Well, I tried a taste of some ohm-slaw and transistor theoryand got up a crazy idea to design and build a set of 450 watt low gain amplifiers for my speakers which are rated at about 600 watts rms. Trust me, I won't need to go as far as 600 watts unless I'm trying to level a building, which can wait till I'm really sure of what I'm doing
... Anyhow, I'd like some opinions from the more experienced people out there on this amp I designed before I attempt to build it. Please look at the attached diagram and say if you thinkit'll even power up without frying and also what might be done to make it a better amp. I biased it so that each of the 2n3055's(except for the darlington drivers) is idling at one amp, making a Q current of 10 amps just for the main O/T transistors. Thanks for
any suggestions or to point out that I'm completely nuts!
Duo
Well, I must say that schematic barely resembles anything suitable for a 450W amp. It looks more like something you'd find in a tiny cheap radio or a Gameboy or something. I don't think that design would work so well for a power amp, but you should take it from the more experienced people here than me. I see it's class A too, you're going to be dissapating a LOT of heat if you do it that way. Plan on cooling it with a pair of truck radiators? I've never heard of a 450W amp with that much bias current. Also, why don't you use differential input stage? I see you have no global NFB either. You also need a way higher supply voltage, you need plenty of headroom, it's not based just on ohm's law alone. You need a lot more than ohm's law and transistor theory to design a 450W amp. There are so many things that modern amps incorporate into their designs that your circuit lacks.
To get a good foothold in audio design, you should check out web sites like http://sound.westhost.com, he has a page there on audio amp basics, and read some books on amplifier design, like Randy Slone's "High Power Audio Amplifier Construction Manual" and Doug Self's "Audio Power Amplifier Design Handbook". I remember when I was where you are, it wasn't so long ago. You just need to read up a little more.
Good Luck!
P.S. I'm building some 450W amps too, but they're for 2 ohm operation. I have quite an extensive thread on it ("1kW MOSFET amplifier") you should check it out. My reputation here still suffers from this project, (it was worse before I decided to make 14x450W amps and 8x112W amps instead of 8x900W into 12 ohm amps) so watch out if you dive into something big too early, people will critisize you mercilessly if you don't follow their every bit of advice.
P.P.S. It's great to have more young people like myself here. I'm 16. Even though I think it scares some of the elders.
I've never heard of a 450W amp with that much bias current. Also, why don't you use differential input stage? I see you have no global NFB either. You also need a way higher supply voltage, you need plenty of headroom, it's not based just on ohm's law alone. You need a lot more than ohm's law and transistor theory to design a 450W amp. There are so many things that modern amps incorporate into their designs that your circuit lacks. To get a good foothold in audio design, you should check out web sites like http://sound.westhost.com, he has a page there on audio amp basics, and read some books on amplifier design, like Randy Slone's "High Power Audio Amplifier Construction Manual" and Doug Self's "Audio Power Amplifier Design Handbook". I remember when I was where you are, it wasn't so long ago. You just need to read up a little more.
Good Luck!
P.S. I'm building some 450W amps too, but they're for 2 ohm operation. I have quite an extensive thread on it ("1kW MOSFET amplifier") you should check it out. My reputation here still suffers from this project, (it was worse before I decided to make 14x450W amps and 8x112W amps instead of 8x900W into 12 ohm amps) so watch out if you dive into something big too early, people will critisize you mercilessly if you don't follow their every bit of advice.
P.P.S. It's great to have more young people like myself here. I'm 16. Even though I think it scares some of the elders.
HI Duo. This is a pretty safe forum for putting forth ideas you might consider "crazy". The people here will only "criticise you mercilessly" if you know little *and* shoot your mouth off arrogantly. Knowing little is not in itself a bad thing. Also if someone unjustly criticises you, watch and see if hordes of others don't come to the rescue.
