How do you test a diode?

[Muse]

My microwave stopped functioning and I posted a thread in this forum. It seems to work as always except it no longer heats stuff. People said I should test the diode. I did and thought it was dead and have one on order at my local indy e-supply. However, now I think my diagnostic is wrong.

I was told to take a multimeter and measure the resistance and then reverse the position of the multimeter's probes. In one position I should see near zero resistance, in the other near infinite. I was getting apparently infinite resistance in both cases and deduced that the diode had died. I have another dead microwave and tested it's diode, same thing. I decided both were bad.

However, I have a working microwave oven and just opened it up and I get the same thing. I must be doing something wrong. In the case of the non-working ovens, I tested with the diode in place and removed. In the case of this working microwave, I'm testing with the diode in place. I don't see why that should make any difference. If there's near zero resistance in one direction it should show that even though it's hooked up. One end of the diode goes to ground and the other is evidently connected to the capacitor.

These diodes are HVR-1x 3 and two HVR-1x 4's. I tested on 200 range and 20k range, same thing.

What's up here? Thanks for any help.

 

[Demon-Xanth]

Are you sure it's a diode?

 

[Calin]

You should only test with the diode removed - the circuit can fool the measuring (imagine you test a resistence inside the <what kind of equipment would be> that have several resistences in parallel. You won't read the resistence you want to read, but all of them.
If your multimeter is using more than some 0.7V while measuring resistances, the resistance in one direction would be less than some 100 ohms. As the voltage increases, the perceived (measured) resistance decreases

 

[FoBoT]

yeah, so remove it from the circuit and check it with an ohm meter, one way it will appear as ~0 ohms and the other it will appear as ~infininte ohms , if its good. if it is open, it will be infinite both polarity, if it is shorted, it will be ~0 ohm both ways

 

[Muse]

Originally posted by: Calin
You should only test with the diode removed - the circuit can fool the measuring (imagine you test a resistence inside the <what kind of equipment would be> that have several resistences in parallel. You won't read the resistence you want to read, but all of them.
If your multimeter is using more than some 0.7V while measuring resistances, the resistance in one direction would be less than some 100 ohms. As the voltage increases, the perceived (measured) resistance decreases

If measuring several resistances in parallel, I can't see how I could get a reading of more than the component with least resistance, right? For instance, if A, B, C and D are in parallel, if my diode is D and I'm reading infinite resistance from one lead of D to the other, the resistance of D is infinite both ways, regardless of the resistance of A, B or C. The less than infinite resistances of A, B and C could only reduce my readings. The multimeter is showing "1", the same thing when measuring pure air.

 

[Muse]

Originally posted by: Demon-Xanth
Are you sure it's a diode?

Look up HVR-1x and you will see it's a high voltage diode, typically used in microwaves.

 

[Muse]

Originally posted by: FoBoT
yeah, so remove it from the circuit and check it with an ohm meter, one way it will appear as ~0 ohms and the other it will appear as ~infininte ohms , if its good. if it is open, it will be infinite both polarity, if it is shorted, it will be ~0 ohm both ways

I already removed the suspect diodes and tested them. They test dead, as reported. I don't really want to remove the diode from the good microwave because I'll just have to reinstall it, a real hassle, and I don't really like messing in there, considering there's a highly charged capacitor. What's wrong with my analysis? Shouldn't I get less than infinite resistance when testing it in place based on my reasoning? I realize something's wrong in my testing here. Can I safely assume my other two "dead" diodes are truly dead?

 

[Muse]

OK, I removed it. It still tests infinite, both directions.

I didn't remove it entirely. One end is attached to the case (i.e. grounded) but the other end goes to nothing. That should be totally tantamount to removed. Nada. What's wrong? Why doesn't it test good. This microwave's heating stuff just fine.

 

[Muse]

Ah! I just did a Google search on the string "how to test a high voltage diode" and the first hit is a page called How To Test the High Voltage Rectifier (Diode) Used In Commercial & Residential Microwave Ovens

The deal is many multimeters will show infinity both ways. They have to use at least a 6 v. battery to do otherwise...

"This test requires an ohm meter with at least a 6 volt battery in order to accurately measure the front to back resistance of the diode. Meters with insufficient battery power may read infinite resistance (open) in each direction, mistakenly showing a good diode as being open."

I used two different digital multimeters, here, different species. I'll try my analog multimeter. It uses a 9 volt battery... Didn't know I'd ever need it again, but guess I do!

