AutoBiasing Transistors

[UTMonkey]

Hi Nigel,
1+2 no problem.
3. You say 5, KChriste says 10. as long as I am between those values am I ok?

 

[audioguru]

Design the circuit for the typical hFE of 200. Then calculate the voltage at the output when the transistor has a minimum hFE and when it has a maximum hFE to make sure the circuit still works.
When you buy these transistors you can't order the hFE amount, you get whatever they have low, medium or high.

Some transistors have selected reduced ranges of hFE. The BC547A has fairly low hFE, the BC547B has medium hFE and the BC547C has high hFE. They are selected by the manufacturer so they cost more than a BC547 that has a range of hFE from 110 to 800.

 

[Nigel Goodwin]

3. You say 5, KChriste says 10. as long as I am between those values am I ok?

10 times is better - but 5 times is about the bare minimum - bear in mind the lower value the resistors you use, the lower the input impedance will be.

 

[UTMonkey]

Thanks everyone.

AudioGuru, you asked me to calculate the transistor at 100 and 300. How does my Hfe of 100 look?
If i calculated for 300 what may i expect to see?

What is interesting that with the "improved" Hfe=200 circuit you referred to a gain calculated by Rc/Re = 400/150 (ish) thats just over 2 - nothing like 200.

Am I not grasping this properly?

Best Wishes.

 

[audioguru]

The value of your emitter resistor is so high that it reduces the voltage gain of the circuit and makes the output voltage change only a small amount when the hFE is changed.

Connect an electrolytic capacitor across the emitter resistor then the voltage gain is equal to the collector load resistance (including the external load in parallel) to the internal resistance of the transistor. A voltage gain of about 160 and very bad distortion at high level.

 

[kchriste]

UTMonkey,
You can use a lower value emitter resistor and adjust the bias resistors so the collector is still at 1/2Vcc and get a reasonable gain with a reasonable distortion figure. This could also be done with two emitter resistors in series and a electrolytic capacitor across one of the emitter resistors. The reason for adding a emitter resistor and dropping the gain is that the transistor's internal Re is nonlinear and introduces distortion. So we add a swamping resistor in the emitter to reduce the distortion.

What is interesting that with the "improved" Hfe=200 circuit you referred to a gain calculated by Rc/Re = 400/150 (ish) thats just over 2 - nothing like 200.
Am I not grasping this properly?

There is a difference between the DC beta (hFE) and the Small Signal Current Gain (hfe) of a transistor. Without a swamping resistor, the AC voltage gain is dependent on the emitter current (which controls the transistor's internal Re) and the load resistor on the collector.

 

[UTMonkey]

Wow, ok I definitely won't claim to understanding what you meant.

The emitter I calculated is based on the PD voltage -0.7 divided by collector current of 10mA.

If I was to reduce Re to something like 15 ohms that would mean that 150mV would be fed into Re (15 ohms x 10mA) which would also mean that the PD would have to bias 850mV?

But even if I was able to do that wouldnt it mean a gain of (Ic = 400 / Ie 15) of 26?

hFE? hfe? what is it I have been working out?

I think I am close thanks to everyones help but those last two messages have thrown me slightly.

Best regards

 

[UTMonkey]

I also took the values from the "new" 200 gain calculations, coupled the input with 10uF cap and put a 500mV AC signal through it.

Look at the screen shot of the results.

Any comments?

 

[audioguru]

Your transistor works. Its input DC voltage is very close to 2V. Its output DC voltage is very close to 6V. Its current is 0.01A which is 10mA.

Its input signal is 500mV and its output signal is 1539mV so its voltage gain is slightly more than 3.

 

[kchriste]

If I was to reduce Re to something like 15 ohms that would mean that 150mV would be fed into Re (15 ohms x 10mA) which would also mean that the PD would have to bias 850mV?
But even if I was able to do that wouldnt it mean a gain of (Ic = 400 / Ie 15) of 26?

