Transistor as a switch

[champ1]

I am trying to understand how NPN transistor BC547 work as switch.

Transistor can be three state
Cut off
Threshold
Saturation

I'm trying to understand when the BC457 turns on or off

 

[crutschow]

I'm trying to understand when the BC457 turns on or off

The base-emitter junction looks like a forward biased diode, so it starts to conduct significant current when the base-emitter voltage reaches about 0.6V.
At that point the transistor is conducting collector-emitter current with a Beta (current gain for base current to collector current) ranging from 110 to 800 (per the data sheet), depending upon the particular transistor you have.
When the base current reaches about 1/20 of the collector current (as determined by the collector load resistance and voltage), then it is fully saturated ON as a switch.

To turn it off, you reduce the base voltage and current to zero.

Make sense?

 

[AnalogKid]

It turns on or off based on a combination of the base current (current moving into the base and out the emitter) and the base voltage - more correctly, the voltage across the base-emitter junction. The datasheet has tables and charts showing the relationships.

The base-emitter junctions acts as a standard signal diode. When the voltage across it exceeds the forward conduction voltage (Vf), unlimited current moves through it. Usually, the current is limited to a safe value by a resistor or combination of resistors and other components in series with the base.

When the junction is conducting, the current into the collector and out the emitter is proportional to the base current. The proportionality is called the gain h(fe). As shown in the charts, the gain is nowhere near constant. It varies as a function of temperature, collector current, the voltage between the collector and emitter, atmospheric pressure (not much, but still ...), etc.

ak

 

[champ1]

My guess is that if i give a 5v supply across the Base to emitter junction, transistor will be damaged.

 

[Harald Kapp]

We have a resource on this.

My guess is that if i give a 5v supply across the Base to emitter junction, transistor will be damaged.

Without base resistor: yes.

 

[ramussons]

Keep in mind, that all Bi Junction Transistors, NPN or PNP are Current Controlled Devices. The Base current.

 

[CircutScoper]

My guess is that if i give a 5v supply across the Base to emitter junction, transistor will be damaged.

Current flow into the base/emitter junction of a bipolar transistor is an (approximate) exponential function of the applied voltage, increasing by roughly a factor of 10 for every 60mV above threshold, which for silicon is roughly 0.6V.

So yeah. Since 10^((5 - 0.6)/0.06) = 10^73, it would be fairly reasonable to predict that something noticeably dramatic would happen if 5V were applied to the base. (!!!)

Something like: Say -- wasn't there a transistor here a (nano)second ago?

 

[crutschow]

If you understand what happens when you put 5V in the forward direction across a silicon junction diode, then you will know what happens when you apply 5V to the transistor base-emitter junction in the forward direction.

 

[ivak245]

Resistors are your friend here. You could put a trimpot in the base circuit, then measure the collector current as the base CURRENT is varied. At some point your transistor will get very hot and let all the smoke out, this is the best way of understanding how they practically work. This is what we did in the lab at school, just make sure you have your safety glasses on!

 

[CircutScoper]

Resistors are your friend here. You could put a trimpot in the base circuit, then measure the collector current as the base CURRENT is varied. At some point your transistor will get very hot and let all the smoke out, this is the best way of understanding how they practically work. This is what we did in the lab at school, just make sure you have your safety glasses on!

Ah yes -- the classic and ever popular "smoke test." I was also fond of a similar criterion: When the leads become incandescent.

 

[danadak]

Keep in mind, that all Bi Junction Transistors, NPN or PNP are Current Controlled Devices. The Base current.

Transistors, bipolar, are V controlled devices.

The classic relationship of I and V is (Ebers-Moll) -



Regards, Dana.

 

[Bluejets]

https://toshiba.semicon-storage.com...n, the collector,considerably varies with VCE.

Basically, a bipolar transistor amplifies a small current entering the base to produce a large collector current. It is a current-driven device since the collector current is controlled via the base current. The current gain varies with the collector-emitter voltage (VCE). In the active region shown in the right-hand figure, a bipolar transistor provides a gain called a DC current gain (hFE) . In this region, the collector current remains almost constant regardless of the collector-emitter voltage (VCE). On the other hand, in the saturation region, a bipolar transistor exhibits a DC current gain of only 10 to 20, where the collector current considerably varies with VCE.

 

[danadak]

I think the device physics support its the field in the Vbe junction that controls
the transport across the base..

The region look like -

As pointed out here the current controlled model is an over simplification of the bahviour. Simulators
use the Eber-Moll model for accuracy, not the "beta/alpha G" model.

https://electronics.stackexchange.com/questions/252791/why-is-a-bjt-considered-current-controlled

This will give you migraines -

https://ia902704.us.archive.org/32/...uctorDevices3rdEdition-S.M.SzeAndKwokK.Ng.pdf


Regards, Dana.

 

[crutschow]

Transistors, bipolar, are V controlled devices.

True, the theoretical physics of the BJT shows it as a voltage-controlled device.
And I imagine some of the pedantic types gnash their teeth whenever anyone says they are current-controlled.
But for analog designs, hand calculation of the bias points is generally more easily done if a black-box current-control model is used, including the Beta current gain, and viewing the base-emitter junction as a forward-biased diode.
There's a good reason the value of the BJT current gain is usually shown in their spec sheet.

