# FT in transistor

1. Apr 29, 2012

### baby_1

hello my friends
i searched a lot in the internet and i couldn't find my answer.could you tell me what is exactly FT do in transistor?
what are the formula that join ft to another transistor elements for designing an oscillator or mixer?
if good book tell me more i will appreciate you.

i read the wikipeida before.but i asked many people about ft to choosing a transistor for desire job.for example if i want to create a collpits oscillator with 100Mhz frequency what transistor should i choose?what is the formula according the ft of transistor and my frequency?
could you help me ? its not very important to tell me in English.please tell me with your fresh language
its important to me to understand and know about electronics in RF environment

as you see in many oscillator with frequency between 90-108 Mhz they use c945 with 150 ft,how it works fine?
does we have an equation that explain the formula between Voltage Gain and ft in colpitts oscillators?
because i asked somebody else but tell me that "current gain * BW product = fT "
but if i use emitter-common that Current gain is approximately 100 and for oscillate my circuit in 100Mhz i shoud use transistor with ft 10Ghz?!!!

Does any books tell me moe about designing collpits oscillator with all steps(such as choosing transistor , Capacitor , ....)? theory and experimental?

2. Apr 29, 2012

### AlephZero

Ft is the frequency when the transistor small signal gain $\beta$ falls to 1. If you are operating at a frequency below Ft, the gain is > 1 so you can make a working oscillator.

Ft is not a "constant", it depends on the transistor's operating point. This explains how Ft depends on other transistor parameters, on page 27 onwards: http://www.allenhollister.com/allen/files/physics.pdf [Broken]

Last edited by a moderator: May 5, 2017
3. Apr 29, 2012

### yungman

FT depends on operating condition. The data sheet always give the condition on current and all.
For me, I would use a higher frequency transistor to do a 100MHz oscillator. I don't want the parameters of the transistor to get in the way. Low frequency transistor that has fT of 150MHz has higher parasitic capacitance and other parameters, those a device dependent and you can't count on it. Get a higher frequency transistors, something like 1GHz and the circuit will behave more like the design.....if you know how to do the layout for RF.

Again, this is my way of doing things, I am not claiming it is the best way. I just want to keep things simple by eliminate as much variable as possible and I had been doing this.

Last edited: Apr 29, 2012
4. Apr 30, 2012

### DragonPetter

Is Ft of a transistor like the gain-bandwidth product in an opamp?

5. Apr 30, 2012

### yungman

No, FT is when β falls to 1, this means there is no current gain. But if you have strong base drive, you can still get gain in a CE stage.

6. Apr 30, 2012

### DragonPetter

Last edited: Apr 30, 2012
7. Apr 30, 2012

### yungman

I don't see the article relates the two, the article is for GBWP of op-amp. They just talk about FT in that article in turn of rolling off.
Also the last part:

In a bipolar junction transistor, frequency response declines owing to the internal capacitance of the junctions. The transition frequency varies with collector current, reaching a maximum for some value and declining for greater or lesser collector current.

You have to look at the model to really work with the transistors. For op-amp, not much can change the GBWP.

8. Apr 30, 2012

### DragonPetter

Yes, the opamp GBWP is determined by its internally compensated capacitance pole(s) that are used for stability if you want to talk about its frequency model, just as you talk about the internal capacitances in the model of the transistor. I'm not going to argue about models or anything like this, I just meant from a purely conceptual meaning, the Ft of a transistor and GBWP of an opamp imply the same thing. The article does relate the two when it specifically calls Ft a gain bandwidth product (even if it didn't, why would the article on GBWP have this Ft section to begin with?).

9. Apr 30, 2012

### yungman

I don't think you should call FT GBWP. I never read anything like this. For op-amp, the definition is very simple, the open loop gain falls to 1, that's that. For transistor, it is a whole different story. You really have to look at the input capacitance, input base resistor, Ccb and the whole works.

Someone that has more knowledge should jump in too as I don't claim I am the expert.

10. Apr 30, 2012

### DragonPetter

An opamp circuit's frequency response is also dependent on input and load impedances, but then you aren't really considering just the opamp by itself, which is what GBWP is supposed to tell you. A transistor's datasheet does not know what the base resistor, input capacitance, etc. is so how can the datasheet give a figure for $f_{t}$ if it is design dependent? Obviously, a transistor is biased under a certain scheme, whereas the opamp does not require current biasing, but this is more of a consequence of them being different technologies rather than anything related directly to GBWP. I think GBWP is universal to linear amplifiers, regardless of what their specific technology implementation is.

As another reference, my textbook with chapters on opamps by Sedra and Smith refers to the GBWP for opamps as $f_{t}$. I don't think there is any denying the connection that I asked about initially.

Last edited: Apr 30, 2012
11. Apr 30, 2012

### yungman

GBWP for opamp is voltage gain. Transistor FT is unity current gain. You can still get voltage gain beyond FT. Read post #5.

12. Apr 30, 2012

### DragonPetter

Now you're arguing semantics and mixing ideas. Gain is gain, frequency is frequency. An amplifier can be a voltage or a current amplifier, with GBWP being valid to either. The relationship between gain and bandwidth is valid for both a transistor amplifying current or for an opamp amplifying voltage.

One last reference:
http://www.williamson-labs.com/480_xtor.htm#miller

I see gain bandwidth product referred to transistor amplifiers here, with no mention of opamps.