Correct. Sort of.
Remember I mentioned that the transistor has it own internal r'e? The formula for small signal transistor like the 2n2222 with a rectangular junction is:
r'e = 25mV/I where I is in milliamps. So the 2n2222 would have an r'e of 25mV/10ma = 2.5hm: when it passes 10ma of current. So gain would equal 400/(15+2.5) = 22.86. But when the emitter resistor is larger, like 130hm: in your diagram, you can usually ignore the transistor's internal r'e.

hFE? hfe? what is it I have been working out?

hFE which is correct for figuring out the DC biasing. Just don't try to use it to calculate the voltage gain of the CE amplifier.
If you want more AC voltage gain from transistor amplifiers, you need to cascade a few stages. If you cascade two amplifiers, each with a gain of 10, you'll have a total gain of 100.

 

[audioguru]

You don't want a transistor amplifier to have its maximum amount of gain because its distortion at high levels is horrible like this:

 

[UTMonkey]

Thank you all, from what you have seen so far do you think I am "grasping" this?

Perhaps I am actually going through this the wrong way, as all the mails on this forum tend suggest "What is it I am trying to do?"

Actually I don't have an application for this circuit - I just wanted to understand transistors more, but after getting valuable comments from you guys it really has peaked my interest more.

So what is it I want this amplifier to do?

Incoming Signal

If I wanted to build the first stage of an amplifier, what do I need to know about the signal I am amplifying?

What is the source signal? should I be interested in its Vp-p?

There is mention of its impedance, how relevant is that to this amplifier stage?

Outgoing (amplified) signal
I could almost ask the same question for the outgoing signal.....

I would be very grateful if someone could for the sake of my learning give me a signal specification and let me me design the first stage of an amplifier to acommodate it.

Many Thanks

 

[muzubair]

hey,
u can also use variable POTs instead.....

 

[Roff]

hey,
u can also use variable POTs instead.....

Hmmm... Gives new meaning to the term "intelligent design".

 

[audioguru]

Make every resistor a variable pot. Then simply tweak their settings until they all smoke.

 

[kchriste]

If I wanted to build the first stage of an amplifier, what do I need to know about the signal I am amplifying?

It's frequency range, impedance and level (Vp-p).

There is mention of its impedance, how relevant is that to this amplifier stage?

It is important as far as loading and power transfer is concerned. For maximum power transfer, as in RF circuits, the source impedance should match the input impedance of the amp. But for minimum loading, as is common in audio circuits, the amplifier input impedance is typically designed to be 5-10 times the source impedance. Telephone lines typically use 600ohm impedances and 600hm: isolation transformers.

Outgoing (amplified) signal
I could almost ask the same question for the outgoing signal.....

Yes, the amp is now the source and something else is the load. Same rules apply.

 

[UTMonkey]

Thanks,

For arguments sake, what kind of impedence and voltage level would a garden variety microphone have?

And if i refer to my previous amplifier what would that mean to my design of the amplifier.

Could you please humour me with this, I would like to take the theory I have learnt over the past posts and confirm it with a "proper" example.

I am not looking for a mega "amplifier" example just something that proves my workings out.

Regards

 

[audioguru]

An electret microphone is powered from a 10k resistor and together are a 3.3k ohm impedance. The output level is 1mV at 3m, 5mV to 10mV at 10cm and up to 100mV if it is very close to a screaming person.
So its preamp will have an input impedance of 17k to 34k ohms or more and a gain of 100 with a gain control. With the mic at 10cm from somebody talking the output will be 0.5V to 1V and can be reduced by the gain control.
For low distortion a 2-stage amplifier will have a gain of 10 for each stage, or a 2-stage amplifier with negative feedback can be used, or a low noise opamp with negative feedback.

 

[UTMonkey]

Thanks for the info.

How on earth did you calculate the impedance (17k-34k)?
Gain control of 100?

 

[kchriste]

For garden variety, the electret microphone module is fairly common. It has a built in amplifier with a typical output impedance of 600 to 4000 ohms. This is the type of mic you'll see in cheap tape recorders, etc. So lets just say it has an impedance of 1000hm:.
So, you'd design your amplifier to have an input impedance of 5Khm:. To raise the input impedance you would lower the collector current because the input impedance is the divider in parallel with the base impedance. They are calculated as follows:
Divider Z = R3//R4 (// = parallel)
Transistor base Z = Beta(RE + r'e)
Input impedance = DividerZ // TransistorBaseZ
EDIT: I've noticed that AudioGuru gives a different impedance. You'll have to find the data sheet for the mic in question to truly know the answer to this.