Hey, I am building 5 Tube Furnances in the near future, is there any interest of getting one, because I dont need 5 Tube Furnances but only one or two
I can say you the price for it later, and also will post the process of making one in this forum. Just want to know if there is general interest in
tube Furnances.
Greetings
AmoledJJay - 10-11-2017 at 10:55
Perhaps. It depends on price, quality, maximum temperature, size, durability, etc. I've built several low-quality tube furnaces. The thought of
building a high-quality one has certainly crossed my mind, but the materials can be expensive and require ceramics skills to work with if you want
high durability and high temperatures.Melgar - 10-11-2017 at 11:07
Sure. Would prefer to trade reagents and/or equipment for one, if possible, as I'm short on cash right now, and have too much chemistry stuff.Amoled - 10-11-2017 at 14:13
Yeah, it will take some time, but theoretically I also would trade equipment.
I will do my best to provide good quality, but I don't know, whats your expactation of good quality and high temperature. JJay - 10-11-2017 at 14:20
1300 C, usable in both vertical and horizontal configurations, capable of regular use for years with minimal maintenance?Melgar - 10-11-2017 at 14:33
I guess it should be able to get pretty close to the maximum safe operating limit for the heating element (kanthal, I'd imagine?), without burning
out. And you should run it for at least a few hours at its high temperature to make sure it all works. I guess as long as the temperature can be
measured with a thermocouple, and the power can be adjusted, that should be fine. You should probably just make a k-type thermocouple built in. I
have two thermocouple readers already.
I don't mind fixing it on occasion, and it should probably not be completely sealed up for that reason, unless absolutely necessary. I probably have
to fix the stuff I made myself more than anything else, so I wouldn't demand higher standards from others.Amoled - 10-11-2017 at 14:39
At the moment it's just planned for horizontal use, but should be possible to change this,
As for now the material that were planned for the furnance would withstand around 1200°C-1300°C, but also this should be no problem, and I should be
able to get materials that are just slightly more costly, but can withstand 1500°C, but another material for the heating coil must be used. The size
would be around 400/450mm
[Edited on 10-11-2017 by Amoled]macckone - 10-11-2017 at 18:37
Quartz tube will work up to 1200C, Kanthal A1 wire up to 1400C, so 1200C max
assuming reasonable loss in transmission into the tube. This combo is actually
cost effective.
For higher temperatures you can go to aluminum oxide 1700C but then
heating stuff becomes more of an issue and you are probably going to want to
go to flame heating. Molybdenum heating elements can go up to 1850C
they are quite pricey and dont wrap around the tube well.
At extra high temperatures you will want to go to arc furnaces.
K type thermocouples also max out around 1200C so planning on
a tube furnace that goes to 1200C is reasonable.
Going to higher temps gets exotic and expensive.violet sin - 10-11-2017 at 19:42
Are you using a commercially available piece for reference? I picked up a desktop vertical model for 80$ used at the pawn shop. Lind
sola-basic. It is a two piece set, oven and controlled box. Quite heavy when you stack one on the other and cary around, but not impossible. I'm
not saying this is pertinent for price, but construction is pretty bulky and industrial feeling. You could use it to beat up a geo metro, and I bet
it would still function.
On the other hand I have seen a bead-makers kiln from an old Stanley toolbox, ceramic fiber mat and a quarts heater element from a room heater.
It got hot and worked for borosilicate.
So it doesn't HAVE to be able to take on a small auto head to head and still be useful. Though with hight temp, more protection/safety would be
most important. This illustrates how having already seen a real model, the knock off may look wimpy. If I hadn't seen the real one, I bet that
toolbox/fiber mat version would have been way more impressive. Home projects versions may be not hold up to experienced users expectations.
I am not discouraging you from building at all, quite the opposite. It's hard to improve your vision if you never make one. Best of luck building,
favorite way of learning.macckone - 10-11-2017 at 20:13
22mm id tubing can be had for around $35
Kanthal wire is pretty cheap $10.
Insulation can be had for $15.
The real expense is the controller and that is something
that is expensive to buy.NEMO-Chemistry - 10-11-2017 at 20:24
22mm id tubing can be had for around $35
Kanthal wire is pretty cheap $10.
Insulation can be had for $15.
The real expense is the controller and that is something
that is expensive to buy.
Can i ask what kind of controller are normally used, and what kind are you thinking of? Ive had an idea for a long time, its from something I did at
school two years ago.
While it didnt involve temperature, I might have a reasonable solution for a controller, no idea how accurate it would be though.
