Triac Failure Modes?

[The Doktor]

Okay, this is another question regarding a very old microwave I am trying to help a guy repair. Since my last post got some suggestions it should probably be replaced instead, I will explain a bit. This is not a little tabletop microwave you buy off Amazon for $100. It is a very large, built in unit which the guy said it would probably take two people to move. Not super easy or cheap to replace. Comments about the magnetrons wearing out our true, but a new magnetron is not super expensive.

I'm sure some of you will wonder why I don't inspect it personally. Well, as I said in the previous paragraph, it is not easily moved. And due to some fairly serious physical disabilities, it would also take great effort to get me to the oven. I don't believe the guy is very well-versed in electronics, but he knows at least a little, and is fully aware of the dangerous posed by the magnetron and its related circuitry. Just in case, I did mention that the high-voltage supply is quite capable of ending a life.

Here is a (very shitty) little schematic I drew up showing the pieces in question. Terminals 1, 2, 4 and 5 are at the connector on the edge of the circuit board. The large triac and transformer are located somewhere in the oven. I drew the circuit board portion to visualize how everything on the output of the optical coupler was connected. I just added the bottom in for the purpose of this post.

What I found on the board was the diode on the optocoupler had gone open circuit. Also, resistors 1, 2, and 3 all had their outer coating burnt off. There were black smudges on the board under all three, but the board itself was not discolored like usually see with long-term exposure to heat. It looks like they probably went quickly with a rather large overload. The resistances on the schematic were exactly what I measured with a Fluke 8840 a/af. The Triac controls power to the high-voltage transformer, the filament has its own separate transformer.

What I am guessing has happened is that the large triac is no longer triggering, which would basically make the little triac in the optocoupler try to supply the power to the magnetron itself, through those three resistors. So I have two questions. The first one, do triacs sometimes fail in such a way that they don't trigger? According to most posts I find, the way they usually fail is by short-circuiting. The second question, do you agree that the likely problem is the large triad failing? And let me add a 3rd question, is there anything else you think would have damaged the resistors? It seems to me that so long as the large triac triggers, there would be no way for access current to flow through the resistors and optocoupler.


Thanks
Ed

[The Doktor]

I actually thought of another question. Is it likely they were using phase angle control to vary the power on this oven?

When I was looking at the board, I just assume the opto was controlling and AC coiled relay. The only ovens I've ever worked on at all had a single transformer that supplied both filament and high-voltage power, and were switched by relays. The "lower power" settings were always done by running at full power for a while, and then switching off For a while. Not really nice for cooking food, nor is a particular good for the magnetron in my opinion but that's what I've always seen. If it does use phase angle control, I will have to change the opto. I chose a part with zero crossing switching, which would kind of break the whole phase angle thing.

[floobydust]

Get the guy to send clear, detailed pictures and the make/model number. How many watts is this? As an example, a Subway Turbochef Tornado is huge 190lbs! Microwave is 3,500W alone yet no phase-control.
Older microwave ovens do not use phase-control, it's usually full blast on/off. Phase-control gives less average power into the transformer, but that would also undervolt the tube's filaments- so you would need a second transformer.

Yes - if the triac is lazy and does not activate, the opto-coupler and parts around it will roast. I have also seen triacs fail where they go less sensitive and want way more gate current to trigger, so they sorta work but not really. It still damages the gate drive.

In your schematic some of the resistor values look way low:
R1+R3 typically is 180Ω (120VAC) or 360Ω (240VAC) in datasheets, limit is Vpk/1A opto peak current.
R2 (gate resistor) is around 330Ω-1kΩ.
Careful some of these resistors need to be flameproof/fusible resistors for sure R1, R3.
I am just ballparking the part values and without oven details they could be off.

I would replace the burned resistors, triac, opto-coupler.
For testing, you can substitute a light bulb in place of the big transformer as a load for the circuit. Unless you want to just try the new board. If the opto carbonized and shorted input LED to output then the control board would be damaged as well.
The new triac will need proper heatsinking thermal paste/pad or the new one will just overheat and croak.

You don't want zero-cross switching with a triac driving a transformer. The inrush current is the highest possible at zero-cross where dV/dt is highest on a sine-wave. I know it's counter-intuitive. Sometimes the control board does the on/off phase angle control to limit inrush and magnetization of the transformer core.
Can you ID the part number of the original opto-coupler? Wild guess MOC3072, MOC3052 etc.

[S. Petrukhin]

Usually TRIACs do not go into an open circuit, but become a short circuit. 

