Hi guys.. 🙂
Looking around in this forum there are tons of amp designs with all kinds of goals they try to achieve. Some are BJT based, others use MOSFETs and yet others use a happy mix of the two.
I'm interested to sort of find out to see if there is a shift in preference. It doesn't matter whether you are a builder or a listener. MOSFET amps have been around a while, but as always with 'new' technology, they weren't perceived to outgun BTJs when it came to listening pleasure. Over the years things have changed ofcourse. Both BJTs and FETs have been developed significantly over the past 30 years and it'll be interesting to see the outcome of this poll.
So go ahead and shoot!
Looking around in this forum there are tons of amp designs with all kinds of goals they try to achieve. Some are BJT based, others use MOSFETs and yet others use a happy mix of the two.
I'm interested to sort of find out to see if there is a shift in preference. It doesn't matter whether you are a builder or a listener. MOSFET amps have been around a while, but as always with 'new' technology, they weren't perceived to outgun BTJs when it came to listening pleasure. Over the years things have changed ofcourse. Both BJTs and FETs have been developed significantly over the past 30 years and it'll be interesting to see the outcome of this poll.
So go ahead and shoot!
MOSFETs are better overall IMHO especially the ones designed for audio purely beause they are easier to implement and can dissapate more power wiout secondary break down and a positive temp coeffcient. To say you prefer FETs over BJTs is silly the chances of people listening to an amplifier that has only output devices changed and the rest of the amplifier capable of driving both without affecting input stage and such like is highly unlikely.
I would bet in a proper blind AB listening test no one would be able to tell between two amplifiers, both of which capable of driving the load competently and I mean two SS amps not SS vs t00bz.
I would bet in a proper blind AB listening test no one would be able to tell between two amplifiers, both of which capable of driving the load competently and I mean two SS amps not SS vs t00bz.
Thanks for your comments 🙂 I'll stay away from discussing an indivual's opinion on preference whatever the underlying reason is; the poll is merely to get an overall impression on device preference.
This is valid for lateral only(actually negative tempco)
I have the amps of both types and I like both equally.
This is a good question.
I like whatever parts and circuits that sound good.
Each has its plusses and minuses. Whatever works.
I like whatever parts and circuits that sound good.
Each has its plusses and minuses. Whatever works.
You might as well ask: "Which are better - apples or oranges?".
Since good amplifiers can be designed using either, but will need to be different designs that optimise the use of the chosen output device, there is no basis for comparison. The pros and cons of each have been written about extensively by many authors - and that's the point, engineering comes down to a chosen set of compromises according to the desired outcome and which devices are chosen is determined thus or simply by preference.
Since good amplifiers can be designed using either, but will need to be different designs that optimise the use of the chosen output device, there is no basis for comparison. The pros and cons of each have been written about extensively by many authors - and that's the point, engineering comes down to a chosen set of compromises according to the desired outcome and which devices are chosen is determined thus or simply by preference.
true, but since we are practicing engineering here, there are three sides we take care of: performance, cost and safety. with some transistors it is cheaper to achieve the same performance and safety.
Member
Joined 2009
Paid Member
for Class A maybe MOSFET, but all else I would pick BJT without a doubt. Except that I like tubes 🙂
Hi,
hmmm, and I always thought that a amp would be a little more than just a output stage
jauu
Calvin
hmmm, and I always thought that a amp would be a little more than just a output stage
jauu
Calvin
I personally am totally crazy about MOSFETs due to their ease of driving and their natural hi-z input impedance. There are some other specifics that make me chose FETs. MF80 uses FETs for output devices and VAS cascodes, but the rest are BJTs because of the topology (fully symmetric). My new project will be completely FET based; due to brand the new topology it's impossible to do the same thing with BJTs, so it's not like I've been going out of my way to avoid BJTs 🙂
Hi MagicBox,
I tend to prefer vertical MOSFET outputs with JFET front ends and BJTs in the middle. However, the emergence of the ThermalTrak BJTs has, in my opinion, revitalized the BJT output approach, and I really like them. Even before the ThermalTraks I always believed that BJTs made great amplifiers once they became fast enough with ring emitter and the like technology.
It was just a personal leaning toward MOSFETs ever since I did my MOSFET amplifier with error correction in the early '80s. I really like the very high speed of the vertical MOSFETs and the freedom from beta droop and ft droop at high current. At the same time, they can be more prone to parasitic oscillations if one is not careful. The equivalent ft of vertical MOSFETs can be in the hundreds of MHz.
The key thing to recognize is that every technology depends on the right care and feeding by the designer. Good designers can make great amplifiers out of either MOSFETs or BJTs, and there are many examples out there. Conversely, unfortunately, poor designers can make terrible amplifiers equally well with either MOSFETs or BJTs.
I think it is also important to realize that among MOSFETs there are two sub-catergories: Laterals and Verticals. They each have their followers and corresponding advantages and disadvantages. Pick your poison. Laterals are robust and difficult to destroy, partly because of their higher Rds-on and their negative coefficient of current vs temperature. Verticals are a bit easier to destroy, but are quite robust when properly used and protected. Laterals got a head-start great reputation with their use in the legendary Hafler amplifiers. Verticals are faster, have higher transconductance and can deliver higher current.
