That is all well and good, but:
1- I started powering it with minimal power and slowly turning it up, I would think the shortest one would glow first and as I turned it up it would glow brighter and the 2nd shortest would start.
2- The wire was cooled as it was in an airflow of a wind tunnel.
3 - A single wire worked perfectly, I was able to adjust the glow by turning the knob, and if I wanted to overpower it and pop it.
That makes sense. Nichrome’s resistance increases with increasing temperature. So when you apply a small voltage, majority of the current is going through the shortest one. But as the shortest one heats up, the resistance increases and thereby now the second shortest one starts carrying more current (makes sense ?)
Its a boundary layer problem, but wires do not cool much with air flowing. We can start a different thread for heat transfer in flowing air, but suffice it to know that its not that great. This is the same reason that dirt builds up on ceiling fan blades. You would think that the air flowing around them would keep blades clean, but there is a boundary layer where air is “still” around the blades and the nichrome wire.
It of course will. You didn’t have the complexity of temperature changes effecting the resistance and hence the current flow in other wires.
Like I said, this is completely expected, since one wire isn’t multiplying the current.
Consider two lengths of wire, one 4x the length of the other. Due to P=\frac{V^2}{R}, the short one will consume 4x the power as the long. But because it’s shorter, it actually consumes 16x the power per unit length. So it’s very easy for the long wires in this setup to be barely warm while the short ones vaporize.
I agree with Sam_Stone that it would have shown a cascading failure, but I don’t think his explanation of the reason is really complete. If the power supply really did supply a fixed voltage, the vaporization of the shorter wires wouldn’t have an effect. However, most likely there is a great deal of internal resistance in the system, from the steel rods to the connectors and so on. So when the short wires vaporized, that reduced the current in the system, which reduced the voltage drop… therefore increasing the voltage seen by the remaining wires, pushing the next wire over the threshold, and so on.
I don’t understand why the lowest wire wouldn’t glow at all, though. It will have the most current flowing through it, so you should have been able to get it to glow before the others did. But your description says it did nothing until you turned up the current, then everything failed.
Unless… The power supply goes from 0 current to somewhat more than the small wire can handle when you start to increase the supply current. Or, there was something about the supply that couldn’t properly deal with the lower resistance of all the strands in parallel. Many power supplies require a minimum load to function correctly (especially switching supplies), and can be unstable if the resistive load is too small. Nichrome wire in small lengths is very low resistance, and if you had a whole bunch in parallel you could have presented the power supply with a below-minimum resistive load. If so, a small resister of the correct power rating in series with the device would fix the problem.
What was the length ratio between the shortest wire and the longest one?
Something along those lines seems likely, but the OP doesn’t seem to recall the details’ Not sure if “old style light dimmer” is referring to a very old resistive type, or the somewhat newer but still old triac type. Or maybe something else entirely, like a Variac.
Shortest was about 14 inches and the largest was about 16.
Maybe this has something to do with it but along with something in the system providing inductance, perhaps in the power supply or whatever the device was used to adjust/regulate voltage.
When the shortest wire heated up to a point before it was glowing, it’s resistance increased, perhaps faster then the current could diminish due to inductance in the system, this would cause a increase of voltage over all the wires, which in turn would increase resistance but could not lower current fast enough then the inductance would allow. Eventually the voltage was high enough across the system to overload them. Does this make sense?
That’s where I was slowly heading with this.
My comment about an old style dimmer, it was basically a potentiometer. I have variacs also, this was just a cheap controller I could leave wired to into a foam cutter. If he used a triac dimmer decades ago it would have been meant for incandescent or fluorescent lights, not sure what it would do with that circuit.
That doesn’t seem like enough of a difference to account for the problem, then. If there were no other design problems, I’d expect you could dial things in so that the long wire just barely glowed while the shorter wire was a bit more orange. Maybe the power supply is the culprit. You don’t remember any more details there?
Yes. For most power supplies, there is something called the Maximum Power Transfer Theorem (MPT). The maximum power supplied by a system is achieved when the internal resistance is equal to the external resistance.
So the power in the external circuit (the nichrome wires) will keep going up as you reduce the resistance but it will reach a maximum and from then on, reducing the resistance will not increase the power.
WHen you had the one wire setup that worked, did you also have loops of nichrome around each steel rod? And was it multiple strands of that wire?
How many total strands would you say you had on the one that failed? Per leg, and total?
16" of Nichrome 60 in 36 ga requires 22.7 V to get to 800 degrees, at which point it will draw about .6A. Multipy the voltage requirement by the number of wires per leg. If it was 10 wires, you’d need 227 V. If you had three legs, it would still be 227V but the current draw would go up to 1.8A.
The difference in current between 14 in and 16 in would only increase the temperature of the shorter wire by 100 degrees. Even with 10 loops, the temperature of the lower one would only go up 130 degrees. So length differences are likely not the issue here.
Sure, more wires means more current, but that’s not a problem. It’s not the total current that causes the wires to melt; it’s the current in each wire. Which should be the same regardless of whether the other wires were present or not. Or if there was any effect at all from the other wires, it should be in the opposite direction: If the power supply wasn’t up to providing that much power, then putting in more wires should drag the voltage down, meaning even less heating in each individual wire.
Give me all the details you can remember. I found my foam cutter supplies, I have plenty of nichrome wire in gauges from 18-21 and several power supplies. I can assemble this in the next couple of days and try it out.
