• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Mosfet??

tromba

Super Member
So what is MOSFET? I looked up and for this:
Although the MOSFET is a four-terminal device with source (S), gate (G), drain (D), and body (B) terminals,[1] the body (or substrate) of the MOSFET often is connected to the source terminal, making it a three-terminal device like other field-effect transistors.

But that's not deal helpful to me. So bottom line, why would I want a MOSFET amp, or why wouldn't I want one? Are they better for some application than others? It's it just the "smoother" sound that's always mentioned? Is MOSFET TOTL?
 
Register to hide this ad
This is an age-old question: Bipolars or MOSFETs. The short answer to the question "Are MOSFETs TOTL" is.....it depends.

MOSFETs are usually used in multi-channel applications, such as HTA, because they produce less heat. In two channel applications, bipolars are usually preferred.

Now I'll sit back and let the fun begin...:lurk:
 
Go Bipolars !

Many amplifiers have always used bipolar transistors in their output stages for several significant reasons. When compared to other types of amplification transistors, bipolars offer:
Wider bandwidth
Lower distortion
Excellent reliability and a longer life span
High Efficiency
High resistance to ESD (electrostatic discharge)

:banana:
 
Many amplifiers have always used bipolar transistors in their output stages for several significant reasons. When compared to other types of amplification transistors, bipolars offer:
Wider bandwidth
Lower distortion
Excellent reliability and a longer life span
High Efficiency
High resistance to ESD (electrostatic discharge)

:banana:

I'll agree with you about bipolar devices' reliability, efficiency and ESD resistance, but I beg to differ on the other two points. As long as the driver stage has enough current to charge and discharge the gate capacitance, a MOSFET circuit can be at least as wideband (or fast, if you prefer) as a bipolar circuit of comparable power rating. The I-V characteristics of MOSFET devices are also more linear than the I-V curves of bipolars, so amplifiers built using those devices will exhibit lower distortion before negative feedback is applied.

I've owned push-pull tube amps, bipolar amps and one MOSFET power amp. The one that sounds best to me and has earned a long-term home in my main system is an Audio by Van Alstine MOSFET amp. Second best, interestingly, is a Heathkit AA-1640 with a quasi-complementary circuit (all-NPN outputs, complementary driver), and nipping at the AA-1640's heels is a late-1950s Sherwood S-5000 integrated amp with an EL-84 push-pull tube output stage.
 
IGBTs uber alles! Down with the nanny state of TOPFET! MOSFET and IGFET are conspirators! HEXFET is unnatural!
 
Yeah, this sort of thing can be really confusing if you don't have a solid electrical engineering degree. Here's what I believe is a good discussion of the good and bad of MOSFETs written by Nelson Pass some 30 years ago: http://www.passdiy.com/pdf/citation.pdf.

I'd suggest you read at least the first 3-4 pages on the device tradeoffs. After that, he delves into a well-known modification for the old HK Citation 12 power amps. BTW, I've got one of those as a project here at Casa dShoaf. I also have both bipolar (HK Citation 16) and a MOSFET (Hafler P500/DH-500) amp in the line up here along with tube power amps (HK Citation II and Mac MC-225). Bottom line: They're all excellent in their own right.

Hope that takes some mystery out of MOSFETs.

Cheers,

David
 
Last edited:
A sound company I used to work for in the '80's used to use Metron A4000 amplifiers for their sound systems....
These are 350watt @8Ω or 600 @4Ω.......
The amplifiers run flat out all the time as you would expect, but the Metrons kept blowing up, they ran hot, and didnt like the low impedance loads.....Very unreliable at full power.....
So we modified (which meant simplified) the driver stage, and fitted MOSFET's to the output section......
These amps stayed in service for another 10 years, and I can tell you from 32 of these amplifiers, we then had VERY few problems.......
They were then VERY reliable, and could drive 2Ω Nominal loads with ease....
I have one of these amplifiers I use on my studio monitors currently....

Go MOSFET's!!

img0584l.jpg
 
FETs are the way to go in high power situations. One of their intrinsic properties is to heat spread. In a nutshell that means as one part of the transistor die gets hot it loses its gain. The load current then moves away to a cooler spot until that spot gets hot. The process occurs continually helping to distribute heat across the transistor die.

BJTs exhibit the opposite effect. When the spot on the die gets hot then that spot increases its gain. The increased gain causes more current to flow through that spot. If the heat sink and/or feedback control isn't strong enough to keep the BJT in check then it will eventually destroy itself by thermal runaway.

Here is a bit on Wiki.

http://en.wikipedia.org/wiki/Thermal_runaway#Power_MOSFETs
 
I know I'd love the MOSFET sound and a MOSFET amp, like a Sansui AU-X1111 MOS Vintage. Just one problem which holds me totally back from making a MOSFET amp purchase: It's hard, to impossible, to get some MOSFET repair parts, for many amps.

Gotta check into that before you purchase the amp that you desire to purchase. Are there repair MOSFET parts?

Expensive amp, no MOSFET parts availiable. Now you have an amp with dead MOSFETs.

If exact replacement MOSFETs are availiable for the amp you want...buy some extra MOSFETs and get the amp! If the amp has 6, I'd buy 8.

Something to consider.
 
Last edited:
I didn't even know what the question WAS!!! So I gather the opposite of MOSFET is bipolar. Now on to the reading suggested by others in the thread.

Thanks for all the replies so far.

Not really the opposite, but 2 different kinds of transistors. Constructed differently. JFETs and VFETs are others. All have somewhat different characteristics, and IME sound a bit different to one another.
 
MOSFETs do not run cooler.

The dissipation of a device is determined by the voltage across it multiplied by the current through it. For a power amp with a given rail (power supply) voltage and a bias current+ signal current, the dissipation is the same, unless your talking about Gate current vs. Base current. You can calculate that, but it's not enough to say that MOSFETs run cooler.

Yes, MOSFETs can be more reliable in a basic design, but it's also kind of a negative as well. If an amplifier is delivering big current to a low impedance load, the output device(s) heat up. In a bipolar transistor, the current gain goes up (ohmic value goes down). In a MOSFET, the ohmic value goes up. Yes, that increases reliability, but it also reduces output. Kind of a built-in VI limiting circuit.

Pluses and minuses.
 
Last edited:
MOSFETs do not run cooler. . .

I never said that they run cooler. I said they inherently spread heat across the die. The process is exactly as you described except I described it for infinitesimal die areas and you described it as a lumped parameter.

The heat spreading property of FETs is also what enables FET dies to be simply paralleled. That is impossible with BJTs because one of the devices will hog the current and overheat while the other BJTs sit there like bumps on a log.
 
I've tried a number of SS amps over the years and I recently realized that my favorite in the low power side of things is the Creek 4330 while my favorite in the higher power side is the Conrad Johnson MF 2300A. The one thing they have in common is output MOSFETS so I guess there must be something to that sound "character" that I find appealing. I guess it must be the lack of a "fatigue" factor.
 
Back
Top Bottom