As for your amplifier, I think you are asking *way* too much of '3055's. Look up "safe operating area" or "SOAR". Your'e gunna pop them for sure. Thomas Edison said that invention is "1% inspiration, 99% perspiration". Power electronics, including amplifier design is all blood sweat and tears, all three factor vying for supremacy. Other than that, keep thinking amps, dude! You'll get there!
GP.
"With power electronics you are always only a microsecond away from misery".
As for your amplifier, I think you are asking *way* too much of '3055's. Look up "safe operating area" or "SOAR". Your'e gunna pop them for sure. Thomas Edison said that invention is "1% inspiration, 99% perspiration". Power electronics, including amplifier design is all blood sweat and tears, all three factor vying for supremacy. Other than that, keep thinking amps, dude! You'll get there!
GP.
"With power electronics you are always only a microsecond away from misery".
Duo, since you are 16 I suspect that you know not as much as a pro.
First of all: If you have an idea and can't analyze it theoreticly test the circuit with small signal transistor. If can't get it work, it probely won't work with higher voltage and bigger transistors.
My opinion of your idea:
You have no control over the DC-level (at the output) and working currents.
The input impedance is practically nothing (may be a problem if the signal source can't deliver enough)
The output stage is working in class-A (is it by purpose?), lots of heat! Class A is not a bad thing if yoou really want it.
Worst of all: The distortion will probely be horrible.
Conclusion: You must create stable working conditions. Chosen transistors are slow, the amp may be too slow.
My best advice is to look around and see how have others have done. Check my stuff, check Nelson Pass' stuff, Check Rod Elliott's stuff, check Lars Clausen's amps, check schematics from commercial amps (I have learned of lot by that). Try also to get knowledge about each transistor circuit and feedback theory, otherwise it will be only a struck of luck if you succeed. Start to learn how a simple standard amp works, then it will be easier to develop ideas of your own.
http://www.lcaudio.dk/
http://www.passdiy.com/
http://www.sound.au.com/
First of all: If you have an idea and can't analyze it theoreticly test the circuit with small signal transistor. If can't get it work, it probely won't work with higher voltage and bigger transistors.
My opinion of your idea:
You have no control over the DC-level (at the output) and working currents.
The input impedance is practically nothing (may be a problem if the signal source can't deliver enough)
The output stage is working in class-A (is it by purpose?), lots of heat! Class A is not a bad thing if yoou really want it.
Worst of all: The distortion will probely be horrible.
Conclusion: You must create stable working conditions. Chosen transistors are slow, the amp may be too slow.
My best advice is to look around and see how have others have done. Check my stuff, check Nelson Pass' stuff, Check Rod Elliott's stuff, check Lars Clausen's amps, check schematics from commercial amps (I have learned of lot by that). Try also to get knowledge about each transistor circuit and feedback theory, otherwise it will be only a struck of luck if you succeed. Start to learn how a simple standard amp works, then it will be easier to develop ideas of your own.
http://www.lcaudio.dk/
http://www.passdiy.com/
http://www.sound.au.com/
Hmmm
OK, I know what you guys mean, that's kinda what I was thinking about, overcurrent for the transistors and problems with control... I wasn't sure about the 3055's being slow because I can't find any good spec sheets on them... By the way, the design was intended to be class A and run quite hot, (I calculated that it would idle about 2500 watts at the connection to the electrics... Quite a light dimmer, i'd need a dryer outlet for it...
Anyway, back to the point, even if I had some good spec sheets(with curves) and some literature, I'm pretty sure I could make something that's work... I'm just real bad at finding that kind of information.
OK, I know what you guys mean, that's kinda what I was thinking about, overcurrent for the transistors and problems with control... I wasn't sure about the 3055's being slow because I can't find any good spec sheets on them... By the way, the design was intended to be class A and run quite hot, (I calculated that it would idle about 2500 watts at the connection to the electrics... Quite a light dimmer, i'd need a dryer outlet for it...Anyway, back to the point, even if I had some good spec sheets(with curves) and some literature, I'm pretty sure I could make something that's work... I'm just real bad at finding that kind of information.
