Simple BJT bias but simulated results aren't desired.

Hi all,
I designed BJT circuit using BC547 with conditions: supply voltage: vdc=3.95V; Vbe=0.745V; Vce=0.4V; Iled=8mA;Ic=10mA. Based on them, I calculated value of devices as figure. But, when I simulate on Proteus; results are very difference. I don't know reasons; so I can't adjust values ( resistor) to my expectations. I'm looking for your opinions. Thanks

- - - Updated - - -

https://obrazki.elektroda.pl/5594883900_1471361333.png

HI,

Q4 is is saturated.

A design to extremes is difficult to anylze precisely.

Every change of a single value (e.g. Supply voltage) will have huge influence on all other values.
In reality even the temperature dirfts of the PN junction will cause huge changes in values.

Klaus

Datasheet shows only typical values of Vbe and Vce, with wide variation range. It's pointless to calculate an expected operation point in this circuit exactly.

And vdc in your simulation is apparently different from claimed 3.95 V.

KlausST said:

HI,

Q4 is is saturated.

A design to extremes is difficult to anylze precisely.

Every change of a single value (e.g. Supply voltage) will have huge influence on all other values.
In reality even the temperature dirfts of the PN junction will cause huge changes in values.

Klaus

Click to expand...

You talked exactly. I want't Q4 is in saturation region. Because I want Led is on when Vdc>= 3.95V and Led off when vdc<3.95V. I tried to changed resistors but results aren't my desire.

- - - Updated - - -

FvM said:

Datasheet shows only typical values of Vbe and Vce, with wide variation range. It's pointless to calculate an expected operation point in this circuit exactly.

And vdc in your simulation is apparently different from claimed 3.95 V.

Click to expand...

Actually, I designed base on figure about relationship between Ic with vbe, vce (showed in datasheet). Besides, Vdc is 3.95V, but it isn't show in figure (because i đin't put voltage marker at there). It is difficult to explain.??

Hi,

If you just want to switch ON the LED, then R17, R18 and Q4 are about useless.
Connect another 1N4148 instead of R18 and get almost the same results.

Btw: in case the input voltage increases a bit, your LED will soon be killed by overcurrent.

Klaus

Besides, Vdc is 3.95V, but it isn't show in figure (because i đin't put voltage marker at there).

Click to expand...

I know that the funny Proteus simulator doesn't show voltage sources by default. Thus I calculated the voltage drop across the 30 ohm resistor and found that Vdc can't be 3.95 V. Or there's current flowing into other components omitted from the schematic.

More generally speaking, you shouldn't expect that the Proteus BC547 model is exactly behaving according to the (typical) characteristic curves in your data sheet. It hopefully fits the specified operation points.

But even if the model is exactly complying to the data sheet specification, type variations indicated by the min and max columns can thwart your calculations, also a different ambient temperature.

KlausST said:

Hi,

If you just want to switch ON the LED, then R17, R18 and Q4 are about useless.
Connect another 1N4148 instead of R18 and get almost the same results.

Btw: in case the input voltage increases a bit, your LED will soon be killed by overcurrent.

Klaus

Click to expand...

https://obrazki.elektroda.pl/6083992700_1471446821.png
This is overall circuit which I'm designing. I want to green led on when voltage is above 3.95v and off when voltage is under 3.95V. Meanwhile, red led is on when voltage <3V and off when voltage > 3V. Moreover, voltage are always smaller than 4.2V; so I don't care much about over current to led.

- - - Updated - - -

FvM said:

I know that the funny Proteus simulator doesn't show voltage sources by default. Thus I calculated the voltage drop across the 30 ohm resistor and found that Vdc can't be 3.95 V. Or there's current flowing into other components omitted from the schematic.

More generally speaking, you shouldn't expect that the Proteus BC547 model is exactly behaving according to the (typical) characteristic curves in your data sheet. It hopefully fits the specified operation points.

But even if the model is exactly complying to the data sheet specification, type variations indicated by the min and max columns can thwart your calculations, also a different ambient temperature.

Click to expand...

https://obrazki.elektroda.pl/4930126200_1471447832.jpg
Now, I'm studying about structure of BJT; I'd like to know why and how can I fix it. I'm looking forward your opinion.

Hi,

Why don´t you give the whole circuit in the first post?
For sure with the new circuit the bjt makes sense...

*****

Moreover, voltage are always smaller than 4.2V; so I don't care much about over current to led.

Click to expand...

Maybe you should..
a rise in voltag from 3.95V to 4.2V may increase the LED current from 20mA to max. 28mA. Can your LED withstand this? We don´t know..

*****
Your circuit lacks on noise immunity and precision.
A mV on voltage change may cause your LEDs to flicker. Even the change in current when it switches from green to red may cause oscillations. (depens on circuit and currents)
A change in ambient temperature of 10°C may change the threshold voltage by 60mV and the green LED current by 2mA...maybe this is tolerable for you, maybe not.

*****
You write "simple".
Maybe a three pin device like LM809 can do the job...(more precise, more noise immune)

Klaus