SparkFun Forums

[Go0bur]

Hi, I'm currently collecting the parts to build a reflow toaster oven. I was wondering a few things:

1.Is there any advantage in using two relays (one for each AC path) over just using one relay? I notice most AC power plug components have switches that disconnect both paths.

2. As of right now, the 5V relay is running off the same AC-DC regulator as the uC. THe switching of the relay is going to cause voltage spikess, I'm just wondering how big the bypass capacitor should be beside the switching MOSFET that controls the inductor. (I think large, because relays switch slow compared to most ICs which use 0.1-0.01uF as a bypass capacitance). I was originally going to go with a 12V relay, but that would have required an additional 5V LDO and a zener voltage clamp (to clamp the 12V to <9V so the LDO can accept it) in addition to the switching AC-DC converter. I don't need the extra parts with a 5V relay...so it would seem this is the only advantage of a 5V relay over 12V, since I can't provide enough 5V current directly from the uC to directly switch the relay anyways.

3. A flyback diode is supposed to be connected across in reverse bias to the relay's coil terminals because shutting off the relay causes voltage spikes ending current flow across the relay's coil inductance?

4. The third pin on AC outlets is used as a ground for faults right? It's purpose is to provide a low impedence path to ground for short-circuits and other faults so that less current will flow through someone touching surfaces of the device where fault current is running? (Via the fact that for two parallel resistors, the smaller resistor dominates the larger resistor). I'm just not sure what to with it, whether it needs a relay or not. Oh, my oven design is supposed to act like a smart power bar/timer with temperature sensor that sits between the wall and oven...so I don't have to open up the oven. That should affect the answer.

5. Some people have suggested using TRIACS because you can control the current rather than just on/off. They have also said that you can interface it directly with a uC. Right now, I am trying some 5V relays, but of course, the current is too high for an uC to supply so I have to use a MOSFET to switch the current. BUt how can it directly interface with the uC? My understanding is that if +V is applied it will pass one polarity of the AC waveform, and applying -V will allow the other polarity to pass, so to pass everything you need to switch between a +V and -V...this can't be done directly with a uC. Am I misunderstanding something about TRIACS?

5. Any ideas how to get the uC to track the AC half-cycles in order to properly switch the TRIAC to control current?

6. Suppose I am really lazy and forgetful, any ideas how to get the oven to open itself to cool down the PCBs at the end of reflow? (and maybe even how to tap the oven on it's own to jerk the components into alignment at the proper time? Servo motor? maybe something more consistent and reliable so the oven door doesn't drop open and jerk uncooled solder?

7. I have never used a convection oven...how strong are the fans? As strong as the air flow out of the back of a typcai PCB tower? An oscillating fan?

[Kuroi Kenjin]

I can answer a few of these:
1) I believe the double relay is ment for complete electrical disconnect for safety reasons. I don't know what other advantage it would give.

3) Yeah, a schotkey type diode is usually recommended for a flyback. Although I have a ton of 1A regulator diodes and they seem to work well too.

4) Yup, your correct, usually this is attached to the metal casing and/or through a 1M ohm in parallel with about a 0.1uF cap to bleed off excess charge (or at least on the few projects I worked on). Also, won't you need to take out the internal controls of the oven? Won't they try to counter your control wth it's?

5.1) I thought TRIACs just need a current pulse to turn on (+V) and a reverse pulse to turn off (0V or less). I may be wrong as well.

5.2) The micro can be hooked directly to the AC via a 1M ohm resistor or other fairly massive resistance divider. The 1M or large resistors should keep the current sufficiently low to keep power dissipation below 1/4 W or so. Also this'll keep the voltage (depending on configuration) well with in the PIC's A/D Range. Also remember that the line voltage is very slow compared to the A/D speed of the PIC.

6) I did have an idea to put a fan in mine to cool off the inside fairly quickly, but the hole would have probably drawn off enough heat to reduce my overall peak temp. Just put on a "hey stupid!" buzzer.

