Fluctuating voltage from a bad transistor

[danny davis]

If a circuit is fluctuating voltage, how do you know which transistor or IC chip is fluctuating the voltage up and down or at random?

How would a troubleshooting tech, test and pin which transistor or IC that is cause this kind of problem?

What is this called when a transistor or IC is fluctuating the voltage? the transistors is not shorted or open , so what is this called? or what state is the transistor or IC chip in?

Using Freeze or a Heat gun on the transistor won't help troubleshoot this problem, I have tried that

How would you guys approach this issue?

 

[Audioguru]

The output voltage of an amplifier will fluctuate up and down making a motorboat sound (putt, putt, putt) when a power supply filter capacitor value becomes too low.

 

[danny davis]

a power supply filter capacitor value becomes too low.

If the filter cap is 470uf, u mean it changes the value of the filter cap, or you mean the voltage is low on the filter cap?

- - - Updated - - -

a power supply filter capacitor value becomes too low.

If the filter cap is 470uf, u mean it changes the value of the filter cap, or you mean the voltage is low on the filter cap?

- - - Updated - - -

some filter caps , filter out the ripple frequency and other filter caps storage the power supply voltage

I'm not sure how to tell the different between a filter cap that filters out the ripple frequency VS filter caps that store the power supply voltage

 

[chuckey]

If the reservoir cap (connected to the diodes) goes low in value, you will get a steady hum. If a power line decoupling capacitor goes low, then there can be low frequency feedback via the impedance of the PSU lines, this results in "motorboating". This often changes frequency if the volume or tone controls are adjusted. It used to happen when the batteries ran down on old transistor radios.
Semiconductors are the most reliable devices used in electronics, they can fail, but hardly every go intermittent. If there are "fluctuations" in the voltages in a circuit, I would look at the soldered joints first, especially if there are any large or long component soldered in. Due to vibration and temperature changes the component can put a physical strain on the PCB and pull the connection out of a soldered joint. Some times inspecting the underneath Of the PCB soldered joints, you can see a circle around a pin in a soldered joint. This can mean that the soldered joint has failed, resulting in an intermittent connection. Just touch up any suspect joint with a soldering iron. Once I fixed a colour TV, that intermittently lost the colour. When I put a soldering iron on the 4.4 MHZ crystal connections , and I actually heard it go "click" as the joint melted and the wire moved to its natural (unstrained) position,that fixed the fault!
Frank

 

[godfreyl]

The output voltage of an amplifier...

This being one of danny's threads, I doubt he's talking about audio amplifiers. Unfortunately he didn't say what sort of circuit he is talking about, or what voltage in the circuit is fluctuating.
:roll:

 

[danny davis]

Unfortunately he didn't say what sort of circuit he is talking about, or what voltage in the circuit is fluctuating.

Well it shouldn't matter, because if a transistor internally is causing fluctuating, what internally is wrong with the transistor? and what kind of state is the transistor in when it's fluctuating on the collectors output?

- - - Updated - - -

The Transistor is Partially Intermittent I guess you can say

1.) The transistor is either fluctuating voltage AC or DC at the repetitive pattern
2.) The transistor is fluctuating voltage either AC or DC at a random pattern

 

[betwixt]

I think we are all circling around a specific condition where the circuit has gone unstable but this is not what the original queston asked.

Yes, assuming the circuit is designed properly and worked before, an open circuit or high ESR capacitor can be the cause of fluctuating output but in that condition the fluctuation tends to be at a regular rate and pattern.

In this case though, I thing RANDOM fluctuations are what the question refers to and danny davis is asking how to troubleshoot it. There is no single answer in that case but the general approach would be to look for clues that narrow down the area of fault and then work backwards toward the start of the circuit to see which stage intruduces the fluctuation. For example, if it was a stereo amplifier the first test should be to see if one or both channels exhibited the problem. If it's only one it would eliminate the power supply and any other circuitry that was common to both chanels. Then work from where the fluctuation can be measured to earlier stages in the circuit, if the output of a stage fluctuates but the one before it is stable, that stage is the faulty one. There is a lot of 'gut feeling' in faultfinding, it isn't a set solution every time and with complex IC designs there is sometimes no option but to substitute parts or use a 'trial and error' tactic.