In a nut shell we did some stuff on light and light detection, but it crossed my mind that with 3 different leds, it just might be possible to measure
pretty high temps.
I guess its a bit like a IR method used in IR thermometers but simpler. JJay - 10-11-2017 at 20:50
22mm id tubing can be had for around $35
Kanthal wire is pretty cheap $10.
Insulation can be had for $15.
The real expense is the controller and that is something
that is expensive to buy.
The controller is the easy part
I'd just use an oven dimmer (this one for example, although Sulaiman can surely find a better one: https://www.ebay.com/itm/4000W-AC-110V-120V-SCR-Voltage-Adju...), perhaps with an Arduino with a temperature probe and a relay. macckone - 10-11-2017 at 21:59
Jjay, that will work as long as you don't need close tolerance control.
For close tolerance work, you need a PID controller.JJay - 10-11-2017 at 23:17
I am a human PID controller
But yeah, digital control of the dimmer would be nice....Melgar - 11-11-2017 at 01:27
Just program in some fudge factors with an Arduino, and I bet it'd work as well as any PID. Possibly even better, since you'd be calibrating it to
your specific hardware.NEMO-Chemistry - 11-11-2017 at 04:11
There are loads of Arduino PID libraries around. Chances are you wouldnt be far off.JJay - 11-11-2017 at 06:46
By adding an LCD module to the Arduino, the controller could emit helpful messages such as, "Decrease power to 30%," and "Turn the power on dummy."
You can get some dimmer-like action out of a relay by cycling it on and off, but I like to avoid that as much as possible. I hadn't given much thought
to what would go on the Arduino, but you could use a recurrent perceptron library or something if you really wanted to be hardcore and lazy at the
same time.Amoled - 11-11-2017 at 08:01
I wanted to work with a PID-Controller, coz I never really worked with an Arduino, only with a Raspberry Pi, and an Arduino seems to be somehow
completly different, I think I should look into all this Arduino things, because all these project I've heard of, you could possibly do, sound very
interesting.Melgar - 11-11-2017 at 09:42
You can get some dimmer-like action out of a relay by cycling it on and off, but I like to avoid that as much as possible. I hadn't given much thought
to what would go on the Arduino, but you could use a recurrent perceptron library or something if you really wanted to be hardcore and lazy at the
same time.
Lots of circuits do that, except they use MOSFETs instead, because they don't have any moving parts. The main advantage of relays is being able to
isolate power sources, and also to easily switch AC.
I've had to take multiple undergraduate and graduate classes on control systems, and they're often quite sophisticated. For example, your
thermocouple is presumably outside of the tube and your object in it, right? Say you want to program it to heat the object to a certain temperature.
Most heating devices would heat until the thermocouple read that temperature, then maintain that temperature. But it could take a lot longer for your
object to reach that temperature. So you could initially overshoot the temperature based on the size and material of the object (entered at the
beginning) and then stop heating at a certain point such that the object's temperature will be rising as the tube's is falling. So you reach your
target temperature much faster that way.
Another way you can do that is mechanically, so that your thermocouple is an adjustable distance from the tube. That way the thermocouple temperature
more closely matches the inside of the tube. Less control, but a lot easier.
But Arduinos could also have programs to, say, gradually increase temperature over an hour, hold it there for three hours, then gradually decrease it
back to room temperature over the course of an hour. Or cycle temperatures, even. Or program it to turn off if it detects flammable gases. Or hook
up an outlet, so you can plug something into the the outlet, that the Arduino can turn on and off. Say a fan, or an ultrasonic wave generator.JJay - 11-11-2017 at 11:07
I kind of assumed solid state relay (I think they are based on MOSFETs, but they might as well contain powdered goblins for all I know). I don't like
them cycling on and off because of the uneven heating and erratic magnetic fields. I am a competent programmer, and I personally would not use a
recurrent neural network to address temperature changes, but that's one way to do it....Melgar - 11-11-2017 at 11:28
I don't like them cycling on and off because of the uneven heating and erratic magnetic fields.
You really have to do it that way though, or with a PWM system. Reason being, a component doesn't generate much waste heat in either the
totally-closed or totally-open states, but if you try and partially close the component... well, now it's a resistor and is wasting a lot of heat and
probably burning out.JJay - 11-11-2017 at 11:35
Pretty much... I mean, you could probably use a DAC with op-amps and diodes, but I'm not about to attempt to put together a circuit like that given my
level of experience.NEMO-Chemistry - 11-11-2017 at 16:31
Triacs are used with AC, pref one with zero point crossing detection, that way you can safely switch upto 50 or 60 times a second (depending on
country).
