Heat Treatment Oven

There would not be enough infomration in your question. One factor to consider is the resulution of your sampling, this may require 8, 16 or 32 bit processing. In this case only a limited number of Arduino dedicated Emmbedded processors maybe required.

Will you use sensor that operate under Linux or via I2C or Network based, as these are about the only way you can you can use Raspberry Pi.

Thank you. The tech stuff is a bit over my head sorry. Basicaly what i want is a insulated airtight box with heating ellements and a gas inlet and out let. I want to be able to controll the temp automaticaly with a thermocouple and have it ramp up or down the temp as orderd by a controller. Say sit at 700 c for 30 mins incress to 900 c for 2 hours then feduce down to 700 c for 30 mins. Plus i would like the controler readout to be able to displaythe lpm of the gass flow. Ohh and i only have a pc computer running windows

Except in really special applications, I recommend getting a high-quality temperature controller from (off of?) EBAY.  Use one with an external SSR, so you can scale to your load.  Thermocouples aren't really a good place for noobs, my opinion, nor is PID.  Similarly, I'd buy a furnace and modify it rather than scratch build, probably.  The item you need to switch your gas on is called a solenoid-operated valve.  These will be relatively small, so they 'desire' a hard-landing.  Very large valves sometimes need a soft-landing, a discussion probably outside of  our scope here.  You need to shunt your solenoids with a reversed diode.  The current rating (plus design margin) of the diode is that of your solenoid.  The point of this is to deal with the EMF created by the collapsing magnetic field of the solenoid.  You want to place this and a capacitor (in shunt w/ load) with about an (decimal) order-of-magnitude of energy storage relative to the solenoid inductance proximately as possible to the device they protect.  The diode should be fast.  The capacitor should be low-ESR.

Over-rate your drive Qs and isolate them from your logic supply.

I would have to agree about buying the ready made kiln and then alter it.

I think this project is not soley for noobs, You are dealing with not only making high but also designing it.

If I was going to make this project I would use something different to a arduino but it would. A Raspberry Pi wouldn't cut it as you require analog sensing which Raspberry Pi does not have.

In designing you will need to source what sensing component you are going to use to sense the temperature, the Gas Flow. Forom there consider which development board to use.

have you thought of using a psoc5 board, all the add ons are click and drag, to build and program it.

might make life a bit easier.

go to psoc5 tutorials.

ckeck out this link.

some are connecting psoc5 up to arduino sheilds.

Overview, week 7

Here is a link for thermocouple boards:

Reference Designs: Thermocouple Temperature Measurement | Analog Devices

This is a better start for a relative novice.  Unless phrases like 'Seebeck Effect' and 'Cold-Junction-Compensation' come trippingly off of your tongue, it would be good to start here.

==========

Little typo in my prior post:  You can put a small capacitor in shunt with the solenoid.  This makes a nice first responder before the diode changes state.  If you keep the energy of this capacitor down to a few percent of that stored in the solenoid, or watch what happens on a triggerable scope, you're good.  Use MMLC

What I was trying to say is that one needs to put bulk capacitance across the solenoid supply.  Use Al e-lytic.  This should have 10-100X the energy of the solenoid.  It needs to be physically and electrically proximate to the anti-spike diode and the coil.  We do not want to create an un-intended flyback supply!  Electrically proximate means low-Z.

Wide, flat, short trace, use two-ounce Cu material if possible.  Prototyping, returns made of solder braid can be pretty fast, the weave minimizes skin-effect, the Cu is thick and pure.

IIRC energy stored in a capacitor (condenser, localized concentration of the charge field)  =  VVC/2

       energy stored in an inductor (coil, localized concentration of the magnetic field)         = IIL/2

==========

where V is the potential drop across the device in Volts.

          I is the current through the device in Amperes.

IMHO, I would NOT put a capacitor in parallel with the coil of a solenoid.

You normally switch the solenoid with a transistor (or mosfet) rated for say 2-10 times the current the solenoid uses. Adding the capacitor will increase the current at the instant of turn-on by enormous amounts.

So you're using the transistor outside its recommended operating range.

If the diode is too slow, the voltage will rise until the transistor starts to conduct again.  Nowadays that is an allowed operating mode (for mosfets) that is allowed under circumstances. When the amount of energy stored in the coil is less than what the transistor is allowed to dissipate in this mode, you could forget the diode completely. Using the diode will probably, even if it is slow, reduce the amount of energy the transistor has to cope with by a huge amount.

So... My advice is: Use a diode, don't worry about the possible spike that occurs before the diode kicks in. (I did some measurements last Friday and the scope couldn't see the spikes due to the slowness of the diode. )

(I tried blowing up the transistor by giving small pulses to the solenoid, that someone claimed blew up his transistor. No such luck... The transistor held up.)

Don, the formulas, although written awkwardly here on the forum (couldn't get the forum to do it better myself), are correct. But how do you figure out the inductance (L) of a relay. I could add a current-sensing resistor and watch the current increase over time when suddenly turning it on. But from the datasheet?

I'm with Roger on this - if for some reason you MUST tune out the coil inductance with a capacitor than put  a resistor in series with it (not going into calcs here but its value should be the same as the DC resistance of the coil.) If you get the capacitor value right the complete network has a purely resistive impedance (well it would if all the parts were perfect).

Diodes are easier and cheaper, you don't (normally) need anything fast or fancy, a 1N400X (X 1 - 8 denotes max operating voltage) works well in most cases.

MK

Roger Wolff wrote:

 

IMHO, I would NOT put a capacitor in parallel with the coil of a solenoid.

 

So... My advice is: Use a diode, don't worry about the possible spike that occurs before the diode kicks in. (I did some measurements last Friday and the scope couldn't see the spikes due to the slowness of the diode. )

A reverse-biased diode is a variable capacitor.  The initial current from the inductor discharges this capacitor until the diode forward-biases and the capacitance vanishes.  At least that's how I understand it.

For details on the phenomenon: Varicap - Wikipedia, the free encyclopedia

So I agree with Roger and Michael: you should not need more capacitance.

More on flyback diodes: Flyback diode - Wikipedia, the free encyclopedia