I finally broke down and decided to learn about mosfets because of their ease of use in class A designs. I don't really know anything about them so could anybody fill me in here?
I need to know how to bias them and that sort of thing. I made a rendition of the Zen amplifier with an IRFZ34 N-ch Enhancement mosfet and it sounds excellent. I want to make some of my own designs so I need to know how to set these puppies up..
I need to know how to bias them and that sort of thing. I made a rendition of the Zen amplifier with an IRFZ34 N-ch Enhancement mosfet and it sounds excellent. I want to make some of my own designs so I need to know how to set these puppies up..
Hi,
Just search around on IRF's web site. They have some really good info.
...Try This doc (It may be a little on the technical side)
http://www.irf.com/technical-info/appnotes/mosfet.pdf
...or other docs here.
http://www.irf.com/technical-info/papers.html
Basically, a mosfet is a voltage controlled current device (or transconductance device). That means that a voltage from the gate to source will cause a current to flow from the drain to source. Very similar to BJT transistors except that a voltage is needed at the Gate, where in a BJT current is needed at the Base to cause current to flow. The source is basically the same as the emitter, wher the Collector is the same the Drain.
The Z34 is a switching mosfet. You may try to use a mosfet designed for linear applications. I don't know how this will affect sound, but the switching fet may be unstable at times.
-Dan
Just search around on IRF's web site. They have some really good info.
...Try This doc (It may be a little on the technical side)
http://www.irf.com/technical-info/appnotes/mosfet.pdf
...or other docs here.
http://www.irf.com/technical-info/papers.html
Basically, a mosfet is a voltage controlled current device (or transconductance device). That means that a voltage from the gate to source will cause a current to flow from the drain to source. Very similar to BJT transistors except that a voltage is needed at the Gate, where in a BJT current is needed at the Base to cause current to flow. The source is basically the same as the emitter, wher the Collector is the same the Drain.
The Z34 is a switching mosfet. You may try to use a mosfet designed for linear applications. I don't know how this will affect sound, but the switching fet may be unstable at times.
-Dan
Thanks for the info. I see what you mean about how the mosfet works. Actually, this z34 sounds absolutely excellent in the configuration I'm using right now. It's very clean and expresses everything describing the qualities of a class A amlifier. I guess perhaps a linear mosfet may sound even nicer. I'll have to get some...
I think switching mosfets can sound really excellent. You can get really big ones too. N-channel for the most part though, not so many really big P-channels. My current amp uses a single switching fet 500v 32A TO-247max per channel and it sounds clean as a whistle. I cannot hear any sonic deficiencies. Nor do I want to!
N-channel for the most part though, not so many really big P-channels. My current amp uses a single switching fet 500v 32A TO-247max per channel and it sounds clean as a whistle. I cannot hear any sonic deficiencies. Nor do I want to!
I think that some of the newer switching MOSFETs should work just fine in a feedback amplifier. Switching MOSFETs are often designed to be used in swithing power supplies, when very low Rdson and gate capacitance is needed. A low capacitance equates to higher bandwidth, too.
MOSFETs can be harder to stabilize (you may need gate resistors and ferrite beads), and the leads should be kept as short as possible - much more so than BJTs. In Audio Amature magazine in the mid 80's, I think, astronaut Norm Thaggard joined with another author, I forgot, to design an all FET amplifier. I bet you can get a reprint of that article - it had a lot of great info about using FETs in an audio application.
MOSFETs can be harder to stabilize (you may need gate resistors and ferrite beads), and the leads should be kept as short as possible - much more so than BJTs. In Audio Amature magazine in the mid 80's, I think, astronaut Norm Thaggard joined with another author, I forgot, to design an all FET amplifier. I bet you can get a reprint of that article - it had a lot of great info about using FETs in an audio application.
Circlotron said:I think switching mosfets can sound really excellent. You can get really big ones too.
That's not to say that switching mosfets won't work in a linear amplifier, it's just that they may be harder to use. I'm convinced that for the mostpart any mosfet will work in audio applications. They do make some really big RF-Amp fets. I've often wondered about these being used as audio amplifiers, however I fear they would tend to oscillate.