After that you can use a simulator that uses the voltage-controlled model to more accurately determine the bias values (although curiously, the Spice BJT model parameters do include a Beta parameter).

Also I think the current-control model makes it easier to understand transistor operation as a switch, since the conditions for saturation use a value for the base current, not the base voltage, to insure saturation.

 

[Harald Kapp]

Please, guys, we've had this discussion at length before. While the equations use V_BE, in practice it is mostly much easier to control the base current (and let V_BE adjust by the transistor's properties).
Repeating this discussion will not help the op.

 

[danadak]

True, the theoretical physics of the BJT shows it as a voltage-controlled device.
And I imagine some of the pedantic types gnash their teeth whenever anyone says they are current-controlled.
But for analog designs, hand calculation of the bias points is generally more easily done if a black-box current-control model is used, including the Beta current gain, and viewing the base-emitter junction as a forward-biased diode.
There's a good reason the value of the BJT current gain is usually shown in their spec sheet.

After that you can use a simulator that uses the voltage-controlled model to more accurately determine the bias values (although curiously, the Spice BJT model parameters do include a Beta parameter).

Also I think the current-control model makes it easier to understand transistor operation as a switch, since the conditions for saturation use a value for the base current, not the base voltage, to insure saturation.

Pretty much agree. Its interesting though, we constantly complain about the incompleteness of
various spice models, yet here we (includes me) sing the praises of an incomplete model.

Like the rule of thumb we use to sat a transistor Ib = Ic / 10, but this one does not always comply -

The human condition.... Destined towards extinction....


Regards, Dana.

 

[crutschow]

Just trying to minimize confusion in the TS's understanding about the theoretical operation of a BJT versus the practical.

Like the rule of thumb we use to sat a transistor Ib = Ic / 10, but this one does not always comply -

Not sure why you say that(?).
The graph does show saturation voltage is greater at higher currents when using a Ib = Ic / 10, but that's mainly due to the intrinsic resistance of the collector-emitter when on.
Of course, at high currents, close to the transistor limits, the beta usually does go down significantly.

 

[danadak]

Just trying to minimize confusion in the TS's understanding about the theoretical operation of a BJT versus the practical.
Not sure why you say that(?).
The graph does show saturation voltage is greater at higher currents when using a Ib = Ic / 10, but that's mainly due to the intrinsic resistance of the collector-emitter when on.
Of course, at high currents, close to the transistor limits, the beta usually does go down significantly.

I say that Ib = Ic / 10, the terminology is forced beta, and many datasheets spec Vcesat
at a forced beta of 10.

From the 2N3904 datasheet -




Regards, Dana.

 

[crutschow]

I say that Ib = Ic / 10, the terminology is forced beta, and many datasheets spec Vcesat
at a forced beta of 10.

Okay.
But I still don't understand your statement:
"Like the rule of thumb we use to sat a transistor Ib = Ic / 10, but this one does not always comply -"
How does it not comply?

 

[hevans1944]

I am trying to understand how NPN transistor BC547 work as switch.

Transistor can be three state
Cut off
Threshold
Saturation

I'm trying to understand when the BC457 turns on or off

What you have done is jump-start a veeery old thread that attempted to discuss how a transistor works. The thread became a tug-of-war between those who said the "transistor is a (base) current-controlled device" and the "transistor is a (base-to-emitter) voltage-controlled device. Nothing was "solved" to the satisfaction of everyone who participated in the thread. Eventually the thread died, and now I cannot even find it in the archival records. The new owners of these forums are probably not anxious to open the discussion again.

There are many states and levels of conduction that a transistor can be in, between "Cut off" and "Saturation". There is no "Threshold" state that I am aware of. The BC457 NPN transistor will "turn on" whenever the base-emitter junction is forward biased with a voltage that makes the base positive with respect to the emitter. Note that some collector current occurs even with small amounts of forward bias. There is no magic "threshold" that must be crossed before the transistor will turn on.

The transistor will "turn off" whenever the base-emitter junction voltage is zero or slightly negative. Meanwhile, the collector-base junction is reverse biased as the transistor begins to conduct in its collector circuit. With a series-resistance (or reactance) load in the collector circuit, the collector-to-emitter voltage will decrease with increased conduction until the transistor saturates: further increases in the base-emitter forward-bias voltage does not cause any significant increase in collector current when the transistor is in saturation.

When using the BJT as a switch, it is desirable that the saturation voltage be very small (usually in the neighborhood of 0.1 volts) and the cut off (collector) current also be very small (usually in the neighborhood of a few dozen nano-amperes.) To accomplish saturation will require an input signal voltage well above the 0.6 to 0.7 volt "knee" in the transfer function between base-emitter bias voltage and collector current. Apply the base-emitter bias voltage through a fixed-value resistor that limits the base current to a value of roughly i_cc/10, where i_cc is the saturation current of the transistor. It is a good idea to insert another, larger-valued, fixed resistor between the base and the emitter to guarantee cut off when no signal is applied.

One final comment on the use of a transistor as a switch: in general, a MOSFET switch will accomplish the switch function faster, better, and cheaper than the BJT implementation. You should look into that.