BTJ or JFET in the gain stage

[tavares_costa]

I asked the GPT guy, but I don't trust him that much, as for several times it answers dubious or incorrect responses... And I think some topics are somehow biased.

So I wanted to know:
BTJ or JFET?
How accurate are its answer?

GPT:

JFETs in the Gain Stage:

• High input impedance: JFETs typically have a very high input impedance, making them suitable for applications where a high input impedance is desired. This minimizes loading effects on preceding stages and helps maintain signal integrity.
• Low noise: JFETs tend to exhibit lower noise characteristics compared to BJTs, making them advantageous in applications that require low noise performance, such as audio amplifiers.
• Simple biasing: JFETs have simpler biasing requirements compared to BJTs, making them easier to implement in some cases.
• Voltage-driven loads: JFETs have a higher output impedance compared to BJTs, which makes them more compatible with voltage-driven loads such as capacitive loads.

BJTs in the Gain Stage:

• Higher current gain: BJTs typically have higher current gain (β) compared to JFETs. This higher gain can be advantageous in applications where higher amplification levels are required.
• Lower output impedance: BJTs generally have a lower output impedance compared to JFETs, which can help drive low-impedance loads such as speakers more effectively.
• Wide availability: BJTs are widely available and offer a wide range of options, making them suitable for various amplifier designs.
• Cost-effective: BJTs are generally more cost-effective compared to JFETs, which can be a consideration in some designs.

In summary, JFETs are often preferred for their high input impedance, low noise characteristics, and suitability for voltage-driven loads. They are particularly beneficial in applications that require high input impedance and low noise performance. On the other hand, BJTs offer higher current gain, lower output impedance, and cost-effectiveness, making them suitable for applications that require higher amplification levels and driving low-impedance loads.

 

[Mark Tillotson]

Quite a lot of gobblydegook in there really:
"which makes them more compatible with voltage-driven loads such as capacitive loads" - huh?
"BJTs typically have higher current gain (β) compared to JFETs" - complete lack of understanding.
"JFETs have simpler biasing requirements compared to BJTs" - doesn't understand the high variability of FETs.

I don't see evidence of these pedestrian lecture notes having passed through a thinking muscle, just rehashing of phrases is enough to explain this output. And what is this mythical "gain stage"? Where is the mention of the exponential response of BJTs v. the quadratic of FETs? Where are the diagrams? EMI susceptibility was missed, although that's arguably a minor issue. Note the telling grammatical error in : "applications that require high input impedance and low noise performance". I want high noise performance!

 

[tavares_costa]

Just what I wanted to know, because I always had a feeling that most of the times its answers were wrong.

 

[rayma]

If it doesn't actually "know" then it just makes up stuff.
And there's no way to tell the difference except by independent checking.

 

[bucks bunny]

"If it doesn't actually "know" then it just makes up stuff."

Same way like real humans do. 😊

 

[tavares_costa]

"If it doesn't actually "know" then it just makes up stuff."

"Same way like real humans do. 😊"

I suppose he was well trained then...

LOL

 

[rayma]

The better humans say they aren't sure, or don't know.
Granted, there aren't that many of them any more.

 

[Mark Johnson]

BJTs in the Gain Stage:

• Lower output impedance: BJTs generally have a lower output impedance compared to JFETs, which can help drive low-impedance loads such as speakers more effectively.

This is simply wrong. What you want in a gain stage is ... drum roll ... lots and lots of gain. To get high gain (thus low distortion) you want the gain stage transistors to operate in a high output impedance regime. This can be accomplished in both JFET-based and BJT-based gain stage circuits, through the well known circuit topologies called (i) degeneration; or (ii) cascodes; or both.

Gain stages work best when driving high impedance loads. That's why many audio power amps arrange for their gain stage to drive a Triple Emitter Follower output stage circuit. The TripleEF presents a very high impedance for the gain stage to drive.

In the passage above, "... BJTs ... which can help drive low-impedance loads ..." is irrelevant and also ignorant. In the context of Gain Stages, the load is guaranteed not to be a low impedance.

 

[Citizen124032]

To be frank, as an admirer of jfets, I voted that single bjt cast, for the simple reason that the transfer steepness, and thus the 'amplification ability', is higher with bjt then with jfet's. I guess that counts for mos also. And I considered the voting about voltage amplification, not current amplification.
So, for high (voltage) amplification in a certain undefined stage, bjt's have some better options, but this is a smash flatten observation, not a verdict. All depends on countless other condtions, also.

 

[Audio1Man]

Read Marks #8 and then read it again.

If I were to carry 10 tons of rock up the mountain, I would not use a VW Bug.

I VOTE FOR THE CIRCUIT DEVICE(s) THAT ARE THE BEST FOR THE PRODUCT DESIGN.

Every part/device has its unique function and limitations. What I choose is my best solution to fit all of the specs and $ costs.

Duke

 

[WhiteDragon]

Some of the statements make sense.

And the majority of hobbyist will stick to BJT for gain stages.

Jfet is mainly for low noise, ironically in a high gain situation.
Such as Phono cartridge amplification or microphone preamp.

 

[bimo]

Only few JFET that have voltage input noise around 1nV/√Hz. Below this, it hard to find. Input noise current very low, so it is better for high impedance input.
BJT can have input voltage noise below 1nV/√Hz, but input current noise more higher. BJT for low input voltage noise mush have very low rBB. To achieve low input current noise, BJT need to have high hFE and operate at low collector current, but voltage input noise will be higher at low collector current.

So, it depend on the application.

 

[jagwap]

JFETs are sometimes less susceptible to RF as they do not have a PN junction on their input that could demodulate unwanted signals.

