Microwave on-off sequence for custom control circuit.

[GTechno13]

I have built a custom control circuit for an old microwave and I just have a couple questions on power up and down sequence I was hoping someone could answer.
I just got a new apartment that didn't come with a microwave so my dad sent me with the one I took apart (when it still works ha ha) so I can't afford to break this one.

The microwave has, and I have control of, a relay and a triac. The relay appears to turn on the fans, light, and supply power to the triac and then the triac controls the magnetron. When microwaves turn on you can hear the magnetron take a second or two to fully power up. My first main question is if I need a delay between the relay on and the triac on in my code to maybe charge the cap or something or can I turn them on at roughly the same time (relay being right above triac in code)?
Another observation is that when the cycle finishes and the beeping begins, the magnetron sounds like it is switched off (triac) but the fans remain going until the end of the beeping cycle or the door is opened (relay). Would this be a fair assessment? Is this for extra cooling of the magnetron even though many people often open the door which stops everything?

Thanks so much if someone can provide a quick answer. I'm ready to test and am a bit impatient. I just don't want to wreck the cap or mag with an improper start up cycle.

Thanks!

 

[KeepItSimpleStupid]

Somethings's wierd.

Light and fan should have two different controls.

1. Light should work when the door is opened and when cooking.
2. When cooking fan, light and triac should turn on at the same time.
3. Triac should cycle, but I think only when percentage of time is used.
4. A thermal on the magnitron should be able to turn off the magnitron if it overheats.
5. There are two sets of interlocks for the door.
6. Set one will short the fuse if misadjusted.
7. The second set controls the light (door open: light on/magnitron off); Door closed (magnetron enabled)

There are a few Caveots for all of this.
There should be a short delay after the triac turns off before it's enabled again (1 line cycle should be OK) (continuously on is OK)
You may have to bang the triac for a full cycle of voltage. Look up triac controls of motors.

Most important is that the % on time is NOT equivalent to % heating. That has to be determined some other way. Not sure how.
The problem is that the filament of the magnetron has to heat up.
I think, you can use temperature rise to figure it out. e.g. Run a few minutes with tap water at 100% measure the rise in temp. Make sure that when you input 50%, it rises 1/2 the number of degrees and at 10%, 0.1 of the number of degrees. You have to experiment with this.

e.g. a 10% on time, might really mean that the actual time on is 15%.

Two methods can be used. One is phase angle fired and one is known as Zero Crossing Pulse. In the latter, a number of full cycles on and full cycles off are used. Phase angle firing is harder because only part of cycle is used.

Because this is an inductive load, and a triac will turn off at zero voltage and because it's inductive, the voltage and current zero crossing will not co-incide, therefore that once a voltage zerocrossing is found, the triac must be continually pulsed to just before the next zero crossing. If a current zero crossing occurs before the next voltage zero crossing, the triac will turn off.

I hope I didn't miss anything. The magnetron filament and high voltage turn on at the same time `because the filament and HV come from the same transformer.

The magnetron, HV and blower should come on at the same time. They must be interlocked.The light comes on when either the door is open or the blower is on.
The interlock consists of both primary and secondary interlocks. One will crowbar the fuse if the other interlock isn't activated. Usually these detect bent doors, so there are one at each end. If one interlock works and the other doesn;t the fuse gets crowbared. The other switch usually enables power and turns on the light. The blower will turn on the light too.

 

[GTechno13]

Thanks, I believe the light will come on regardless of the relay state. I kept all that wiring and just have the relay on the coil side. I copied the triac driver circuit off the old board (optoisolator, snubber, and so on) and just drive the opto from my uC so as not to change any of the high voltage circuitry. This microwave has 3 total microswitches. I believe one is the series stop, one is the fuse shorter and one is the logic one which I have feeding my board. My original plan was to improve the microwave with a humidity sensor, microphone (for smart pop!) and have various power and custom settings but this all relied on a capacitive touch, laser engraved, edge lit piece of acrylic which was going to look way cool but I ended up having so many problems with the touch ICs and managing the 16 closely grouped pads that I gave up for a long time. As mentioned, I need a microwave NOW so I just threw in a rotary encoder and that's it. Turn the knob and it counts up by 5 seconds and down by 1. Push in and turn to set the clock. It's like a digital version of old timers and I really like how intuitive and simple it is. My main concern was if the relay somehow pre-charged the capacitor before the magnetron comes on. I do know its a voltage doubler but I'm not too familiar with anything beyond that but after looking at the diagram now (duh!) it looks as though the triac must be on before anything high voltage is powered. I'm going to go try it now! Thanks!

 

[Nigel Goodwin]

You don't want phase control - which wouldn't work anyway.

Microwaves simply turn ON and OFF over a good number of seconds to adjust the power, there's no 'sequence' or 'startup cycle' about it. You can't use too short a cycle time because of the warm up time of the mag.

Commonly relays are used both for the mag and the rest of the circuitry, but a small number of ovens do use a triac instead of a relay - which seems pretty pointless, as the triacs go faulty, the relays almost never do.

 

[GTechno13]

It works! The light does not come one when the door is open though and I realize that's how this one has always been. I do know that for power regulation they use what could be referred to as PWM on a time base of 10-30 seconds. 5 on, 15 off = 25% on a 20 second base which really does not work well. I heard the new inverter ones might be able to do constant power control which would be nice. Its really hard to melt butter in a microwave without it splattering everywhere.

 

[KeepItSimpleStupid]

Nigel: A phase controlled microwave here: https://www.electro-tech-online.com/custompdfs/2012/09/v22-1.pdf

GT: Exactly, but the 25% duty cycle isn;t equivalent to 25% power because of the necessity for the magnitron filament to have to heat up.
So, if you so some experiments with temperature rise and the hardware duty cycle and create a plot, you should be able to figure out what 10% power vs 90% power really is. You want a 50% that is inputted to mean 1/2 the temperature rise of 100% for the same amount of elapsed time.