[SOLVED] poor stability in a bjt

[jeffrey samuel]

why is their great variations in the β value and the collector current when bjt is used for a prolonged period of time

 

[jasonc2]

One likely cause is temperature-dependent effects as the transistor heats up.

See the charts on page 3 and 4 of the 2N3904 datasheet for example. What kind of variations are you seeing and what transistor are you using?

 

[jeffrey samuel]

in bc107 when i used for a period of 15 min i found variations in the practically measured base and collector current thereby the effective gain is shifted causing a shift in the op of amplifier

- - - Updated - - -

can using a biasing method be the cause for it

i tried fixed bias method

will there be any difference if i used the emitter resistor method

 

[FvM]

As a first point, I want to emphasize jasonc2's clarification: We are talking about temperature and possibly supply voltage dependent OP variation, not time dependent (e.g. aging).

With this clarification, yes amplifier circuits have quite different sensitivity to temperature and voltage variations (and furthermore component value tolerances, which is a somewhat related problem). If possible, you'll want to use a circuit with effective internal DC feedback that has low sensitivity to the said variations.

You'll find the topic discussed in analog design textbooks.

 

[jeffrey samuel]

thank you FvM
but ic can be maintained constant even if temperature varies over prolonged time in emmiter resistor method right

so it can be considered as replacement for the design or will i find the same variation

also tell me if replacing the device (fet / mosfet ) instead of transistor will have any impact on the gain factor
as am yet to try it

do clarify

 

[etmabreu]

Gain variation is a problem you have to live with. You can even have huge differences (say, a factor of 2 or more) if you change your BJT for another one of the same type and model!

Usually we get around this with proper bias.

How are you biasing your BJT? Can you post a schematic?

 

[jeffrey samuel]

i merely used a fixed base resistor that is all

i guess that might be faulty planning on using emitter resistor next time

but still there are theoretical variations when doing the same

 

[FvM]

To avoid any temperature related variation, you'll need compensation or differential design, which is often necessary for DC amplifiers. For AC coupled designs, some variation doesn't hurt.

 

[jeffrey samuel]

so instead of compensation circuitory can we replace transistor by jfet

- - - Updated - - -

in case of dc amplifiers i mean

 

[FvM]

JFETs have a temperature dependant operation point, too.

 

[Darktrax]

In both FET and bipolar designs, having some resistance in the source, or emitter circuit automatically stabilizes the current against temperature-related drift, it being a form of dc feedback. As the current rises, the device tends to bias off, countering the rise. This resistance can be partly, or completely bypassed with a capacitor to restore the gain at higher frequencies.

Even low resistances less than 100 ohms at the emitter can work wonders in stabilizing against temperature drift.

 

[D.A.(Tony)Stewart]

JEFFREY< If you drive a current amplifier with a voltage source say 0.7V , you get huge variations in base current due to the Vbe vs temperature Shockley effect. and thus the drive current changes over time due to input bias problem if that is what you did. -5mV/deg C as I recall.


... Feedback and self bias resistors do "work wonders" to stabilize voltage gain for wide variations in current gain.

When high current loads are applied the gains drop quickly and especially when used as a saturated switch.

Then you can expect 5~50 depending on device and output current, whereas hFE you can expect 50~500 for different types at low current.

 

[FvM]

Some bias variants with different sensitivity to temperature, current gain and supply voltage variations.