Microwave oven: How can I temporarily limit the startup load?

Question

Problem: It's a common problem for "NON-inverter" type microwave ovens to pop circuit breakers, and mine is no exception.

UPDATE Jan 2021: Extension / power strip added, (mentioned below) has still eliminated this problem completely! I think I've only popped that breaker once in two years. Fascinating as this subject was, I'm done. Note that both virtually identical KitchenAid microwaves popped that breaker, but the other same sized circuit breaker circuit, (slightly longer than this run's 1 foot away from the main panel), has never popped the breaker with either microwave. So, AFAIK, this problem has been confirmed as a design defect, and not a malfunction, on a circuit that is on an unusually short run. And further, the added small amount of impedance added, eliminated these design symptoms.

There could be a huge number of reasons for this. Even though I thought I had really researched microwave ovens extensively, have replaced magnetrons in two nearly identical ovens, I found this one resource that taught me more than I had ever known. It covers design goals, troubleshooting, and things you'd never know, unless you both designed these ovens and also repaired them as a full time job:

Notes on the Troubleshooting and Repair of Microwave Ovens, written and compiled by Sam Goldwasser. I highly suggest you read it carefully. It will likely save you a ton of time, even if you are a master IEEE designer.

Based on what I've learned, I have new questions and new concerns:

-- Using an add-on power strip could be dangerous, if not properly designed. (This "fixed", masked or worked around other problems for me.) My strip is well built and I'm not concerned. But be aware that the magnetron is 5000v hot on one side, and grounded directly to the case. Lose the ground and /or a lousy partially floating neutral, (which I've seen before), and you could get killed from the case becoming some fraction of 5,000 VAC, through you to earth ground.

I've got two identically designed ovens, both of which deliver the same solid 700W, computed from testing heating water. (Litton manual says 1 liter 1 minute -->> deg C x 70 = Watts -- mine 18 f = 10 c x 70 = 700 watts). So, I doubt if I have a shorted diode, or have lost some turns in my transformer. But, arcing in the transformer itself, the triac, the diode, the capacitor becoming weak and things going on inside the magenetron, the waveguide itself, food buildup; these and many more factors can ALL affect performance, and can affect the startup and continuous draw of the oven. Inverters just would add to the complexity of determining what's really happening, since adding a switching power supply can add a set of different problems.

Before I added the power strip, throwing the breaker seemed to be getting worse, which could mean that the 20A breaker itself is getting weaker, (which I'd rather not replace). I had replaced magnetrons in two nearly identical ovens, and understand the specific designs of both, which I don't intend to modify at all. It seems to me that the bleed resister across the capacitor that discharges it, (and does work), does a "great job", as intended. But other people have said that this means that startup draw, to charge that capacitor, is the cause of about an extra 5-10A on a 10A microwave, for the short startup time.

Again though, I'd like to add something external to the oven, without opening the case, which might then work for many that have the heavy transformer and not an inverter switched power supply type oven.

My goal would be to design an add-on device that would limit the voltage / current for 0.1-0.5 seconds while the microwave is charging that capacitor (which doubles the DC voltage AFTER it is fully charged). On most ovens, I would guess that the control board's filter capacitor should keep that operational with just a very short term voltage drop, even if a substantial voltage drop to its power supply.

In a standard oven, there's a huge and heavy transformer that is serving as a huge inductor, fed with normal 120 Vac, usually using a solid state triac as a switch. (Panasonic, Sharp and many others use inverter power supplies.)

SIMPLE solution, that has so far, works!

SEE cautions about adding this, mentioned above!! Adding a simple 6-8 FT AWG 14 extension cord to the circuit seems to have fixed the problem. As @Charles Cowey suggested, my oven is plugged into a circuit that is literally one foot from the breaker box. It seems cold weather reduces AC use, increases the available line voltage enough, and puts this right at the edge.

I will have to open the power strip [and report back] to see if there are some added inductors inside. (AWG 14 has an impedance @60 HZ of only 3 ohms / DC resistance 2.5 ohms per 1000 feet. So ~.03 ohms added to an effective 12 ohm oven doesn't seem significant enough to fix it. Ref.: AC/DC Chart / pub. by Anexter ) In any case, if building a universal solution as inexpensively as possible, it seems that simply CHOOSING a good surge protector with an added torroid coil inside, (which I have seen), should solve most of these problems.

"Older ovens with large transformers" (by @Charles Cowey) implies that standard designs have changed, perhaps to switching supplies?? IF so, am I solving a problem that will be phased out in the next decade?

OTHER Possible solutions?:

NOTE -- EXPERIMENTALLY rewiring that 120/240 VAC circuit feed is NOT a good idea--Do it wrong and you could get a 5000 V shock!

