I'm going over all the extruder design information as I prepare to build mine (as I watch glue dry on the unfinished cartbot...), and one thing that doesn't quite make sense is that the nichrome wire's ends are sticking out of the electrically insulating heat conductor (in my case and in the standard, JB-Weld high temperature epoxy) (picture [reprap.org] and page [reprap.org]).

The ends sticking out become hot during operation: "The short pieces of nichrome get hot in operation, so keep them flying free (that is, don't fold them tight against the barrel or the PTFE)." (That's from the same page.)

So, why can't we just solder the ends of the nichrome wire to the power/ground wires (ends of ribbon cable in the standard) before JB-Welding it up - then wrapping the nichrome wire all around the barrel, and then having the attached power/ground wires partially inside the JB-Weld and sticking out? It doesn't make sense to me to have 40mm (two 20mm lengths on each side) of the heater wire heating the air instead of the barrel...

I don't think it's really stated anywhere, but I'm guessing the reason this isn't done is probably that the solder might melt, move, and leave an open circuit. This also doesn't quite make sense to me, as, if you apply the JB-Weld around the thing and let it set, wouldn't it form a barrier? And so wouldn't the liquid solder have nowhere to go, and thus remain in place?

Or would the liquid seep through pores or something in the JB-Weld, perhaps shorting the nichrome before forming an open circuit?

Or is solder perhaps not as electrically conductive when liquid?

Quote

It doesn't make sense to me to have 40mm (two 20mm lengths on each side) of the heater wire heating the air instead of the barrel

Does it matter that much? I guess it isn't very neat, but I don't imagine it has any negative impact aside from the small amount of power it wastes.

Quote

the solder might melt

Will melt, I reckon. But that wouldn't stop you crimping the two bits together and then insulating them if you felt the need.

Quote

if you apply the JB-Weld around the thing and let it set

Remember that the JB weld will perish. It would suck to have a big build fail because the heater wires popped loose. That's an argument against soldering, anyway.

Ru Wrote:
-------------------------------------------------------
> Remember that the JB weld will perish. It would
> suck to have a big build fail because the heater
> wires popped loose. That's an argument against
> soldering, anyway.

What do you mean it will perish? The packaging on the stuff I have says it's maximum temperature is 600

Joshua Merchant Wrote:
-------------------------------------------------------
> Ru Wrote:
> --------------------------------------------------
> -----
>
> Do you mean the nichrome will
> get hotter than that when heating the barrel to
> 240

About how hot do you think it will get, and why doesn't it conduct well enough to the barrel to maintain an approximately equal temperature?

Quote

Do you mean the nichrome will get hotter than that when heating the barrel to 240

Joshua Merchant Wrote:
-------------------------------------------------------
> About how hot do you think it will get, and why
> doesn't it conduct well enough to the barrel to
> maintain an approximately equal temperature?
>
LOL! What part of epoxy is a pretty good thermal insulator don't you understand? To drive energy through a thermal insulator you've got to have an impressive delta T. I wouldn't be particularly surprised if your thermostat was having to drive the actual nichrome temperature to nearly red heat (~1,400 F, iirc) to get enough heat into the extruder barrel to melt the plastic.

Forrest Higgs Wrote:
-------------------------------------------------------
> Joshua Merchant Wrote:
> --------------------------------------------------
> -----
> > About how hot do you think it will get, and why
> > doesn't it conduct well enough to the barrel to
> > maintain an approximately equal temperature?
> >
> LOL! What part of epoxy is a pretty good thermal
> insulator don't you understand?
Actually, I missed that part entirely.

> To drive energy through a thermal insulator you've
> got to have an impressive delta T. I wouldn't be
> particularly surprised if your thermostat was
> having to drive the actual nichrome temperature
> to nearly red heat (~1,400 F, iirc) to get
> enough heat into the extruder barrel to melt the
> plastic.

So... why use a thermal insulator between the nichrome and the barrel?

Quote

So... why use a thermal insulator between the nichrome and the barrel?