Here’s how I think you should have done it, to achieve your goal of a temp gradient:
Use the 2 steel rods, but mount them straight vertical, parallel to each other. Have several pieces of the nichrome wire, all the same length. Place the first one near the bottom, wrapped around the rods at either end. Then a few inches above that, place 2 wires across, close together (but not touching), each wrapped around the rods. A few inches above that, place 3 nichrome wires. And so forth. All wires the same length, same resistance, and thus each absorbing equal shares of the current and dissipating it as heat.
When finished, you will have a device with 1 heating element (wire) at the bottom, and more heating elements, closer together as it goes toward the top. Which should result in your desired temperature gradient in the airflow.
This. Please stop confusing the discussion with misinformation @MrDibble.
You are I suspect getting confused by thinking that because more wires in parallel reduces the resistance (and hence increases the current) across the whole circuit, it somehow decreases the resistance (and increases the current) across each wire. This is patently wrong. The resistance across each individual conductor does not change - and so for a given voltage the current does not change - merely because conductor is/is not in parallel with another conductor.
No, I don’t think this. But I did misread the OP, and thought that it was just the whole circuit that blew, not each individual wire. So it’s not that simple, but I still think sticking them all in parallel is a factor.
I would suspect that the problem lies in the power source. In order to vaporise the nichrome wires the voltage needs to rise to something significantly greater than intended. For a reasonable range of voltages the nichrome wire will provide a negative feedback effect, rising in resistance and lowering the overall amount of power being dissipated. Clearly the voltage delivered is more than just a small jump over the amount desired for operation.
With a perfect power supply everything should work fine. A constant voltage source will deliver whatever current each leg of the device draws, and there should be no failures. In reality all power supplies have some internal resistance. Regulated supplies have effectively a very small source resistance, mostly due to the resistance of the leads.
Laboratory power supplies are usually notionally constant (if variable) voltage, and many have settable current limiting. Such a device should cause no problems.
Other power supply possibilities include a rheostat, or Variac. (A potential divider isn’t going to be useful.)
A rheostat is just a series resistor, the more parallel nichrome wires there are and the greater the voltage drop across the rheostat will be, requiring it to be adjusted for each setup. But add more wires, and for the same setting, there will be lower voltage on the output terminal of the rheostat, not more.
A Variac has a mix of reasons for variation of output voltage, but again, the lower the resistance of the nichrome wire setup, the lower the output voltage will be for a given setting due to internal losses expressing themselves as a series resistance.
One out-there idea is if a Variac was used, and the input current limit was reached, if the Variac had a thermal cut-out on the input, it would open circuit the Variac suddenly, which might lead to a significant inductive spike being delivered to the apparatus. However I’m far from convinced that there would be enough energy in the system to vaporise the wires. Such a scenario is a bit of a stretch, but not impossible. I would also be really worried that anyone would use a Variac in such an experiment. They are not designed for such use, and are intrinsically dangerous devices, being quite capable of electrocuting the unwary.
A cascading failure brought about from a poorly behaving rheostat, one with jumpy behaviour may well be the problem. Set at one resistance, good for a single wire, but unable to manage a smooth transition to the lower resistance needed for a set of parallel wires. So you turn it on, then up, and then - bang - it jumps down in resistance to something fatal and the whole thing blows.
All I ever used is a single wire that I started by wrapping it on one rod till it was: 1: at the proper height, 2: tight enough that I would not expect it to move/fall down the rod and that it had a good amount of contact with the steel. Then I sent it over to the second rod and wrapped it till it was ‘tight’ enough again. For a single wire run I was done at this point. If it were a multiple wire run attempt I would wrap it till I was at the proper height for the wire to return back to the first rod. This process was repeated till the array was complete.
Though I did try a double strand of wire twisted around each other, I basically stayed with single wires. IIRC the twisted wire broke sooner than the single wire and the twisted wire I think heated unevenly. Whatever the case the single wire seemed to work better.
The total array tried was about 7 wires about 3/4 in apart, though I did do variations of it, including a 2 wire array that also popped.
This is incorrect, the array is in parallel, the current would increase the voltage would be about the same (a bit different due to the V shape).
This is awesome and thanks. It really is a very simple array, which was to simply get a temperature gradient in a stream of moving air. The moving air part is not required though a fan could be used if desired.
Again it is a \=/ shape where \ is a steel rod and / is the other steel rod. Those rods were inserted into the wind tunnel frame to hold them in a V shape (I drilled holes in the tunnel walls for this). But basically anything non-conductive that holds them in the \=/ shape would do. NiCh wire was wrapped around them as described in post 38. Alligator clips were attached to the top of the steel rods. At this point the array itself is complete.
The power supply is what I’m not sure of and drawing a blank, It very well may have been a old style dimmer for a wall switch, current could have been AC or DC, but provide enough voltage to be able to vary the brightness of a glowing NiCh wire from none to bright to popped. It also may have been a laboratory style supply. Again I have a blank on that part.
IIRC I did try this or a variation of it, I remember redrilling the holes so that the rods were parallel and the spacing of the array got tighter as it went up. Though it wasn’t perfect, as the rods were thin enough to bend slightly as I needed the gradient in the lower portion of the wind tunnel. It did not work and I was already thinking how to move on and get the experiment done in a different way. So yeah it popped too.