How to Test Diode & transistor leakage

[danny davis]

Re: Testing Diode leakage & transistor leakage

3. to test leakage you apply a test voltage with the device in a non-conducting state and measure how much current flows. In theory nothing flows, that's why anything you meausre is called 'leakage'.
4. for diodes, you apply the voltage in the non-conducting direction and measure the current.

What do you mean in a non-conducting state? Once your apply a Test voltage to it, then it's conducting the junction and state and current will flow

Most DVM meter output a TEST voltage , is this high enough? on the Diode Mode it has a TEST voltage and on the Continunity Mode it has TEST Voltage. Are these TEST voltages high enough to test for leakage?

5. for bipolar transistors, connect the base and emitter together so it cannot have any bias current then measure any current passing through it.

You short out the base and emitter? and how to connect the DVM meter in series to measure the current if the base and emitter are shunted or shorted together?

6. some tests for bipolar transistors are done with the base left disconnected but still measure collector current.

So I put my DVM meter in series with the collect lead and the output load to the transistor? and turn the circuit on?


Can't transistor testers or semiconductor testers, test for leakages in circuit?

I think the Huntron 2000 can test semiconductors for leakages in circuit

 

[Audioguru]

Re: Testing Diode leakage & transistor leakage

What do you mean in a non-conducting state? Once your apply a Test voltage to it, then it's conducting the junction and state and current will flow

Is a junction supposed to conduct when the junction is reverse-biased?
Do you know what forward-biased and reverse-biased mean?

Most DVM meter output a TEST voltage , is this high enough? on the Diode Mode it has a TEST voltage and on the Continunity Mode it has TEST Voltage. Are these TEST voltages high enough to test for leakage?

Why do you talk about voltage when leakage is a current, not a voltage?
I guess you did not look in the owner's manual of your meter to see how much current it provides for a Diode Test. Then do you know how much current it would provide to dimly light an LED? 2mA?
In Diode Test, what does the display show, voltage or current?

How much leakage current would mess up the circuit? 2uA? 2nA?

for bipolar transistors, connect the base and emitter together so it cannot have any bias current then measure any current passing through it.
You short out the base and emitter? and how to connect the DVM meter in series to measure the current if the base and emitter are shunted or shorted together?

If you connect a meter to a transistor with nothing else, what do you expect to measure? Its temperature?
You have no clue about these things do you. Maybe you want to measure how much beer it leaks?

 

[danny davis]

Re: Testing Diode leakage & transistor leakage

non-conducting state = Reverse biasing a junction?

In Diode Test, what does the display show, voltage or current?

Voltage

Why do you talk about voltage when leakage is a current, not a voltage?

Yes I know leakage is current , but which MODE on my DVM meter do you want me to use , Diode check or Continuity checker mode? both of there test voltage and test currents are different

So you want me to test the diode and transistor in reverse bias ( non- conductive state ) to see if there is current passing through and conducting

The Huntron 2000 has a TEST voltages that are higher which will give u higher current output too than a DVM meter

 

[Audioguru]

Re: Testing Diode leakage & transistor leakage

which MODE on my DVM meter do you want me to use , Diode check or Continuity checker mode? both of their test voltage and test currents are different

If you use Diode Test then the meter will show about 0.6V for a forward-biased junction, OL (overload) for a small leakage current or a very low voltage for a very high leakage current (almost a short circuit).
If you use Continuity Check then the display will show the very low resistance of a short circuit.

So you want me to test the diode and transistor in reverse bias ( non- conductive state ) to see if there is current passing through

Yes, but you do not know how much leakage current is normal and you do not know how much leakage current will upset the circuit, so the test is useless.

The Huntron 2000 has a TEST voltages that are higher which will give u higher current output too than a DVM meter

High voltages and high currents damage semiconductors. Leakage current is supposed to be tiny.
The Hunter 2000 does not measure leakage current anyway. It injects pulses to a semiconductor so you can see on a 'scope a forward-biased junction conduct and a reverse-biased junction block the voltage or break down. When the junction breaks down then it is damaged.

Why are you obsessed with leakage current that rarely occurs?

 

[danny davis]

Re: Testing Diode leakage & transistor leakage

If you use Diode Test then the meter will show about 0.6V for a forward-biased junction, OL (overload) for a small leakage current or a very low voltage for a very high leakage current (almost a short circuit).

OL overload is for Reversed bias junction ( not forward bias junction )
Low voltage when Reversed bias junction ( not Forward bias junction )

If you use Continuity Check then the display will show the very low resistance of a short circuit.