Probably due to the difficulty of finding a good thermal conductor which is also an electrical insulator which can withstand that much heat. Something has to keep the heating element in contact with the barrel, after all, and I don't believe that the nichrome is electrically insulated?

Joshua Merchant Wrote:
-------------------------------------------------------
>
> So... why use a thermal insulator between the
> nichrome and the barrel?
>
Because the number of materials that have very good thermal conductivity and very poor electrical conductivity approaches zero.

Ceramic?
[en.wikipedia.org]

Wait a minute, I'm using an aluminum barrel... why can't we (I! I mean I, really!) just anodize it? [en.wikipedia.org] says anodising can be used to increase the thickness of the natural oxide layer (which in aluminum is normally very thin).
We'll have a dielectric coating between the aluminum barrel and the nichrome wire which also happens to have a high thermal conductivity (well, at least for a dielectric: 40 W/m K compared to aluminum metal's 237 W/m K; I couldn't find the thermal conductivity of JB Weld for a comparison - anyone got an estimate?).

So, anodize the aluminum barrel, then drill the hole in it (so the interior has only the standard thin film of aluminum oxide), then wrap the nichrome wire around it (with my soldered wire!), then apply epoxy around it to keep it in place (though I suppose you could use something else), and then heat the nichrome up to, say... 300

Edited 1 time(s). Last edit at 09/06/2008 12:42PM by Joshua Merchant.

Quote

why can't we (I! I mean I, really!) just anodize it?

Sounds like a reasonable ida. I think you should try it. Let us know how it goes

Quote

and then heat the nichrome up to, say... 300

Ain't no way I'm going to be wrapping nichrome around an aluminum barrel, anodized or not.

Aluminum oxide, aka Alumina, is the predominate ingredient of most of the high temperature adhesives (Cerastil, Resbond).
If you can figure a way to oxidize a thick enough layer onto an aluminum barrel it might work. Make sure to use fuses if you experiment with this!

Also, solder will not stick to nichrome, that's why you need to leave a bit out to crimp on to.

It all seems like a lot of hassle when you can use a number of different materials Forrest (iirc Resbond) Nophead (Cerastil) and I use fire cement/furnace cement. I think these are all rated into the +1000

---

Ian
[www.bitsfrombytes.com]

Forrest Higgs Wrote:
-------------------------------------------------------
> Ain't no way I'm going to be wrapping nichrome
> around an aluminum barrel, anodized or not.

Uhm... why not? Type II anodized aluminum (which I will probably be using) has a nonconductive surface (according to a few non-wikipedia sources found via google).

Of course, I would take a multimeter and thoroughly test the barrel before wrapping it with the nichrome, just in case, but it really shouldn't be a problem.

The way I'm planning on doing it (Type II, no dye, hot deionized water-sealed [actually I might use distilled water if it works - I'm just like that]), it should have a well-sealed, moderately thick non-conductive surface. I suppose if you grab the screw with a pair of pliers and whack it against a steel table, you might take off a bit of the coating (Type III hard coating could prevent this by making the oxide layer thicker, but it is quite a bit more of a hassle), but I really don't think this should be a problem... I mean unless you have some really whacky mental disorder that involves sudden, uncontrolled limb movements.

SnailRacer Wrote:
-------------------------------------------------------
> Aluminum oxide, aka Alumina, is the predominate
> ingredient of most of the high temperature
> adhesives (Cerastil, Resbond).
> If you can figure a way to oxidize a thick enough
> layer onto an aluminum barrel it might work.
It's a simple matter of building an electrolytic cell with the aluminum barrel as the anode and sulfuric acid as the electrolyte. Aside from safety considerations (for handling and storing sulfuric acid), it's high school level chemistry.
> Make sure to use fuses if you experiment with this!
Why? So it doesn't burn out my PWM driver in case there is a short or something?

