You don't need AC

Oops!. I meant to say "You don't need DC". Sorry.

I've cut styrofoam with a piece of .005" (#36) nichrome wire, that reads about 7 ohms per foot. I used a 7.5V, 1A wall-wart, and just held alligator clips in my hands, and it worked astonishingly well.

You might try a simple (read: crude) current regulator circuit:

+13V ------+----------------+ | | | [nichrome wire, ~5 ohms] [10K] | | / c [5K pot] e | | | [1 ohm] | | +13 V RET -+----------------+ But, of course, all of the usual disclaimers apply - I don't guarantee this circuit, it has no protection, etc, etc, etc - I can only say it's worked for me, but I used care while using the circuit.

You'll probably want to heatsink the transistor.

Good Luck! Rich

Thanks for the inputs.

It sounds like all I need to do is take the output from the DC power supply to an LM317 that is controlled by a variable resistor. I could then put a meter on the downside of the LM317 to monitor output voltage.

Could this cause "feedback" to the power supply and damage it?

Not really. It would work but the 317 is rated at a max of 1.5 amps with heat sink. Depending on what current you need it could be to much. The 317 will dissipate whatever voltage you are dropping across it times the current your using. IN any case the 317 is a voltage regulator and it is something you don't need. Even if you used an external transistor its overkill(The DC power supply probably already has the same circuitry in it for regulation anyways).

Once you work out the details you need then you can just use either selectable resistors(Not sure the technical term but you can just use a rotary switch to insert resistors to increase the total resistance) or a power transistor(easiest way I think).

The power transistor essentially just sits between the power supply and the hot wire and you can control the amount of current flowing into it by adjusting the base of the transistor. You'll still need to know the maximum current needed to heat up the wire to around what you'll be using.

Thanks for the replies.

I guess what I need to do is take the 12V DC from the regulated power supply to a small circuit with a LM317 controlled by a variable resistor. That would allow me to put a meter on the output side to monitor the V to the heating wire.

Could this create any kind of "feedback" to the power supply that might damage it?

v/r Jerry

Along these lines he could hang a 1K potentiometer across the PS output, and use its variable output to feed a Darlington emitter follower (connected to a heat sink). With a 2 or 3 W pot, he could go as low as 100 Ohms and get a bit more linearity.

Tam

Please don't use linear voltage regulation; it's inefficient and unnecessary. That would create lots of waste heat. You have a constant load so you could accomplish the same thing simply by having a chunky rheostat or fixed power resistor in series. You want to control a heating element, which simlifies things; you don't need voltage regulation per se, just some way to increase and decrease power. You have a constant load.

I would use pulse width modulation for the power control. PWM control varies the power in your heating element by turning the power source on and off rapidly at a variable duty cycle. It takes time for a heating element to cool down between on/off cycles, so you can use a relativeley low frequency. You can build a fairly simple circuit with a duty cycle adjustable from zero (off) to 100% (continuously on) using a comparator and a few discrete components. I'm a hobbyist and was able to build a pwm heater control based on advice I got on the electronics newsgroups:

Any jelly bean comparator similar to the 2903 would work. I found nichrome wire at Surplus Sales of Nebraska.

I've been having trouble getting things posted, so if this is a double post forgive me.

Please don't use linear voltage regulation; it's inefficient and unnecessary. That would create lots of waste heat. You have a constant load so you could accomplish the same thing simply by having a chunky power resistor in series with the load. You want to control a heating element, which simlifies things; you don't need voltage regulation per se, just some way to increase and decrease power. You have a constant load.

I would use pulse width modulation for the power control. PWM control varies the power in your heating element by turning the power source on and off rapidly at a variable duty cycle. It takes time for a heating element to cool down between on/off cycles, so you can use a relativeley low frequency. You can build a fairly simple circuit with a duty cycle adjustable from zero (off) to 100% (continuously on) using a comparator and a few discrete components. I'm a hobbyist and was able to build a pwm heater control based on advice I got on the electronics newsgroups:

Any jelly bean comparator similar to the 2903 would work. I found nichrome wire at Surplus Sales of Nebraska.

The power per inch dissipated by a hot wire can be adjusted easily by (1) using a different wire diameter (fatter = hotter) or (2) using a different wire length (shorter = hotter)

So if you start with a long wire that doesn't get hot enough, you can adjust by clipwire shunt that bridges part of your hot wire. If the apparatus seems too bulky, buy some thinner wire.

Electronic controls are also possible, with some amusing variants (like using the temperature-coefficient-of-resistance of the wire to make a thermostat). Easiest is just to use a variac (variable autotransformer) on the input of your DC supply (this will work best if the DC supply is unregulated).

Build this, it is no more complicated than any add-on to your existing supply and is tailored for the job:

Depends on what you want to do, but there are inexpensive ones already made with variable power supplies and all the tools required for cutting styrofoam..look at

True, but the idea was to have a little fun building something myself ( and hopefully learn something along the way). I was looking at several different designs but they basically have a bridge rectifier converting AC to DC then add the variable on top. I already have DC so I just wanted to add to it. It seemed the most logical thing was to build a "plug in" type device to put between the DC supply and the wire. The supply I have is a regulated 13V/20A supply. That is a pretty strong current. I didn't want to snap the wire with too much V and As. The wire I'm using is SS.

and now the key bit of information ...

Your heating effect will be I*I*R and by not using nichrome (or other high-resistance) wire you are limiting your options somewhat. One problem with hot-wire foam cutters - and yes, i have successfully built some - is that the foam cutting pulls quite lot of heat off the wire and cools it. So you get a slowing of the cutting action and the cut doesn't remain clean unless you are actually controlling the temperature, which is quite difficult.

For low-resistance wire types, you need to focus on two things.

1. Use as thin a wire as you can, consistent with mechanical strength, to achieve a sensible resistance figure. If the wire becomes ductile at operating temperature, the wire is unsuitable.
2. Use a current-control approach. How complicated that gets will depend on the wire.

12V should be fine with stainless steel (I've done that before), if it gets too hot use a longer piece.

A regulator isn't needed.

That stuff increases in resistance as it heats up so don't be fooled by its cold resistance

Assuming the wire you have is about 0.5mm diameter about 1m is probably a good length for a first experiment, you'll probably want to go shorter in the end.

If you want to regulate the temperature of the wire regulate the voltage.

Bye. Jasen

If you use a wire with a significant temperature coefficient of resistance, you can use the resistance of the wire itself as a sensor to determine temperature. Use a Wheatstone bridge with the heater wire itself serving as the "unknown," or variable resistance. Use a comparator triggering a one-shot so that each time the resistance in the bridge reaches the point of balance, the one-shot will cause a brief lapse in the power applied to the bridge, turning it off for 100 milliseconds or some amount of time which works best as you may determine by experiment.

While the wire remains below a certain temperature power remains on. As the wire gets hotter its resistance increases; at a certain temperature the resistance in the heater will will cause the bridge to balance -- the comparator will trip the one-shot, turning off the power for an instant, just long enough to let the wire cool a few degress. This cycle will repeat itself, and the temperature of the heater wire will cycle in a narrow range.