Brian.

 

[danny davis]

In this case though, I thing RANDOM fluctuations are what the question refers to and danny davis is asking how to troubleshoot it. There is no single answer in that case but the general approach would be to look for clues that narrow down the area of fault and then work backwards toward the start of the circuit to see which stage intruduces the fluctuation. For example, if it was a stereo amplifier the first test should be to see if one or both channels exhibited the problem. If it's only one it would eliminate the power supply and any other circuitry that was common to both chanels. Then work from where the fluctuation can be measured to earlier stages in the circuit, if the output of a stage fluctuates but the one before it is stable, that stage is the faulty one. There is a lot of 'gut feeling' in faultfinding, it isn't a set solution every time and with complex IC designs there is sometimes no option but to substitute parts or use a 'trial and error' tactic.

Yes I understand this approach this is basic 101 troubleshooting

What I'm asking is when a transistor is fluctuating , I'm not talking about a circuit or a transistor in circuit, i'm just talking about the transistor by it self , when it's fluctuating

1.) what internally is wrong inside the transistor?
2.) What causes a transistor to get like this?
3.) How would you know if a transistor is stable or fluctuating?

- - - Updated - - -

I would first use a function generator and inject a .5 volts AC sine wave on the input of the base of the transistor and than see if the output is fluctuating

How would you guys approach it to find the fluctuating transistor?

 

[dick_freebird]

I don't even know what you mean by "fluctuating". If this was
a real on-the-bench live situation, a picture would be worth
all the words to date.

All transistors make noise. But that ought to be trivial in most
circuits if all terminals are properly driven. An open base or
open gate, though, will respond to things you did not intend
or know about. Such as amplifying ambient EMI or self-
oscillating.

I wouldn't concern yourself with speculating about internal
mystery faults, until you've proven to yourself that another
device of same type doesn't do the same thing when you swap
it into the existing hardware.

 

[Audioguru]

A transistor that oscillates causes a fluctuating output voltage. Motorboating is an oscillation.

A transistor with a fluctuating output voltage might have a fluctuating input voltage (mains hum or interference pickup).

I have never seen a transistor that has something wrong with it that causes a fluctuating output voltage.

 

[betwixt]

They exist Audioguru - believe me!

There can be several failure mechanisms including:
destruction or damage to part of the si/ge crystal due to overheating or mechanical stress.
fracture at the welds due to stressing the pins or repeated hot/cold cycles.
corrosion, particularly along the wire of the pins, it sometimes reaches as far as the junctions. (Devices running hot can actually be more reliable because of this)
impurites in the junction materials
metal migration across pins and to the can if its made of metal.

Brian.

 

[FvM]

They exist Audioguru - believe me!

There can be several failure mechanisms including:
destruction or damage to part of the si/ge crystal due to overheating or mechanical stress.
fracture at the welds due to stressing the pins or repeated hot/cold cycles.
corrosion, particularly along the wire of the pins, it sometimes reaches as far as the junctions. (Devices running hot can actually be more reliable because of this)
impurites in the junction materials
metal migration across pins and to the can if its made of metal.

Yes of course, this happens. The problem with this thread is, that the OP is essentially guessing a transistor or IC is causing the fluctuations, but he didn't tell why he comes to the conclusion.

So all we can say, yes it may be like he guessed, or completely different.

 

[danny davis]

No, its the transistor for sure because I have to change the bad fluctuating transistors

I'm saying if you are testing 100's of the same boards and each board has 50 transistors on it and one of them is fluctuating but you don't know which one out of the 50 is doing it , how about you find it and when troubleshooting skills and techniques would you do to track the bad transistor?

Would you start by doing what?

 

[betwixt]

Use the technique in '101' basic troubleshooting I gave in post #7. There is no magic finger that points to faulty parts, you have to use judgement and to some degree guesswork. You have never mentioned what type of boards or transistors you are talking about but the same basic rules apply - the fluctuations will start somewhere and appear downstream (after) that point but less likely before it.