> Also, solder will not stick to nichrome, that's
> why you need to leave a bit out to crimp on to.
Yeah... I figure I can just do that in close proximity to the barrel and inside the insulation covering, instead of out 20mm away from the barrel floating in the air. I think it would be safer this way if it works. If all else fails, I think I can wrap the wire around the nichrome, put a blob of solder to form an electrical connection, and then cover it with a non-porous (if necessary) material that can stand the heat (to contain the blob of solder when it goes liquid). This split electrical and mechanical joint (on both ends of the nichrome) will then be wrapped with the rest of the nichrome around the barrel and held in place (if the same material is used to hold the nichrome in place as was used to hold the solder in place, it can probably all be done at once).

Ian Adkins Wrote:
-------------------------------------------------------
> It all seems like a lot of hassle when you can use
> a number of different materials Forrest (iirc
> Resbond) Nophead (Cerastil) and I use fire
> cement/furnace cement. I think these are all rated
> into the +1000

Quote

I think I can wrap the wire around the nichrome, put a blob of solder to form an electrical connection

Just crimp the condutor with the heater. Twist em up a bit and squish with some pliers. You could always use a fold of some suitable metal crushed on top if you were worried about it coming apart.

Quote

I think with my method it could be cheaper than cerastil and more efficient than any of the current options.

But how much more efficient? How much power does the heater draw now, for example? I'd be surprised if it drew more than a few amps as 12V, and it isn't like it is running on a 100% duty cycle. Is a small saving really worth it? I guess any saving is a bonus if we ever want to run this thing from a battery, though.

Interestingly, there is an epoxy with boron nitride and aluminum oxide [www.arcticsilver.com], which is used for... well, I don't know, but the company makes heatsink pastes (they say that this specific epoxy "IS NOT intended to be used between a CPU and the CPU heatsink"). Boron nitride is a dielectric with a higher thermal conductivity than aluminum oxide (AFAICT) which can be added to ceramics (it can also be added to our thermoplastics to improve properties [possibly also reducing warping], see [en.wikipedia.org]).

Anyway, I think that such a mixture might be a suitable replacement for JB-Weld on an aluminum heater barrel.
They also sell a less permanent thermal compound [www.arcticsilver.com] which might be used, say, to cover the outer portion of the barrel (after wrapping the nichrome), to allow a semi-permanent version of the barrel (i.e. the nichrome could be removed). I don't know whether this last part would work (I think the compound might drift down to the tip of the barrel).

On a similar note, a home-made mixture of sodium silicate (water glass) and aluminum oxide in powdered form was proposed a bit over a week ago [forums.reprap.org][at the bottom] (though the idea is, of course, not original [not that it matters]). Has anyone made any progress with this or thought of something to add?


I wonder if it would be feasible to make the entire heater barrel out of alumina (via slip-casting or some other ceramic-forming technique)?

Would it be able to conduct enough heat from the nichrome to the thermoplastic? The thermal conductivity is about 1/6 that of aluminum metal, but perhaps if I reduced the wall thickness (the amount of material between the nichrome and thermoplastic, by, say, 6 times) it would work.

Perhaps I could try slip casting the inner barrel, which has the 3mm inner diameter, wrapping it with the wire, then casting an outer barrel around it, and then forming the insulator around the outer barrel.

It might be best to have a ceramic insulation around this barrel, using a different material than the alumina (perhaps one with a much lower thermal conductivity, but a similar firing temperature, etc). (Not that there's anything wrong with just wrapping it with fiberglass insulation... or is there?)


Alternately, instead of making the entire inner barrel out of alumina, have an inner barrel of aluminum, cast a thin middle layer of alumina around it, wrap the nichrome, then cast the outer alumina layer, then do the insulation.
This would basically be the same thing as anodizing, except creating a thicker layer of alumina (more of both types of insulation, electrical and thermal, which is good and bad, respectively), and trading the problems of an electrolytic cell (with a sulfuric acid electrolyte) for the problems of casting and firing ceramics. Although, if you're going to cover the outside of the nichrome with cast alumina anyway, you'd already be dealing with the ceramic problems (but you could just anodize and then cover the outside of the nichrome with something else, like a thermal insulator like JB-Weld, though this might sneak in between the nichrome and the alumina layer and reduce the thermal connection).