Yes the lower the resistance the higher the leakage

Yes, but you do not know how much leakage current is normal and you do not know how much leakage current will upset the circuit, so the test is useless.

True, but why would leakage current upset the circuit?

IC op amp have leakage current on the output pin, so does transistors have normal leakage current , etc.

So what would upset a component or circuit with leakage current? what does it do when you have to much leakage current , it does what to a circuit?

Why are you obsessed with leakage current that rarely occurs?

The Service manuals at work tell me to do these tests, and i have to do them for the test procedure

When a transistor is temperature dependent from freezing it or heating it up, the leakage current raises up with freezing or only with heating it up?

I have to do Temperature tests on each Metal can Transistor and IC op-amp , freezing them and heating them up to see if there voltage changes on the Emitter, base , collector , input pin, output pin , etc .

 

[Audioguru]

Re: Testing Diode leakage & transistor leakage

OL overload is for Reversed bias junction

That is what I said for a small leakage current. Obviously the junction is reverse-biased.

Low voltage when Reversed bias junction

No you are WRONG.
A reverse-biased junction with low leakage current produces a high voltage (OL overload) on Diode Test because it conducts almost nothing.
A forward-biased junction produces about 0.6V to 0.7V.

True, but why would leakage current upset the circuit?

Because a circuit is not designed for a faulty part with leakage current.

IC op amp have leakage current on the output pin

No it doesn't.

So what would upset a component or circuit with leakage current? what does it do when you have to much leakage current , it does what to a circuit?

You need to learn about circuits.

When a transistor is temperature dependent from freezing it or heating it up, the leakage current raises up with freezing or only with heating it up?

I have never seen a transistor with leakage current and I have never seen a transistor or IC that is affected by temperature changes.

I have to do Temperature tests on each Metal can Transistor and IC op-amp , freezing them and heating them up to see if there voltage changes on the Emitter, base , collector , input pin, output pin , etc .

I think you are making them too cold and too hot so you are making them faulty or you are destroying them. Their datasheets state absolute minimum and maximum allowed temperatures. How do you control the temperature?

 

[danny davis]

Re: Testing Diode leakage & transistor leakage

The Hunter 2000 does not measure leakage current anyway.

YEs it does, the manual shows the signatures of when there is leakage, the signatures are bent and bending or curving

It's in the manual

I have never seen a transistor or IC that is affected by temperature changes.

Are you serious? I get boards all the time from out in the field that have transistors that are flaky or drift up and down from temperature changes , plus they are leaky

It says it in my service manuals too at work, they are written in the 60's and 70's aerospace

 

[betwixt]

Re: Testing Diode leakage & transistor leakage

As I stated before: leakage is measured in reverse biased mode (non conducting direction) AT A SPECIFIED TEMPERATURE AND VOLTAGE, see the data sheet for the device to see what the test voltage and temperature are and how much CURRENT the device is allowed to pass.

A DVM in diode test mode measures the VOLTAGE across the junction while the current through it is being limited, normally the voltage source is the battery inside the DVM which is probably 9V or less. Most diodes are tested for leakage at much higher voltages. When you see 0.6V it is the FORWARD VOLTAGE drop, not the reverse current. If the meter says over-range it doesn't mean no current is flowing, it means the voltage isn't being dropped low enough to be in range. That has nothing whatsoever to do with leakage current. The only test that would remotely give an indication of leakage would be the resistance of the diode/transistor measured on the DVM Ohms scale, any resistance would indicate a curent is flowing through the device when it shouldn't be. The test will not be at the correct voltage and a reading in Ohms would not relate directly to leakage current but it would give an indication that something might be wrong.

Think of leakage as being a resistor connected across the device. It lets a little current flow through it when there shouldn't be any. How that influences the circuit operation depends on how sensitive to the current the circuit design is. For example, a 100W audio amplifier would not malfunction if one of the output transistors passed an extra uA or two but if the same leakage was at the input of a high gain instrumentation amplifier it might knock it's output right off the scale. Taken to it's extreme, you think of a short circuited device as having 100% leakage.

I would caution you about using too much heat and cold, especially on very old components. You run the risk of damaging them yourself.

Brian.

 

[danny davis]

Re: Testing Diode leakage & transistor leakage

Think of leakage as being a resistor connected across the device.

Yes the leakage is a parallel resistor or there is leakage that is a series resistor

So you measure the resistance of the semiconductor in REverse Bias direction? to measure the resistance?

I would caution you about using too much heat and cold, especially on very old components. You run the risk of damaging them yourself.