JFETs' very high input impedance can be useful. Having near zero input bias current has it's uses. Guitar amp front ends for example.

BJTs are usually cheaper. But it depends on part choice.

BJTs can be lower noise on low impedance sources, and JFETs can be lower noise on high impedance sources. But it is all about part choice and circuit context.

After that it is really how you use them.

 

[Mark Johnson]

I think you probably meant that JFETs don't have a forward biased PN junction on their input. JFETs certainly do have a pair of reverse biased PN junctions on their input: one from gate to drain, the other from gate to source. That's what the "J" in JFET signifies

Junction
Field
Effect
Transistor

If you insist upon using semiconductor devices with absolutely no PN junctions on their input, you can choose IGFETs (Insulated Gate Field Effect Transistors). By far the most popular IGFETs are the ones which employ silicon dioxide as the insulator and monocrystalline silicon as the body. These go by the name MOSFET: Metal Oxide Semiconductor Field Effect Transistor; and big electronics distributors sell thousands of different models.

 

[tavares_costa]

here's BARD answer for this topic:

BJT

• Pros:
  • Low cost
  • Wide availability
  • High gain
  • Good linearity
  • Can be used in a variety of applications
• Cons:
  • Lower input impedance than other types of transistors
  • More susceptible to noise than other types of transistors
  • Can be more difficult to bias than other types of transistors

JFET

• Pros:
  • High input impedance
  • Low noise
  • Can be used in a variety of applications
• Cons:
  • Lower gain than BJTs
  • Less linearity than BJTs
  • More susceptible to voltage changes than BJTs

MOSFET

• Pros:
  • Very high gain
  • Low noise
  • More tolerant of voltage changes than BJTs
  • Can be used in a variety of applications
• Cons:
  • Lower input impedance than JFETs
  • More susceptible to noise than JFETs
  • More expensive than JFETs

IGBT

• Pros:
  • Very high gain
  • Very low input impedance
  • More tolerant of voltage changes than BJTs
  • Can be used in high-power applications
• Cons:
  • More expensive than other types of transistors
  • More difficult to drive than other types of transistors

Ultimately, the best type of transistor to use for the gain stage depends on the specific application. If you need a low-cost, high-gain transistor that is widely available, then a BJT is a good choice. If you need a high-input impedance, low-noise transistor, then a JFET is a good choice. If you need a very high-gain transistor that is tolerant of voltage changes, then a MOSFET or IGBT is a good choice.

Are Bard's answers more accurate than GPT?

 

[tavares_costa]

I don't think Bard knows what he's saying as well.. I'm mot an expert on this subject as most of you are, but I don't think 1943 is a good candidate for the gain stage:

Bard:
The 2SA1943 is a popular choice for the gain stage of an audio amplifier because it has a high gain and low noise. It is also a relatively inexpensive transistor, which makes it a good option for budget-minded audiophiles.

The 2SA1943 has a gain of about 400, which is higher than many other transistors. This means that the 2SA1943 can amplify a signal with a small input voltage to a much larger output voltage. This can be useful for audio amplifiers, as it allows the amplifier to produce a louder signal.

The 2SA1943 also has a low noise figure, which means that it does not introduce much noise into the signal. This is important for audiophiles, as they want to hear the original signal as accurately as possible.

Overall, the 2SA1943 is a good choice for the gain stage of an audio amplifier. It has a high gain, low noise, and is relatively inexpensive.

Here are some other transistors that are similar to the 2SA1943:

• 2SC5200: This is a PNP transistor that has similar specifications to the 2SA1943.
• MJE15032/33: These are NPN transistors that have slightly higher gain than the 2SA1943.
• 2N3904: This is a NPN transistor that has a lower gain than the 2SA1943, but it is also less noisy.

The best transistor for your application will depend on the specific requirements of your amplifier. If you are looking for a transistor with high gain and low noise, the 2SA1943 is a good option. However, if you are looking for a transistor with lower noise, you may want to consider one of the other transistors listed above.

 

[Mark Johnson]

BARD is confused about what a "gain stage" actually is. BARD is also confused when she says "2SA1943 can amplify a signal with a small input voltage to a much larger output voltage". Because BARD is completely ignorant of the difference between current gain and voltage gain. BARD is completely wrong when she says "The 2SA1943 has a gain of about 400". Here's a (link to the manufacturer's datasheet) and the relevant portion, in my opinion, appears in the image below. See whether you can find the number 400 anywhere.

_

 

[tavares_costa]

Thanks for the reply Mark.

I spent some time talking to Bard and GPT, they both give wrong answers about everything, and I've been reading Datasheets as well and most of their answers are somehow biased from where they get the info, mixed with something like; I don't have a really good answer, so ill just throw some values to keep this guy happy.

I know AI is still in an early stage, but to release something like this, that gives you anecdotic answers to say the least, what is the purpose? I suppose they are good for making a "summary" from a text, or help you write an excuse for your boss saying why you missed job for two days, but apart from that, I don't see any advancements in a near future.

I tried before GPT in coding languages like Java, C++ and Python, 90% of the time the code did not come to any good thing for complex things,

 

[VivaVee]

My experience with ChatGPT is that it happily repeats/reports outright lies/fabrications (and this is only for science/engineering questions).
I assume that this is because it has no sense of authority (as in what is a trustworthy source) and has no ethical constraints
There are other AIs that provide their sources so that you can check the authority

 

[AKSA]

If we use GPT and Bard to run our world it may be even more bizzarre than it is now!
I have never found a 2SA1943 with a gain (hfe) of 400. Hmm.......

HD