While I'd love to test all these possibilities (I lost count), I suspect I only have the time and focus to test one (maybe two if the first fails).

Joshua Merchant Wrote:
-------------------------------------------------------
> Wait a minute, I'm using an aluminum barrel... why
> can't we (I! I mean I, really!) just anodize it?
> [en.wikipedia.org] says
> anodising can be used to increase the thickness of
> the natural oxide layer (which in aluminum is
> normally very thin).




Joshua,

Apparently anodizing is doable in a home-shop environment, though care seems indicated -- Hydrogen gas, acids, etc. (Esp. keeping sparks away from the H2!) Apparently a fairly wide variety of electrolytes can be used (including Coca-cola as a source of phosphoric acid -- no s__t, google it!)
To grow a thick (in anodizing terms) oxide layer, keeping the setup cold sounds important, as is keeping the solution acidic.
An outside container (bucket) with ice sounds sufficient for the temperature issues.

According to
[electrochem.cwru.edu]

a common electrolyte for Al anodizing is
dilute Sulphuric acid. So one could buy battery acid....
Myself, I'd like to avoid that, if possible. A possible substitute is Sparex (#2), whose active ingredient is Sodium BiSulphate. This is apparently available as "Ph Decreaser" from pool supply stored. I was just quoted $9 US for 22 Oz., which should last my likely anodizing career. (Bi here doesn't mean two, it means Hydrogen (a bit of chemistry's historical/obscure nomenclature) -- so this is a mix of Sodium Sulphate and Sulphuric acid, but crystalizes out as a solid. Still acidic; I'll have my gloves/safety googles on if I get to this stage.

Some articles on home-shop anodizing recommend using Lead for the cathode. Um, no thanks for me. Apparently Al, stainless, steel, or another piece of Al can be used. And since we're only interested in anodizing a small area, I may try Al foil or misc scrap Al.

-- Larry

Joshua, et al,

I wish I'd pondered a bit more before my msg (above) re anodizing.
Although anodizing can probably be made to work, here's another idea that (if it works) doesn't require an Aluminum barrel, or mucking about with nasty chemicals.

A thin MICA insulator between the nichrome and the barrel might do the trick, if we can get mica thin enough not to crack, when bent to the barrel's outer diameter. These are commonly sold as electral insulators (and simultaneously thermal conductors), for cases where 2+ transistors need to share a common heatsink. We want exactly this combination of characteristics, and mica is apparently good up to ~1100 degrees F.

I haven't gotten any to try yet.
The best source I've found (in the US) is $0.15 each at mouser.com
Mouser Part #: 534-4661
Manufacturer Part #: 4661
Manufacturer: Keystone Electronics
Description: Hardware MICA INSULATOR TO-3

-- on my list for the next time I order from them.
I *wish* I could get the blanks before they punch holes in them, but I haven't been able to find thin mica sheet in any other form in small quantities.

Anybody have ideas for sources of suitable mica?

-- Larry

Hi Larry,

i have some very thin MICA-sheets which are normally used as isolator for high-power-transistors - you can bend them to fit on the surface of a coffee-mug, but not much more or they will break.

Modern much more elastic heatsink-insulators are made from silicone with some embedded fabric - maybe very thin mineral-fibres.

Maybe you can use glass-fabric (as used with epoxy for building 3D-housings for model-planes): wrap a sheet around the barrel, wrap the nichrome-wire around, cover with a second sheet of glass-fabric and soak with waterglass. When the waterglass is crystalized and tempered, it's heat-resiatant above 1000

why go to so much trouble of anodizing it yourself when mcmaster-carr will sell you a 6' section of anodized tubing with an inner diameter of .120" for less than $7.00
.120" id 6' L anodized aluminum tube