Why would this cause damage? how so

 

[betwixt]

Re: Testing Diode leakage & transistor leakage

It is not a physical resistor, it's an unintentional one caused by impurity in the semiconductor materials and junction construction. Leakage means current has a path to flow through when it should be completely blocked. You can think of it as being a parallel resistor because it has a similar effect but it isn't really one. There would be no point in measuring leakage current in the forward direction, it is supposed to conduct that way.

I will repeat myself and others, leakage is not a big problem and never has been, it's just something to be aware of in sensitive applications. Most transistors are tested at around 30V and typically have leakage of less than 5nA (nano-Amps), in other words the effective parallel resistor to achieve the same current would be higher than 6,000,000,000 Ohms (6,000 M). Compare that with the value of the surrounding components and you will see it is insignificant.

Like most things, if you get them too hot or too cold they suffer damage because of expansion and contraction beyond their stress limits. It tends to break the wire bonding inside the device, especially as the packaging may not have exactly the same expansion coefficient as the wire or semiconductor material.

Brian.

 

[FvM]

Re: Testing Diode leakage & transistor leakage

I still believe that your service manuals will somehow imply if the suggested leakage tests are performed in-circuit or after disassembing parts from the circuit. There might be a reading problem.

You also clarified that the purpose of the said tests is localization of defective parts, so obviously alternative tests serving the same purpose can be considered.

A dedicated in-circuit tester like the said Hunter 2000 is such an alternative method. It requires that you either have a known good board for comparison, or a good understanding of electronic circuits so that you can interprete the signatures as such.

I also agree with the comments about relation of ISO 9000 quality management to staff qualification. If the service technicians don't get sufficiant guidance from the available service manuals, an engineer with respective knowledge of electronic circuits has to define suitable service procedures.

 

[danny davis]

Re: Testing Diode leakage & transistor leakage

it's just something to be aware of in sensitive applications.

How can you tell on a schematic when the circuit or stage is sensitive? any signs of telling?

Most transistors are tested at around 30V and typically have leakage of less than 5nA (nano-Amps)

Yes but when transistors are flaky and intermittent , the leakage is higher right?

 

[betwixt]

Re: Testing Diode leakage & transistor leakage

There is no absolute answer to your first question because it depends entirely on what the circuit does. In general, leakage is more likely to cause problems:
1. when very low currents are supposed to be present, the addition of an additional small current will tend to have more effect than in a high current circuit.
2. when very low voltages are present and are followed by high gain amplifiers. The small influence of the leakage may be magnified by the amplification.
3. in very high impedance situations where there may be no paths to safely drain away even the tiniest of leakage current.

Going back to the 6,000M typical leakage resistance I mentioned, take an average resistor value from your circuits, work out how much it would be changed by adding 6,000M in parallel and compare it with say 1% manufacturers value tolerance. It will almost certainly be insignificant in comparson. VERY few circuits will be influenced by the normal leakage currents experienced in diodes and transistors.

A transistor/diode may fail with high leakage current but it is extremely rare and one of the least likely failure mechanisms. Because it is primarily caused by the manufacturing processes it tends to be fixed throughout the components lifetime.

Brian.

 

[Audioguru]

Re: Testing Diode leakage & transistor leakage

Hmmm. "60's and 70's aerospace". A long time ago.
Many defective transistors and ICs. Poor quality parts and/or poor design.
Putting the two together I get Russian or Chinese very old aircraft.

My first job after university was working for Philips first repairing then designing car radios on their production line and in their engineering lab. They made a color TV in the same factory and in 1971 I bought one (my first TV). It had one IC, many transistors and a few vacuum tubes. I kept repairing it (replacing the vacuum tubes) and improving it (higher resolution and making a pay TV de-scrambler) then replaced it in about 1989. IT WAS VERY OLD but it was designed properly and used high quality parts. In its 18 years of life it was used for hours every day and only its vacuum tubes failed.

 

[betwixt]

Re: Testing Diode leakage & transistor leakage

Believe it or not, there is still at least one millitary system that uses a manually programmed processor. By that I mean it uses a large core memory, so big that it is programmed not by current pulses but by someone plugging magnets into a pegboard! The same system was upgraded about 15 years ago, an amplifier board which previously held about 30 transistors was replaced with a new design using a single 741 op-amp. It looked very lonely sitting alone on a 12" x 12" circuit board. This is what we have protecting our countries! The joke used to be that the missile this controlled was wire-guided, not to make it hit it's target but to pull it back when it missed.

Brian.

