Hi,

I am building a transister test rig to check what the gain would be at specific voltages and specific currents. I know the formula, but as each transister is different I want to know what the gain would be of each particular one before I stick them in a circuit.

I have attached a basic diagram of what I am goin to build, there would be more than one resister to switch between on the actuall test rig to work out the gain difference at different Amps/Voltages.

The voltage source is a DC variable power supply I have.

The main questions I have are

1: Is there anything I am missing that would not make it work correctly?

2: Is there anything I can add to make it more accurate?

Thanks

 

There is always the curve tracer - www.techdoc.kvindesland.no/radio/complex.../20051028154625797.pdf

Your diagram did not attach.

 

Couldn't figure out how to attach to previous post, diagram should be attached now.
Of course will attach Amp meter to circuit to get readings, just didnt think it needed to be in this diagram.
Got a could not find messege after clciking on your link.

 

The transistor is going to get very hot very quickly when the base drive switch is closed so don't dally around making your readings.

You do realise the actually heating will affect the gain of course?

 

Yea, I have a nice variety of heatsinks with fans on them ready to keep the transistors cool and even a temp prob with a digital display to make sure it doesnt get to hot.

 

Your circuit is a thermometer. As the transistor heats it turns on more.
Transistor manufacturers test transistors with pulses of current that have an extremely short duration so they do not begin to get warm.

 

That is what a curve tracer for, preventing the transistor to heat up during measurement. The standard pulses for measurements are either 300us (common) or 80us.

 

ok but, what you are saying is why I want to test it the way I am. They only test it in pulses, but if a transister is used so it never goes into cutoff or saturation, but varies the current inbetween the 2 then I need to know what the gain would be in real life use with a heatsink attached, running at real life temps.

If I was using it for a pulsed circuit goin into cutoff and saturation then I would just use the graphs, but as im not, I need to know what it actually does outside of a lab test.

 

If you design a transistor circuit properly then the temperature does not affect it.
But if you design the circuit poorly then it might not work if it is cold or hot.

Why test the hFE gain of each transistor? Transistors with one part number have a range of hFE. Some have a low hFE and others have a high hFE.
You are supposed to design a circuit that works perfectly with a transistor at any temperature that has any amount of hFE that is specified in its datasheet.

 

basically I am building a circuit with 6 boards, each one needs to have its own current limiter. The transisters I will be testing will be the main part of the current limiter circuit. There can be a slight variation between them, but if there is to much then it wont work, so I need to know what the actuall gain of each transister will be to change the current goin into the base of them so they are at least close to each other.

Plus I am wondering why everyone keeps goin on about the temp? I know the temp can have a impact on it, which is why I said I am adding the heatsink and fan. Its a transister that can handle 15A and im only goin to be using it upto 2A. So with the heatsking and fan I dont think that will be a problem.

 

Unlike Mosfets, Bijunction transistors are prone to a phenomenon called thermal runaway.

This is because the gain/temperature graph shows a positive slope. So the more the current heats the transistor the more the gain increases, increasing the current.... until the the BJT self destructs.

Mosfets on the other hand have a negative gain v temperature slope.

 

So what everyone is saying is that only use these transisters for circuits that pulse them off and on?

because what your implying is that you can only use a transister as a switch otherwise it will overheat!

If thats the case then every website needs to change what they think a transister can do.

 

A transistor overheats only if your design makes it operate too hot.
We said that temperature affects the hFE gain of a transistor and its Vbe.

You should design a circuit so that the temperature and Vbe do not affect how the circuit works.

 

I know that, that is why I said I have a heak sink and fan to keep it cool, and a temp probe to make sure it doesn't get to hot.

The thing im confussed about is, the subject of the temp has been mentioned, I said I would keep it cool and monitor the temp. But still, everyone keeps goin on about the temp part of it.

The temp was not a specific question I asked about, and when it was mentioned to me, I gave a reasonable answer that I have cooling in place and a monitor for it. But still everyone keeps goin on about the temp.

If someone would please help me with my circuit instead of repeating the same thing I have already answered ( apart from putting the thing in a freezer! ) I would be most gratefull, as the main purpose of me posting this thread has been lost for the discussion of what will happen if you don't keep a transister cool enough!

 

Well, with only one base current available, the rig looks more like a stress tester. At least 3055's have a great ability to run at just under red heat and still keep functioning.

We have all had to deal with the effects of thermal runaway in semiconductors. Avoiding it like the plague is second nature anymore.

To be able to do some differentiation between transistors, though, you will need to vary some part of the rig. Like several different base current limiting resistors to show a couple of levels of gain.

 

Indeed, I am planning to have a selection of resistors values to switch between. I did mention it on my origanal post, but maybe wasn't that clear.

I am goin to have a selection of resister values connecting to the base and the collector pin to have a wide range of testing.

I am wondering as the circuit seems a bit to simple, is there anything missing from it that would make it not give at least a rough reading, or anything I could add that would be beneficial?

 

I fully understand what you are attempting and you have had your answer.

It will work and supply your measurements. I can only assume you have worked things out to test around your intended operating conditions. This is good and sensible. Further you do not need pulse testing, particularly if your intended application is for instance an analog power regulator.

The reason many testers use(d) pulse testing is because providing a reliable continuous test supply of several amps is(was) no mean feat and added considerably to the complexity and cost of the tester. My old Philips Instruments transistor tester provides up to 3 amps at a variety of voltages, and weighs (and cost when new) a ton.

However we are only trying to help here, so don't take umbrage, BJTs and in particular power BJTs are prone to thermal runaway. So much so that cooling alone is not enough. You need also to provide monitoring and limiting circuitry in your design against this. Such will not be necessary for short term tests, that you will be monitoring in any case.

 

I dont understand why you are saying " you have had your answer " and dont take umbrage "?

I am just looking for help on my circuit, the problem of the temp has been mentioned to me, but as no solution was been mentioned, I tried to change it back to my origanal post!

I am guessing maybe its a culture difference, so I wont say any more about it.

But this transister will be used at these levels in the end circuit for long periods of time, so if you could advise of possible protection circuitry or something different, maybe a diff transister? that could handle 2-3A running through it I would be most gratefull.