 

[stephenw22]

If my multimeter didn't have a setting for diode testing, or if the diode had a voltage drop that was too high for it to handle, what I'd do is set up a test circuit for the diodes. Find out what kind of voltage drop they have and what their rated current is. Then make up a simple circuit to test it.

Example: I have a diode with a forward voltage drop of ~2.2V (because is has multiple junctions) and a reverse breakdown voltage of ~60V. I want to see if it is ok.

I'd use a 12V power supply. You could use a computer power supply to get this voltage, you won't need much current. Connect the diode in series with a suitable resistor that restricts the current so you don't cook the psu, and dissipates the power if the diode is a short circuit inside.

Connect the diode so that it should block the current. The entire 12v should be across the diode. If there's a significant voltage across the resistor, the diode is faulty.

Connect the diode so it would pass the current. In this case, you'd see a 2.2v drop across the diode, and a 9.8v drop across the resistor. If it's much more or less than this, the diode is faulty.


Note: Most diodes in small electronics have a 0.7v forward voltage drop. I just happen to deal with a lot of stuff that has a 2.2v drop.

 

[Spencer278]

Does your multimeter have a diode testing mode?
If not find out what the Von for the diode and the current for that voltage then get a DC source greater then Von and a resistor to limit the current and measure current thru the resistor.

 

[Muse]

Yup, my analog multimeter just earned it keep. There's nothing wrong with any of my diodes.

Edit: My HVR-1x 3 is measuring about 85,000 ohms and both HVR-1x 4's a bit over 50,000 ohms. Infinite, the other direction.

Edit2: I think it's weird that all these Anandtechers thought it would be close to zero resistance in one direction. That's not how it works.

 

[Muse]

Originally posted by: Spencer278
Does your multimeter have a diode testing mode?

I don't know. My digital's have something saying NPN and PNP, something that looks like something you plug into with letters like a clock face: E, E, B, C, E, E, C, B. The analog meter has no such thing. But it does have a 9v battery.

 

[Heisenberg]

Originally posted by: Muse

Originally posted by: Spencer278
Does your multimeter have a diode testing mode?

I don't know. My digital's have something saying NPN and PNP, something that looks like something you plug into with letters like a clock face: E, E, B, C, E, E, C, B. The analog meter has no such thing. But it does have a 9v battery.

The PNP and NPN stuff is to test transistors.

 

[Spencer278]

Originally posted by: Muse

Originally posted by: Spencer278
Does your multimeter have a diode testing mode?

I don't know. My digital's have something saying NPN and PNP, something that looks like something you plug into with letters like a clock face: E, E, B, C, E, E, C, B. The analog meter has no such thing. But it does have a 9v battery.

Most digital have a diode setting which looks like this, --|<--, symbol. just plug in the diode and it will beep if current can flow. Your ditial meter might not be using enough voltage when measure ohms to turn on the diode.

 

[Muse]

Originally posted by: Spencer278

Originally posted by: Muse

Originally posted by: Spencer278
Does your multimeter have a diode testing mode?

I don't know. My digital's have something saying NPN and PNP, something that looks like something you plug into with letters like a clock face: E, E, B, C, E, E, C, B. The analog meter has no such thing. But it does have a 9v battery.

Most digital have a diode setting which looks like this, --|<--, symbol. just plug in the diode and it will beep if current can flow. Your ditial meter might not be using enough voltage when measure ohms to turn on the diode.

Sombitch! Well, looking, it turns out all my digital multimeters DO have that symbol. However, they test negative for these diodes. I think it's because these are high voltage diodes and these meters don't accommodate them. Only my analog meter is capable of testing these as I said in the previous posts, is my conclusion. Yeah, as stated in previous posts, those meters aren't using enough voltage. The analog meter has a 9v battery, so it does work. I used to have no real use for that analog, which was my first and from a number of years ago. I'm definitely keeping it! Digital is nice, but the analog can do some stuff my digital ones can't.

 

[Calin]

Originally posted by: Muse
Yup, my analog multimeter just earned it keep. There's nothing wrong with any of my diodes.

Edit: My HVR-1x 3 is measuring about 85,000 ohms and both HVR-1x 4's a bit over 50,000 ohms. Infinite, the other direction.

Edit2: I think it's weird that all these Anandtechers thought it would be close to zero resistance in one direction. That's not how it works.

If you give to a normal diode (0.7V drop) a 3V voltage in the "good" direction, your current will be limited by the internal resistance of the multimeter. This means the apparent resistence of the diode is not important during measuring. The actual resistance your multimeter read might be in the range of several ohms to tens or hundreds of ohms. If you use a higher voltage for measurement, the aparent resistance of the diode will appear lower (the "real" resistance will be lower also)