1.2 ohm resistor, in-line, before oven's 120 V ordinary line cord
One simple method might be to simply add an in-line resistor. A 10A microwave @ 120 V is drawing 1200 W, with an effective impedance of 12 ohms. Simply adding a ~1.2 ohm resistor in series would cut the voltage/power by 10%, as well as the surge power, which is tripping breakers, which might do the job. While running of course, that resistor would be creating the heat of a 100 W lightbulb.

Lightbulb: NO!
Unfortunately, I need a lightbulb to use as a resister, which would have less resistance as it heats up, which of course incandescent bulbs do not. Nevertheless, this would be the perfect inexpensive method to use as the main component of this device.

??
So, what could I build that would accomplish this??
What I need is a resistor that starts as a 12 ohm resistor and becomes a 1.2 ohm resistor, or less.

After research, there are inexpensive ($5 for 2) NTC (Negative Temp. Coefficient) devices that do ICL (Inrush Current Limiting). Unfortunately, they work by heating, and won't cool in less than 30 seconds, and so won't work in ovens that cycle on and off every 6 seconds (@ 50% power setting).

Amtherm company focuses on RTC devices and designs, and seem to be the best resource for designing circuits that utilize RTCs.

As a sidenote, these NTCs are using in virtually all of today's variable speed motors used commonly in HVAC systems. These RTCs often fail, and commonly the AC tech unknowingly recommends replacing the whole motor at a cost of > $1,000 rather than the < $5 RTC replacement!

Answer 1

In an older model microwave with a huge heavy transformer, the inrush current may be due to the transformer saturating when power is switched on at the peak of the voltage waveform. If that is the case, very little additional impedance may be required. I have seen the problem in a location where the distribution panel was in the basement directly below the kitchen counter. There may have been only about six feet of wire between the circuit breaker and the microwave. In that case, the problem was solved by replacing the standard 20-amp breaker with a 20-amp, high magnetic trip breaker.

It is possible that only a tiny impedance to the power feed, perhaps 15 feet of 12 AWG cable may be sufficient.

Answer 2

3 recommendations

• Test or replace bulk Film cap (test by RLC meter @120Hz)
• insert an Inrush Current Limiter (ICL) for rated current eg 120Vac, 10A a) 7 Ω to 0.08 Ω or b) 20 Ω to 0.18 Ω. (Slow response)
• add series 12 Ω power R and Time delay SPST 10A relay to shunt R

My Panasonic oven may have a soft start circuit as I can tell by the delay in lamp dim after it starts the turntable, fan and lamp first.

But in my experience , the 1st choice is the best for units beyond warranty and worked for me with exactly your symptoms.

You may also choose an ICL with your specs based on the bulk capacitance at 120V.

The R+ Relay solution also avoids the surge current.

However power distribution in the home and thermal response of wires fuses and breakers are normal never exceeded unless there is a component degradation or fault. Adding some short cable resistance only reduces the surge current a bit.

Therefore the universal solution is repair the unit and #1 recommendation is the most common.

Answer 3

What happens is, the transformer saturates on startup. The magnetization current normally passes through zero at about 1/2 way through each 1/2 cycle. If it gets turned at the beginning, the mag. current starts from zero instead of -peak, so it heads towards double peak and saturates the core. The coil then behaves like an air core coil, so the current goes through the roof, and all the excess flux passes straight through the saturated core to the steel case underneath, causing that characteristic THUNK. The transformer needs to be turned on 1/2 way through a 1/2 cycle. Don't worry about the capacitor on secondary. The secondary is loosely coupled & the flux can bypass it through the magnetic shunts if necessary. Checkout my answer on this site for how the circuit works. What are the bare minimum parts needed to operate a magnetron?

As far as I know, all of the microwave oven processors use the supply frequency as their clock, which means, the relay pickup will be synchronised to the supply. If there is no little silver can with a crystal in it, it will be, especially if it THUNKS every time the relay picks up. If so, you could try this super simple delay circuit I designed in place of the transistor that drives the transformer relay. Remove the transistor, set the pot fully CCW and install this circuit in place of it. Bend the wires as shown so they will bend if bumped and not push the tracks off the main board, then, with the power set low, so the relay drops in & out, adjust the pot until the transformer just stops thunking. Then count the turns until it just starts again, and set the pot ½ way between. That will switch the transformer on in the middle of a ½ cycle when the magnetisation current would be crossing zero. 2N7000s have a Vth spread of .8-3V and I don't know what voltage the processor feeds out to the base resistor, so if you run out of adjustment, change the capacitor to .022uF. Let me know is it works. Cheers!