I got a circuit for a pic micro with a decent triac switching circuit, its no big deal to swap out the pic for Arduino.macckone - 12-11-2017 at 13:43
Buying an off the shelf PID controller would be ideal if you have unlimited funds.
However an arduino, k-type thermocouple, and a triac with a zero crossing detector could be cost effective.
k-type thermocouple - $10
arduino - $25
call it another $15 for other parts
That is $50.
That puts the whole project at about $110.
Of course scrounging and getting some of this stuff for free is possible.yobbo II - 13-11-2017 at 13:31
That PID is really good price but the probe is a rip off.
The probe cost is already covered under the shipping, which is outrageous.wg48 - 14-11-2017 at 03:20
That PID is really good price but the probe is a rip off.
The probe cost is already covered under the shipping, which is outrageous.
Yes excellent price for the controller as it comes with a 40A SSR and heat sink. The heat sink looks small for a 40A SSR probably fine at 13A.
However judging by the suffix on the model number (from the pic) it’s a reverse controller which I think means when the temperature is low the
output is off/low. It needs to be high to drive that SSR.
The description for the controller states "the detective temperature range is 0- 400C" if correct its low for a tube
furnace.
Check the full model number against the spec for no surprises.Fulmen - 14-11-2017 at 04:20
Triacs are used with AC, pref one with zero point crossing detection, that way you can safely switch upto 50 or 60 times a second (depending on
country).
I got a circuit for a pic micro with a decent triac switching circuit, its no big deal to swap out the pic for Arduino.
It’s probably not worth the effort to construct a PID controller given the cost of the "off the shelf ones".
What we need is a ramp and soak PID with burst fire output. So you can ramp slowly to the soak temperature. The hard on until your almost at the
required temperature is tough on the furnace insulation and heating elements. Now that type of controller costs £100 and up. Ideally it would
interface to a cheap optical pyrometer or possibly a tungsten resistance temperature sensor ie a tungsten quarts tube lamp (my idea for a cheap
temperature sensor to say at least 1300C I hope)Fulmen - 14-11-2017 at 08:24
Agreed. If you can build a cheaper/better ramp-soak controller that's one thing, but trying to compete with commercial PID's are pointless.yobbo II - 14-11-2017 at 17:33
The probe cost is cheap (very) imo. I f you can find a Pt/Rh tc for anything like tho cost I would love to hear about it. Is it a good quality probe
though?
Some wire here, an outrageous price. https://www.ebay.com/itm/Pt-Pt-Rh-wires-for-thermocouple/112...
Pt100 probes are not tc's btw, but a platinum resistor. No good for tube furnace.
I love the W filimentt probe. Sounds great!
Molybdenium elements are no good in air.macckone - 14-11-2017 at 20:57
A K-type thermocouple will be just as good and much cheaper, If you
want to go above 1200C you are going to have to go with molybdenum
elements or something similarly exotic unless you go with direct flame.
And at that point you are also getting an exotic tube material, exotic
insulators and then the Pt/Pt-Rh makes sense. But you can get bare
wire way cheaper than that directly from omega engineering. Which
given the application may be a better choice than a sheathed probe.
RTD like PT100 or tungsten filament work fine for tube furnaces,
provided they have the correct substrate. Most of the PT100 are only
rated at 600C. yobbo II - 15-11-2017 at 09:34
Afaik pt resistance devices only go to 900c or so http://www.ni.com/tutorial/7115/en/
I may be wrong. Anyone got a source of reasonably priced pt100 type devices that go to 1200 or so?
When you say molybdenium i think you are referring to molybdenium disilicide.
Molybdenum is hopeless in air and more or less fails completely around 800c.
Molybdenum
This material was first used in industrial furnaces around 1930, with increasing usage
after 1940. It is available in a wide range of forms, with the most common element shapes
employed being wire, rod, strip, and on occasion, tubes.
This material has a strong affinity for oxygen at elevated temperatures and can only
be heated in the presence of a vacuum, reducing (dry hydrogen or cracked ammonia), or
pure inert atmosphere. Molybdenum starts to oxidize between 250 to 300°C with the
formation of molybdenum dioxide (MoO2) that will offer a limited amount of protection
against further oxidation up to around 600°C. At this point the molybdenum dioxide
converts to molybdenum trioxide (MoO3), which becomes dominant.
Molybdenum trioxide is very volatile and readily boils off, exposing the base metal
to further oxidation. At around 800°C catastrophic oxidation occurs with clouds of
molybdenum trioxide being generated. With respect to atmosphere purity, an argon...