Myself, I tend to use the IRFP240/IRFP9240 fets. They have good current and voltage capability, along with a nice power rating (125W ea, at Tc = 25C or so)
Jeff R is right about them being harder to stabilize. Becuse they require only voltage to change current, any noise coupled to the gate will be amplified greatly. In the case of a IRFZ34 mosfet, that change in current will be almost double that of an IRFP240 because the 240 has a lower I/V characteristic. (Both devices gate to source voltage around 5 volts). However, if the Z34 works, use it, it should have yield a better frequency response.
-Dan
Yeah... I see your valid points. That's one thing about the Z34 I really noticed, very good frequency response. It was almost phenomenal considering the fact that my amp design is simply the one fet operating class A with a 2SC1116A operating in the current source. Heh, this is my rendition of the Zen since I only felt like using one mosfet and made the current sink out of a bipolar. Perhaps the sound would be better with a mosfet current sink in there.
Later I accidentaly fried my Z34 by disconnecting an output from my computer and causing a bad static discharge right into the gate. I replaced it with the only other mosfet I had here, an IRF730. The 730 sounds good too, but it's made for low current and about 400 volts, not really class A material especially since my setup idles at about 1.5amps. Either way, I'm impressed with the ease of use and simplicity of mosfets.
I see what you mean about any little noises getting amplified greatly. I had this problem when I tried to connect the input of the amp right up to the computer. All that terrible sound card noise was amplified and was very loud and irritating!
I must also ask. what's a good topology to use to make a class A SE amp with BAL inputs? Is this feasible with simple circuitry like that found in the zen????
I want to build a mosfet amp with common drain outputs but still use mosfets in the VAS but I don't want feedback or complimentary output devices. Is it too much to bother running this topology idea in BAL mode?
Thanks
Later I accidentaly fried my Z34 by disconnecting an output from my computer and causing a bad static discharge right into the gate. I replaced it with the only other mosfet I had here, an IRF730. The 730 sounds good too, but it's made for low current and about 400 volts, not really class A material especially since my setup idles at about 1.5amps. Either way, I'm impressed with the ease of use and simplicity of mosfets.
I see what you mean about any little noises getting amplified greatly. I had this problem when I tried to connect the input of the amp right up to the computer. All that terrible sound card noise was amplified and was very loud and irritating!
I must also ask. what's a good topology to use to make a class A SE amp with BAL inputs? Is this feasible with simple circuitry like that found in the zen????
I want to build a mosfet amp with common drain outputs but still use mosfets in the VAS but I don't want feedback or complimentary output devices. Is it too much to bother running this topology idea in BAL mode?
Thanks
Well, I thought of a good design to use now. I'm actually going to mix the Zen amp with the Son of Zen amp. Basically, I want a differential amp like the s of z, but I want current sinking like in the Zen. That way, I have a more linear style of amp and it's going to be balanced like I want. That way, I'll have a nice low noise amp and still be in class A and SE modes as well!
What do you guys think of this idea???
What do you guys think of this idea???
Duo said:
Hey, any amp you build will probably sound better than a cheap one purchased...
As for fets, they make good output devices, but suffer from nonlinearities and changing DC offsets with changing temperatures (Unless you are using a matched pair, which reduces Offsets)
I think that it is a good idea to use a differential input, and negative feedback to reduce nonlinearities at large signal swings. However, unless you need high input impedances, It may be worth your time to investigate a BJT differential input stage (inherently less offset due to mismatching of transistors).
As for the complementary devices, if you run the FETS in Class AB, with around 50 to 150mA bias current, good linearity can be achieved. Also, negative feedback will help greatly.
As for the current sink, I'm not sure how it will affect sound. From what I've been playing with over the last few days, I'm finding Bipolar current sinks to be pretty good (At least when using a current mirror as a current sink. Again, what I'm playing with are at small signal levels (I out = 3mA). What I'm seeing is around 2-20 uA (F = 500Hz to 500Khz) of change in current for swinging outputs of 20Vp-p (Staying well within the current sink's compliance). At 50Khz, The numbers work out to be 5uA/3Ma or around 0.16% regulation. Not too bad.