Actually, I thought of something I should of mentioned earlier..
I'm not really aiming for perfect audio characteristics just yet,
even if I got it to power up and add some dirty current to a signal, I'd be quite happy. Also, the main reason I picked the 3055's is because I have about 30 of them in a tray on my bench.
Though I do have 4 pairs of A1492 & C3856, do you guys think those transistors would be a good choice for such great power output? Also, my intention was to drive the amplifier from the output of another hifi amplifier (since they are designed for low impedance loads) and even if it would work for a sub amplifier (100hz and down) I'd be satisfied. I could later add things to make it better and better until I am sure of how everything works.
I'm not really aiming for perfect audio characteristics just yet,
even if I got it to power up and add some dirty current to a signal, I'd be quite happy. Also, the main reason I picked the 3055's is because I have about 30 of them in a tray on my bench.
Though I do have 4 pairs of A1492 & C3856, do you guys think those transistors would be a good choice for such great power output? Also, my intention was to drive the amplifier from the output of another hifi amplifier (since they are designed for low impedance loads) and even if it would work for a sub amplifier (100hz and down) I'd be satisfied. I could later add things to make it better and better until I am sure of how everything works.
Duo,
Education and Understanding is key.
2 books: Pick 1 of the 2. For you, I would say the Randy Slone book. He is more PA and musical instrument amp oriented. Seems to understand the need for high power.
Randy Slone: Audio Power Amplifier Design Handbook
Doug Self: Audio Power Amplifier Design Handbook
A very good book and worth every penny: You should ge this.
Paul Horowitz, Winfield Hill: The Art of Electonics
Aud_Mot
Education and Understanding is key.
2 books: Pick 1 of the 2. For you, I would say the Randy Slone book. He is more PA and musical instrument amp oriented. Seems to understand the need for high power.
Randy Slone: Audio Power Amplifier Design Handbook
Doug Self: Audio Power Amplifier Design Handbook
A very good book and worth every penny: You should ge this.
Paul Horowitz, Winfield Hill: The Art of Electonics
Aud_Mot
Re: Hmmm
Try http://www.questlink.com a goldmine for pros and amatuers.
Check all link collections, check MY link page! Check this site's link collections!
Try search engines. Google is good, IX search is good, Dogpile OK, Altavista OK
http://home5.swipnet.se/~w-50719/links_pa.html#searchengines
Duo said:
Try http://www.questlink.com a goldmine for pros and amatuers.
Duo said:
Check all link collections, check MY link page! Check this site's link collections!
Try search engines. Google is good, IX search is good, Dogpile OK, Altavista OK
http://home5.swipnet.se/~w-50719/links_pa.html#searchengines
Duo said:
Hey, hey, hey, you really ARE crazy!
Seriously, go for the concept AMPLIFIER, not a buffer. First, if you anyway will try that, how much gain do you need. I'd say not more than 5. Now you have 2000 (haven't done any calculating). Roughly 1 mA + 1kohms = 40, 10 mA - >400. 50 mA -> 2000. The speed of the transistor is around a couple of MHz. This means bandwith approx. a few kilohertz!
The DC at the output? Avoid big caps in series with the load. >10000 uF/50-100V.
If you are going to do a bass amp, it's VERY easy with MOSFET's because the demand very little drive power. The amp I have to my subwoofer needs only 4 mA at 160 V! delivers 500 Watts
Do I shoot my mouth off arrogently? Nah, if you want to hear that, go follow SliceMaster101's threads. My problem has just been that people would say I couldn't do it and I was asking for trouble (I had this direct-line rectification thing, 900W amps with +/-160V supply, series-parallel speakers, and stuff) and I wouldn't take it from them. Just don't be like that and you'll be fine. I've just been kind of think-headed, not arrogent, I think people will agree. Now that I'm on the right track (well, better), some people can't quite grasp the concept. I'm no expert, but I do try to help with whatever I do know.
