I have a KST42 BJT npn transistor and im trying to figure out how to calculate the offstate transistor impedance to figure out the total offstate current draw. There are many transistors on this circuit board and the impedance will add up to have somewhat of a offstate current draw. I've used a ohmmeter but each terminal of the BJT in any combination of polarity has a resistance in the kilohm range.
Thanks for your help in advanced.
-Vince

 

i've also looked at the spec sheet and couldnt not find anything about the offstate impedance.

 

Very strange, look here:
http://www.fairchildsemi.com/ds/KS/KST42.pdf
Collector cut-off current 0.1uA.

 

thanks for your reply.
How would you relate that to the offstate impedance?

 

very very high, so you don't need to worry about this anymore.

 

Just to know this in the future how would you carry out the calculations accurately. And would resistors connected to the base and collect be added in parallel or in series to the transistor impedance?

 

Well modern small-signal BJT have very small leaking current.
So I really don't understand your problem.
Can you describe in more detail your problem, or what you want to achieve.

 

The problem is that i dont want the off state current to introduce bias to the amplifer. Theres about 100 transistors on this board and i dont know if this will affect the amplifier.

 

Im trying to get a feel for how much error there would be.

 

But amplifiers don't work in cut-off.
In classic linear amplifier all BJT work in active region.

 

"Small" is a relative word. Sometimes leakage of a microAmp or nanoAmp is significant.

I would do this....

Emitter to ground. Base to Ground. Some voltage (2 to 6V or so...) through a 1 MegaOhm resistor to the Collector. Then measure the voltage over the resistor. Note that the multimeter may affect the measurement at this impedance, so know your multimeter. (I have one that has extremely high impedance that is good for this.)

That can show your leakage current. You may vary the voltage and see if that affects it much, too.

I have found some silicon BJTs with significant leakage current, and Ge transistors with pretty high leakages. Some MOSFETs leak a microamp... It adds up if you have a tight power budget or working with small signals.

 

can the added current from these off state transistors be enough to bias the amplifier differently from what is intended?I have an LM2902M op amp. All these transistors (both PNP and NPN) are connected into a multiplexer that chooses which voltage to supply to the amplifer.

 

the multiplexer output is connected to 2 LM2902M negative terminals and the output of the amplifier is connected to another LM2902M but this time each output is connected to a positive and a negative terminal of the 3rd amplifer. The 3 amplifiers have negative feedback.

 

Thanks for the post sage, but i do not understand how that would help calculate the leakage current. Sorry if im a bit dense, i have minimal experience in real application.

 

Thanks for the post sage, but i do not understand how that would help calculate the leakage current. Sorry if im a bit dense, i have minimal experience in real application.

If you want to measure the collector/emitter leakage current, which is how much current flows when the transistor is supposed to be off, then you shut off the transistor and measure how much current is leaking through.

You measure current as a voltage drop across a resistor, so that's how I set up the little experiment.

I've done this for MOSFETs in the past for ultra-low-power electronics. One reason is simply to know how much current is leaking, wasting power. Another reason is that when I have used large resistor values for pull-up and dividers, I need then to account for even the 1 uA of leakage current, because it will affect the resulting voltages.

 

Sage may i also ask why i need to have a very large resistance such as 1 Mohm?

 

Sage may i also ask why i need to have a very large resistance such as 1 Mohm?

Well, I expect that leakage is probably pretty small, so to measure it, it is useful to use a larger resistance. Suppose you have a 6V power supply and you run that (+) through a 1 MegaOhm resistor to the drain of an N-channel MOSFET and you have the MOSFET gate tied to ground, so you expect that it is off.

If you measure 1.0V over the 1 MegaOhm resistor, that means you have 1 uA (microAmp) of leakage in the off state. If you see 0.2V over the resistor, you have 0.2 uA of leakage.

Leakages are typically pretty small. If you find that the 1 MegaOhm is stretched from the (+) to the ground, then you have a larger leakage and you must then change to a smaller resistor value, like maybe 100K Ohms, and try again.

Last thing, let me repeat to be careful what multimeter you use for this, because some multimeters have an input bias current that will affect the readings. I have a great multimeter from Radio Shack that is perfect for this, the "22-820". But if you don't have a super-high-impedance multimeter then you could also buffer the measurement with a capacitor. You can run, say, a 10 uF capacitor from ground to the drain of the transistor. Then let it reach its equilibrium for about 10 seconds, and then make your reading. You should see the voltage get lower as the multimeter pulls current through the sense resistor, but you can catch the first number and it is pretty close to accurate.

 

What do you mean by "1 MegaOhm is stretched from the (+) to the ground" are you supposed to connect the resistor from collector to ground or from the power source to the collector. When the transistor is off does the Ic still equal to Ie? By using a resistor are we not inducing a certain value of current into the off-state transistor?

 

I still dont understand the concept of choosing a particular resistance. Say if we put a 6 V power source in the collector and a 1 megaohm resistor in the collector we get a 6uA current even when it is off right? So can we change the current by setting different resistances?

 

I believe that these are the questions that are preventing me from understanding. Sorry Sage....