If you have lots of identical boards, you could set a good one and a bad one up in parallel and try comparing voltages between them but in the case of sequential logic circuits you have to be very careful that all the states are set the same. If it's an amplifier or oscillator, you probably need an oscilloscope to see differences but if you are repairing 100 boards of any electronic system I would suggest an oscilloscope was essential to have anyway.

Brian.

 

[FvM]

No, its the transistor for sure because I have to change the bad fluctuating transistors

because what?

 

[godfreyl]

because what?

He knows the transistors are bad because he has to change them.

 

[danny davis]

Yes , the senior tech has to find them and I change them , so it's the transistor for sure
The hard part is find which one is bad

the base to emitter for an npn transistor is not. 7 volts in circuit is this true?

For pnp transistors u have to reverse the probes to get the. 7 volts?

I should turn the circuit OFF no power on and use the DVM diode mode when testing a transistor base to emitter in circuit?

Is there any way to measure the transistors .7 volts across the base to emitter , when the transistor is in circuit?

Because mostly the transistor is "energized" and past the turn on/forward biased voltage

So there is no way for me to measure the .7 volts right across the base to emitter?

Is energized a good work to use when saying it's been forward biased and turned on?

You can't measure the .7 volts when a transistor is PNP because it's only .7 volts from base to emitter using NPN transistors, is that right?

 

[godfreyl]

the base to emitter for an npn transistor is not. 7 volts in circuit is this true?

The base to emitter voltage is 0.7V in circuit.

For pnp transistors u have to reverse the probes to get the. 7 volts?

Yes, that's right.

Is there any way to measure the transistors .7 volts across the base to emitter , when the transistor is in circuit?

Yes. The circuit must be powered on. Set the multimeter to read voltage then connect one probe to the base and the other probe to the emitter.

I should turn the circuit OFF no power on and use the DVM diode mode when testing a transistor base to emitter in circuit?

That's another way to do it. Both methods should work.

You can't measure the .7 volts when a transistor is PNP because it's only .7 volts from base to emitter using NPN transistors, is that right?

You can measure the 0.7 volts from base to emitter of a PNP transistor the same way as for an NPN transistor. The only difference is you have to reverse the probes.

 

[danny davis]

Thanks for the info.

the base to emitter for an npn transistor is not. 7 volts in circuit is this true?
The base to emitter voltage is 0.7V in circuit.

Even when the base has signal on the input of the transistor? it will still be 0.7 volts?

I thought the resistors around the transistor biased the transistor and energized the transistor when an input signal was applied to the base or input of the transistor , so testing the transistors base to emitter you won't get 0.7 volts because of the DC offset from the resistors that are biasing the transistor right? plus the circuit is powered on

Mostly I don't get 0.7 volts from base to emitter it in the volts because it's in circuit right?

If you use a PNP transistor as a switch your don't have to reverse the polarity of the VCC right?

OR I think they use the PNP
1.) to invert a positive DC voltage to a negative DC voltage
2.) to invert a HIGH logic to a Low Logic

But I don't think they power the PNP with a negative power supply , i think it's still using a positive + 5 volts VCC

What else have you seen circuits using PNP transistors for?

 

[godfreyl]

Even when the base has signal on the input of the transistor? it will still be 0.7 volts?

More or less. It may be 0.6V or 0.8V, or somewhere in between.

I thought the resistors around the transistor biased the transistor and energized the transistor when an input signal was applied to the base or input of the transistor , so testing the transistors base to emitter you won't get 0.7 volts because of the DC offset from the resistors that are biasing the transistor right? plus the circuit is powered on

Yes, the resistors bias the transistor to energize it, but they do that by putting about 0.7V between the base and collector. So the DC offset from the resistors is the 0.7V that switches the transistor on.

What else have you seen circuits using PNP transistors for?

PNP transistors can be used for all the same things as NPN transistors - amplifiers, oscillators, constant current sources etc.

PNP transistors work exactly the same as NPN transistors, except the voltages are reversed.