 

[Audioguru]

Re: Testing Diode leakage & transistor leakage

Core memory?
The first computer I worked on in about 1973 used thousands of little ferrite cores strung on a matrix of many wires. That was its RAM. Its operating system and permanent memory used punched cards. Its logic ICs were DTL. I fixed a design flaw by piggybacking a gate IC on top of an existing IC. Then I was rewarded an electric shaver.

 

[betwixt]

Re: Testing Diode leakage & transistor leakage

Yes, it was core ROM rather than core RAM. It worked in a similar way but on a much larger scale. The plug-in bits were permanent magnets so unlike a ferrite core they didn't need to be read and re-written each time they were accessed. It still used the same sense wire arrangement and X-Y wires but they were wound around the peg hole instead of passing through the rings. I'm guessing it sensed the presence of the magnet by seeing if it would accept further magnetizing or not. It probably only had storage of maybe ten bits at maybe ten addresses, just enough to configure the weapon system.

I loved the way those old core memories used to 'sing' while in use. I wish I still had some to play with. I believe they were all hand knitted.

Brian.

 

[danny davis]

Re: Testing Diode leakage & transistor leakage

leakage is more likely to cause problems:
1. when very low currents are supposed to be present, the addition of an additional small current will tend to have more effect than in a high current circuit.
2. when very low voltages are present and are followed by high gain amplifiers. The small influence of the leakage may be magnified by the amplification.
3. in very high impedance situations where there may be no paths to safely drain away even the tiniest of leakage current.

Thanks that helps me out a lot , if you can think of any more , please add to the list would help me out a lot

 

[danny davis]

Re: Testing Diode leakage & transistor leakage

Like most things, if you get them too hot or too cold they suffer damage because of expansion and contraction beyond their stress limits. It tends to break the wire bonding inside the device, especially as the packaging may not have exactly the same expansion coefficient as the wire or semiconductor material.

After I been freezing and heating up the transistors, at first they are intermittent and flaky but after a couple of sprays between 1 to 5 sprays the Transistors will fix themselves and not be intermittent and flaky anymore

The Freeze and heating up the transistor must make its doping or chemicals go back to default

- - - Updated - - -

Also the services manuals say when i'm soldering or desoldering transistors metal can or IC metal can op amps that I have to use a SNAP ON heat sink

http://www.radioshack.com/product/index.jsp?productId=2103242

It says in the service manuals that the heatsink alligator clip don't work, you have to use a Heat Sink Snap On Tool

Why do I have to use a Heatsink alligator clip or snap on tool , when soldering or desoldering a Metal can transistor or op amp?

- - - Updated - - -

TO-220 CLIP-ON HEATSINK
http://www.allelectronics.com/make-a-store/item/HS-7139/TO-220-CLIP-ON-HEATSINK/1.html

- - - Updated - - -

http://www.jameco.com/1/1/323-ht-156-heat-sink-clip-2-7-inches-long-0-7-inches-wide.html

When do you use these? for which components please?
GT-CH51: Clip-on Heat Sink Absorbs Dissipates Heat During Soldering Desoldering Sensitive Parts
http://www.jameco.com/1/1/25581-gt-ch51-clip-heat-sink-absorbs-dissipates-heat-during-soldering-desoldering-sensitive-parts.html

 

[Tunelabguy]

Re: Testing Diode leakage & transistor leakage

Why do I have to use a Heatsink alligator clip or snap on tool , when soldering or desoldering a Metal can transistor or op amp?

The heatsink prevents the desoldering heat from traveling up the leads and heating up the device to the point where it is damaged. It depends on the length of time the heat is on too. If you apply heat high enough to melt solder at the circuit board, that heat travels up the lead and begins to heat up the chip. Since the chip has some thermal mass, it does not immediately reach the temperature of the solder. It might take it 10 seconds or so to get hot enough inside the chip to do any damage. And that depends on the length of the lead too. If the lead is very short, the chip will heat up much faster. If the lead is longer, the chip will heat up more slowly - even though the end of the lead at the circuit board is already hot enough to melt solder. If you are really quick, you can, in many cases, heat up the solder in a through-hole connection hot enough to **** out the solder with a solder sucker and then stop the heating, all without any heat sink clips. But if you fail to get the solder out right away, you should stop and let the whole thing cool off for a while before trying again. But using a heat sink clip of some sort gives you a little more time to do the desoldering. Maybe 30 seconds instead of 10 seconds. Don't assume that just because you are using a heat sink clip you can keep the heat on forever. Even heat sink clips eventually heat up too and then the chip will start getting dangerously hot. As to specifics of how hot or how long before damage happens, there are too many variables to say for sure. With experience you will eventually get an idea of how much heat you can use in any particular situation.