Although I often lean toward vertical MOSFETs, I try to be fair in my book to all 3 - BJT, Laterals and Verticals. In fact, I devote a lot more space to the BJTs because they really are the work-horse of audio power amplifiers.
Cheers,
Bob
Hello
For a class A I would use mosfet, and for class AB I would use BJT.
But here arround mosfet are difficult to find and buying by mails do often get lost wen arrive here arround, so I mostly do only class AB BJT amps.
Bye
Gaetan
For a class A I would use mosfet, and for class AB I would use BJT.
But here arround mosfet are difficult to find and buying by mails do often get lost wen arrive here arround, so I mostly do only class AB BJT amps.
Bye
Gaetan
Ever since I started building output stages with vertical mosfets I have not looked back, provided EC is employed. The biggest disadvantage of the HEC vertical mosfet follower is the extra voltage required for proper drive. If care is given to proper heat removal from the vertical mosfets they can take lots of abuse. Because Zin of the EC driver circuit is so high, a low current (few mA) VAS can be used to drive and bias the output stage. This means that the higher tier rail voltage for the VAS can be provided with a simple voltage doubler followed by a low current voltage regulator. I have found that using a bootstrap to power the driver stages within the EC circuit works great and eliminates the need for the higher tier rail to supply the current for the drivers. This leaves only the need to regulate a few mA so a small signal regulator circuit is all that is needed.
Small signal J-fets are great for driving small signal BJT's. This give you the advantage of the J-fets and the Gm of the BJT. Since the VAS is powered by a higher tier voltage source it is of little consequence to use a cascode VAS. Cascode VAS is the way to go IMO.😉
I have not played much with the Thermal track BJT's but they might be something to re-evaluate using BJT output stages. I can see an advantage with them for large production quantity circuits. It would have a lot less components than a properly implemented HEC vertical mosfet follower. However, even when adding up the cost of all the components in the HEC follower circuit it is still less than the Thermal track BJT's.🙂
Here is an example of the circuit I'm referring too... (B+/B- = +24V/-24V)
Small signal J-fets are great for driving small signal BJT's. This give you the advantage of the J-fets and the Gm of the BJT. Since the VAS is powered by a higher tier voltage source it is of little consequence to use a cascode VAS. Cascode VAS is the way to go IMO.😉
I have not played much with the Thermal track BJT's but they might be something to re-evaluate using BJT output stages. I can see an advantage with them for large production quantity circuits. It would have a lot less components than a properly implemented HEC vertical mosfet follower. However, even when adding up the cost of all the components in the HEC follower circuit it is still less than the Thermal track BJT's.🙂
Here is an example of the circuit I'm referring too... (B+/B- = +24V/-24V)
Last edited:
I'd like to agree completely on this; but probably for slightly different reasons, extended reasons rather:
A vas stage/device does a number of tasks:
-Provides your 2nd tier high Gm
-Provides voltage conversion/output
-Provide a current controlled input referenced against a power rail
A cascode relieves the VAS device from at least one task; the voltage conversion. This is now done by the cascode device. The cascoded device is now tasked by providing gain and a powerrail referenced input only.
Splitting out these individual tasks over more devices is imo what contributes to significant improvements.
In my middle stage concept I'm splitting things even further. The cascoded device is no longer responsible for gain either. All it does is to provide voltage domain conversion through mirrorring. What I mean by that is that unlike powerrail referenced VAS device, my actual gain devices no longer need to be referenced against supply rails either, they can be referenced at any preferential voltage, allowing for a much wider control range and much more flexible adaption on how the gain devices are controlled. It allows these gain devices to work in class A fully and since their drain voltages never change, it's like the gain devices are inherently cascoded. In addition, each side's (north/south) gain element consists of an N/P combination so the gain behavior is 100% identical for both the north and south swing, especially when the Ns are matched and the Ps are matched. No N vs P matching required.
But I think I now explained a little too much 😉
Edit:
Hello Bob! I'm honoured to have you visit here with your thoughts, much appreciated =)
Last edited:
Hi CBS240,
These are all good points. In fact, even BJT amplifiers can sometimes benefit from boosted rails for the input and driver circuits because of the extra headroom they provide for the VAS. With respect to bootstrapping, Halcro also uses that for powering the EC circuit.
I would also point out that, although vertical MOSFETs have a higher turn-on voltage than Laterals (about 3.5V vs about 1V), for higher currents the Laterals need just as much forward gate bias because of their lower gm and higher Rds.
Although EC really makes vertical MOSFETs shine, great amplifiers can be made with them without EC as well. Even a single pair of verticals biased at 150mA will do below 0.03% at 20kHz in a straightforward design. Also, using TMC compensation instead of EC really gets low distortion as well.
One of the nice things about MOSFETs is that they are relatively uncritical about the bias current setting. For them, the more current the merrier as long as dissipation is not too much, unlike BJTs where above-optimum bias can actually increase crossover distortion due to gm doubling. In fact, the reason I applied EC to MOSFETs in the early '80s was because of what I termed "transconductance droop" (sort of the opposite of the term "gm doubling" which was later used to describe the effect of over-biasing a BJT output stage).
The bottom line here is that a bigger class A region is usually available with mosfets because they are happy to be biased at higher currents.
Cheers,
Bob
- Status
- Not open for further replies.