But in your case, the control pin was obviously destroyed.
The powerful TRIAC has not opened, I agree with you. 

I see an obvious danger in your circuit: storage capacitors are not limited in accumulated voltage and its polarity.

1. The inductive load can produce a large surge and it will persist in the capacitors. Then you discharge capasitors into the control electrode through the divider by 2 approximately, creating an excessive current. Despite the short period, it easily kills the delicate pin.

2. Not all TRIACs can allow control in the 4-v quadrant when a negative voltage is applied to T1 and a negative voltage is also applied to the control pin.

Just a couple of days ago, I came across the vagaries of TRIACs and asked a question about it.
I haven't fully checked and made sure of the calculation yet, but it's clear that you can't leave TRIAC s without a snubber.

[The Doktor]

Hey guys, I always hate it when people post questions about something and then never follow-up with how things worked out. As I don't want to do that myself, I am now posting a little more about this project. It has been a wild since my last post as the ovens owner has been busy with some family stuff.

First thing, I found out what the oven is and why fixing it is very preferable to buying a new one. The oven is an old GE Spacemaker Installed new in 1980. I'm posting three links below to show you what this oven looks like. It is installed above the stove just like shown in the first and second link. The problem is the shape of the old oven, nobody makes one like that anymore. It is the same height as the new ones that you can buy in the rear, but at the front it is thinner, and angles down from front to back. The problem is not getting an oven to fit in the existing space, it is that if you put it there it is too close to the front burners on the stove and renders the stove basically unusable. Mounting it up higher requires modifying the cabinet above it, which will cost a bit to get done by the time you are done with modifications and painting etc. Also, the cabinet above it will then not be very tall and will no longer be able to hold the items which are normally stored there.

https://www.pinterest.com/pin/ge-spacemaker-microwave-1980--94364554670268081/
https://artsandculture.google.com/asset/advertisement-for-1978-spacemaker-microwave-oven/GAG4e8ZCiko92A
https://www.reddit.com/r/PcBuildHelp/comments/b2e1sj/ive_acquired_an_old_general_electric_spacemaker/

As to my suspicion that it may use phase control for power, that is definitely incorrect. I asked, and he said you can definitely hear it cycle on and off at lower power settings. Part of the reason I suspected it might be phase controlled is that it does have a separate transformer for the filament. I assume they did this to eliminate the constant cycling of the filament at reduced power settings, which I always thought was a terrible idea.

As to whether to use a zero switching triac or not, I missed a bit of a clue earlier. When I was trying to figure out what the Opto was, one of the post was from "wraper" and I failed to notice, but he provided an SK part number for a replacement of the original Motorola 01S63. This is not a zero switching part, I will replace the part I installed accordingly.

Now onto the fun part, the original large triac. The owner removed it and brought it to me the other day, my DMM seemed to show A1 and A2 shorted, and the gate completely disconnected. Nice to see the part was actually bad, but this would not explain the oven not heating or the small Opto getting fried. So I decided to test it a little further, and it was quite an interesting test. I hope test leads to A1 and A2, and fed them with my Power Designs 6050. I measured the voltage drop at the terminals with a fluke 8040 a/af. Initially, it performs like a low value but somewhat unstable resistor. I think I was getting something like 2V drop at 2A, It would keep going up to about 3 or 4 V, past that it would suddenly drop down to 2V Again. Obviously a defective part, but they should actually make the oven run at full power constantly and would not explain the failure of the Opto.

Well, after I guess somewhere between 30 seconds and a minute, the device was starting to get warm. All of a sudden the current drops to zero, and stays there even when I crank to supply all the way up to 60 V. So I decide to try triggering it. What happens when you do this is completely random. Sometimes the gate will not even draw current, sometimes it triggers like it should but will turn off after a few seconds even with 5A flowing through it. Sometimes it will just turn on and off like a relay as quickly as you can apply and remove power from the gate, and sometimes it must be parking internally or something, as it makes noises like a scratchy volume control while the current meter jumps all over the place. Also, even though the terminals are not at all loose, applying pressure to any of them can cause its behavior to change. Clearly this triac is the source of all the problems.

Thanks to all who offered advice, I will make a final post about whether it works or not when it is done. I am almost certain that it will.

Ed

[Jwillis]

TRIACs can fail in a Half wave condition. It would act like a resistor in parallel with a diode. So it would also measure as shorted but function unpredictable since burned out diodes don't make very stable resistors. So with only half wave the transformer would not function very well if at all.