Hello Everyone
I am a junior year EE student working on a project for personal fun. I'm designing a circuit for a guitar effects pedal using PNP transistors and have a question about biasing them. The circuit I'm planning requires a specific ratio of Beta and I am wondering if it is possible to set the value of Beta somehow by biasing the transistors. Since Beta is ic/ib is it possible to set those values by choosing Rc and Rb to obtain the values of Beta that I want (within the range given on the data sheet). Or would I have to build a small circuit to measure ic and ib experimentally to find a transistor that has a value close to what I want?
PS my cheap multi meter can not measure HFE.

Thank you!

 

Beta is an inherent property of the transistor, like the resistance of a resistor. However beta does change depending on the collector current. So if you measure ic/ib with a collector current of 100mA you will get a different answer if you do that measurement at 100uA.
The table below is from the 2N3906 datasheet and you can see that the minimum beta varies from 30 to 100 at different collector currents. It will also vary depending on temperature and, as you clearly know, from one transistor to another.


Given the temperature dependance, even if you select a transistor with a specific gain, the gain will still vary with temperature. What would be better is to design your circuit so you get the gain you need with as little dependence on the specific transistor as possible.

For instance read:https://en.wikipedia.org/wiki/Common_emitter#Emitter_degeneration

 

Beta is an inherent property of the transistor, like the resistance of a resistor. However beta does change depending on the collector current. So if you measure ic/ib with a collector current of 100mA you will get a different answer if you do that measurement at 100uA.
The table below is from the 2N3906 datasheet and you can see that the minimum beta varies from 30 to 100 at different collector currents. It will also vary depending on temperature and, as you clearly know, from one transistor to another.
View attachment 137311

Given the temperature dependance, even if you select a transistor with a specific gain, the gain will still vary with temperature. What would be better is to design your circuit so you get the gain you need with as little dependence on the specific transistor as possible.

For instance read:https://en.wikipedia.org/wiki/Common_emitter#Emitter_degeneration

Thanks for the reply. Is it acceptable practice to measure values of ic and ib experimentally when designing circuits? I realize this would not work on a large scale because of the potential cost of doing so. I'm sure the best way to design is to reduce dependence on Beta like you mentioned. Or is it more common practice to design a circuit just using those minimum values and setting VCE and ic?

 

It is not general practice to measure beta. Beta can typically be between 100 and 300 for 2N3906.
It is best to design the circuit not for beta, but for voltage gain that is much lower than beta (say 10 for example). In this way you are allowing for headroom and your circuit still amplifies regardless of beta.

 

It all depends.
For instance, an audio amplifier, the gain is not critical as you have a volume control.

 

And one other thing about Beta -it drifts with temperature!

 

Is it acceptable practice to measure values of ic and ib experimentally when designing circuits? I realize this would not work on a large scale because of the potential cost of doing so. I'm sure the best way to design is to reduce dependence on Beta like you mentioned. Or is it more common practice to design a circuit just using those minimum values and setting VCE and ic?

You answered your own question; in a roundabout way...

For boutique designs it's acceptable because your customers are willing to pay for the additional labor, and part rejection cost, and wouldn't care that they couldn't repair such circuits because they wouldn't know what beta was required.

But the more common practice is to design circuits that only depend on worst case minimum beta.

When I worked at HP Labs, I learned that some products cherry picked standard parts; meaning they did 100% testing on parts to select the ones that met the circuit requirements. I became aware of the practice when I went to school stock to get some parts to send to my school and was given tubes of integrated circuits that were fully functional, but failed the cherry pick criteria.

To their credit, they assigned HP part numbers to the parts and mentioned in some part cross reference guides that when selected parts were used in a design. Standard parts could be used in a pinch, but should be replaced with the HP selected part to restore the equipment to it's original specs.

I am a junior year EE student

You highlight one of my gripes about 4 year EE programs. The first two years is all general subjects and you don't find out until 3rd year whether you really have an aptitude for your chosen major.

You should pay more attention to your English classes and use paragraphs to organize your thoughts. In my first college calculus class, the professor routinely took points for poor grammar. His pet peeve was using 'the' when 'a' was more appropriate, and vice versa.

 

For boutique designs it's acceptable because your customers are willing to pay for the additional labor, and part rejection cost, and wouldn't care that they couldn't repair such circuits because they wouldn't know what beta was required.

But the more common practice is to design circuits that only depend on worst case minimum beta.

You highlight one of my gripes about 4 year EE programs. The first two years is all general subjects and you don't find out until 3rd year whether you really have an aptitude for your chosen major.

Thanks. That's what I figured.

I agree there are many flaws in the US education structure. I go to a decent tech university and the way the curriculum is organized with all the prerequisite courses it seems like I am just now getting into courses that feel like I'm doing actual design stuff.

Fortunately for me, I found something that interest me to the point that I want to do projects on my own.