Happy baking!

[Go0bur]

4. All toaster ovens can reach temperatures far beyond reflow temperatures so the oven temperature controls will never kick in and most also have "dumb" modes where they just turn on. If you set it to maximum or whatever, the oven will basically always be on whenever you apply power to it for reflow purposes.

5.1 THat's my point exactly, it seems that a TRIAC needs a negative current or input which I have never seen an MCU be able to directly provide.

5.2. I didn't think a PIC ADC could handle negative voltages, no matter how small...are you sure about this?

[Philba]

too many question to answer all.

I have played with a design for a reflow oven though the skillet approach is a lot easier. I would use TRIACs over a relay any day.

the double relay is, IMO, for unpolarized plugs so that the hot is always disconnected. If you use a polarized or grounded plug it's unnecessary.

a TRIAC's trigger current differs based on the firing quadrant. I think you just need to ensure you deliver enough current for the worst case that you will use. You dont use all 4 quadrants, just 2.

I don't see how you can control the current with a TRIAC. You can do phase control which does control the average voltage. Are you planning to PID control the temperature? hard to do with a relay.

US 120 VAC wiring has 3 wires: hot, neutral (aka return), ground. Ground is earth ground and does provide a second return path. GFIs sense the difference between neutal and ground and then triggers if above a preset value (5mA? i forget). 240VAC has 2 hot, though.

the relay voltage spike - back emf - are due to the collapsing magnetic field. a fast diode should take care of the those. another issue is due to the current pull when turning the relay on - that can be as much an issue. even with 2 VRs, you may see some affect from relay current surge. I would breadboard it and see what happens since so much is dependent on the components you are using. I've run 5V relays off the same rail with no ill effects but the VR was a 7805 - TO220 with a couple hundred uF filter cap. by the way, you don't really need to drive the relay with a mosfet - seems like overkill.

[Go0bur]

Best case scenario is to use PID, of course, but relays are a more available method for me right now so that made PID not a requirement early on, but it never hurts to look into a bit more.

It was my understanding that the during a half-cycle, the triac's conducting state is self-maintaing once started and automatically shuts off at the end of every AC zero-crossing. So, if you pulsed the appropriate control signal partway through the every half cycle, you could control the amount of current "in a way". I guess it's more accurate to say you can tune the power being delivered to the load, not exactly control the current, since you don't really get to decide the level of current. Kind of like an AC version of PWM.

I know MOSFET is overkill, but it's what I have laying around and what I prefer to use.

[Philba]

no, what you are describing is phase control. the triac conducts during the on part of the phase. what you control is the amount of time the load is powered. it's just like a relay in that sense. You have no control over the current. For a resistive load, you will get the effect you want, though. It's just not current control. look here for a decent explanation. http://www.littelfuse.com/data/Applicat ... an1003.pdf

[Go0bur]

It has just occured to me that depending on how complicated the oven is, using phase control might make the oven non-functional, especially if it is using a convection fan, or timing electronics and the like (since I am not splicing into the oven, but just controlling the wall power reaching the oven).

But that is a great article, much better than the bits and pieces I've found from googling. I get what you mean by quadrants now...(never understood what it meant before).

[Philba]

you will have to control the heater seperately from the fan. I think you will want to do that with a relay or triac. not that hard you'll just need to work out what's what. also, you may be able to control the fan speed as well,

by the way, if you are using triacs, take a look at the MOC303x opto drivers. worth the isolation.

[Kuroi Kenjin]

With the AC detection you can use the large resistor and full wave rectify the signal, then use a comparator to ground as the sign bit.

or

Attenuate with the large resistor and make a voltage summing circuit to add an offset voltage bringing the entire waveform into the positive voltage.

[Go0bur]

Microchip has an application note suggesting a method of using a 5Mohm resistor in series with uC pins that have diode-ESD protection. I can't picture in my head how it would work without thinking it through more though. AN521.