Anyway, like Peranders said, you need an amp, not a buffer, and some MJ15003/MJ15004 output devices, or maybe MOSFETs if you really do want class A. Power amps tend to be just like a big descrete op-amp.
Just don't be like that and you'll be fine. I've just been kind of think-headed, not arrogent, I think people will agree. Now that I'm on the right track (well, better), some people can't quite grasp the concept. I'm no expert, but I do try to help with whatever I do know. Anyway, like Peranders said, you need an amp, not a buffer, and some MJ15003/MJ15004 output devices, or maybe MOSFETs if you really do want class A. Power amps tend to be just like a big descrete op-amp.
I would think so, but of course you could still do class A if you really want to, but at 450W, I don't know what you would do with all that heat, unless you use water cooling.
P.S. Aud_Mot, it's not "Randy Slone: Audio Power Amplifier Design Handbook" He wrote the "High Power Audio Amplifier Construction Manual" They are very long titles, I know.
P.S. Aud_Mot, it's not "Randy Slone: Audio Power Amplifier Design Handbook" He wrote the "High Power Audio Amplifier Construction Manual" They are very long titles, I know.
Kilowatt said:
Hi Kilowatt.
I've got to say you have really grown and matured in this art. Your comments to Dug have been helpful and right on the money.
P.S. I never thought you were arrogent, just not very likely to see your 17th birthday. ;^) Not everyone can learn from the experiances of others, glad to see you can.
P.P.S. Can't wait to hear your outdoor system when you get it built. I expect I might be able to hear it from my home in South Carolina. (Joke) "imitating Jocko"
Phil
Sure it is. It's just that Mr. Pass likes single ended MOSFET class A, I've heard those sound sweet, but I haven't actually heard any. Mr. Slone has a complementary 40W class A design that can do 0.0018%THD (and he doesn't exaggerate). You say you have 4 pairs of A1492/C3856, you could make a 2 channel version of something like that and use them, you'd have just enough.
You mean a two channel version at 40 watts or at a much higher level... I don't need to work for things lik .0018% THD just yet but I would like moderate quality... I have heard many class ab amps that sound wonderful, right now, I use a Harman/Kardon HK A402
with 60 watts a side which is AB and it has a very clean finesse to it's sound, I got it from an audiophile with a pair of ATC SCM100 speakers with builtin pure class A mosfet amps, triamped with 400 watts an amp... He says that the Harman/Kardon has comparable quality to the Krell amps in the ATC's. I would be happy to design a push pull amp with more power, one aim is to be able to give my speakers more power so they can really pack their punch, I don't care if the quality of the amp isn't that of pure class A, but I also don't want to fry my speakers, which would also be quite hard to do... I used to have an amp that had four channels, each class AB using an a1302 and c3856 pair which was bridgeable to stereo with 350watts a side into two ohms...
But I don't have it anymore...
with 60 watts a side which is AB and it has a very clean finesse to it's sound, I got it from an audiophile with a pair of ATC SCM100 speakers with builtin pure class A mosfet amps, triamped with 400 watts an amp... He says that the Harman/Kardon has comparable quality to the Krell amps in the ATC's. I would be happy to design a push pull amp with more power, one aim is to be able to give my speakers more power so they can really pack their punch, I don't care if the quality of the amp isn't that of pure class A, but I also don't want to fry my speakers, which would also be quite hard to do... I used to have an amp that had four channels, each class AB using an a1302 and c3856 pair which was bridgeable to stereo with 350watts a side into two ohms...
But I don't have it anymore...
With class A you couldn't get more than maybe 80W total out of those transistors, but class AB you could get like 300 or 400W. I assume you want to go class AB, and class AB amps can be made to have like 0.003% or less THD, but it's fairly complex if you want numbers like that (and numbers aren't everything anyway). Even a really simple class AB amp can sound great though, like that project3A from your other thread, for example. You could get like 700-800W total with those 30 2N3055s (a pair will make a 60W amp), even though they do kind of suck by today's standards. Whatever you do, I hope it turns out great.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.