I'm Not sure how this will be affected by larger current, but I suspect that performance may improve. Also, you may look into a cascode current source, since that should improve regulation as well.
Fets are nice, since you need no drive current other than to overcome gate charge, and gate to drain and source capacitances. For improved frequency response it may be worth looking at a bunch of fets and selecting them with the lowest drain to gate capacitance...
-Dan
Which to use Mosfets or BJT?
Hello, I am new to this DIY stuff just built a LM1875 amp with Class B(AB?) 175 watt Bass section. Am thinking of doing a Class A. Rod Elliot says Mosfets sound like crap. Why does he completely dismiss them? I like the idea of building any of JLH, Zen(FET), or DoZ amps but cannot decide which way to go. Any help on this appreciated. Thanks.
Greg Olsen
Calgary
Hello, I am new to this DIY stuff just built a LM1875 amp with Class B(AB?) 175 watt Bass section. Am thinking of doing a Class A. Rod Elliot says Mosfets sound like crap. Why does he completely dismiss them? I like the idea of building any of JLH, Zen(FET), or DoZ amps but cannot decide which way to go. Any help on this appreciated. Thanks.
Greg Olsen
Calgary
Re: Which to use Mosfets or BJT?
I wouldn't let Mr. Elliot's opinion sway you.
He seems to be a passionate guy and I
wonder if something he measured adversely
biased his listening. (Can the Placebo
Effect swing both ways? I would imagine
that it can...)
Give a listen to something like a Pass Labs
Aleph 3 if you get a chance. Or build the
Zen 4 or the "Aleph-X" that is being talked
about quite a bit in the Pass Labs forum
Check out the Class A Amplifier site if you
want to build one of the variations on the
JLH 10 watter.
(http://www.gmweb.btinternet.co.uk/index.htm)
If you can build both, try them out and see
which you like better. (I'm leaning toward
building an Aleph X and one of the latter
day JLH class A amps... hopefully I can
keep the power supply common...)
Erik
GringoAudio said:Hello, I am new to this DIY stuff just built a LM1875 amp with Class B(AB?) 175 watt Bass section. Am thinking of doing a Class A. Rod Elliot says Mosfets sound like crap. Why does he completely dismiss them? I like the idea of building any of JLH, Zen(FET), or DoZ amps but cannot decide which way to go. Any help on this appreciated. Thanks.
Greg Olsen
Calgary
I wouldn't let Mr. Elliot's opinion sway you.
He seems to be a passionate guy and I
wonder if something he measured adversely
biased his listening. (Can the Placebo
Effect swing both ways? I would imagine
that it can...)
Give a listen to something like a Pass Labs
Aleph 3 if you get a chance. Or build the
Zen 4 or the "Aleph-X" that is being talked
about quite a bit in the Pass Labs forum
Check out the Class A Amplifier site if you
want to build one of the variations on the
JLH 10 watter.
(http://www.gmweb.btinternet.co.uk/index.htm)
If you can build both, try them out and see
which you like better. (I'm leaning toward
building an Aleph X and one of the latter
day JLH class A amps... hopefully I can
keep the power supply common...)
Erik
I built a new revision of my altered zen concept, it sounds very very good. This time I'm using an old IBM computer PSU, 38 volts @25A. The sound is extremely detailed and musical. The amp simply disappears from the sound. The mosfet is still the IRF730 and I use a high quality sanken 2sc1116a for the current sink and a BC 170B low noise transistor for the current sink bias control.
It seems to have a lot of power and is able to drive very low impedances with ease. I could add extra speakers on top of the original one while the amp was playing and I noticed no volume reduction in the first speaker at all. The current handling is simply phenominal.
As far as I'm concerned, the Zen topology and its concepts are very unique and practical indeed.
It seems to have a lot of power and is able to drive very low impedances with ease. I could add extra speakers on top of the original one while the amp was playing and I noticed no volume reduction in the first speaker at all. The current handling is simply phenominal.
As far as I'm concerned, the Zen topology and its concepts are very unique and practical indeed.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.