Very simple Comparator Power Supply.

[Psi]

Where on the universe do you get a Transformer, that gets 100V,90V,GND,-90V (after rectifying), that pulls that out?

If i wanted those voltages at ~1kW for a hobby project i would probably get a 1kW toroid transformer and add too or change the secondary windings.
Most have the secondary on the outside which means you don't have to touch the primary and 1kW toroids have a nice big hole in the center which makes them easy to rewinding. You can pass a small reel of wire through the hole.

[Flunze]

Unfortunately a simple 1kW Toroid transformer costs 139€. And the Secondary is not even yet rewound. If it was cheaper, I would have bought one, because Toroids are most used for Amplifiers.


A Variac would be good enought for that, but unfortunately they are not any cheaper and it is not mains insulated, If I don't play with it and reconstruct it.

[Psi]

mm... Toroids

[Astroplio]

What I have learned so far is that you can't have a "simple", "cheap" and "a 1kW" power supply at the same time.
So now the question is, if the poster is willing to face this reality or not. And if not, then why post in a public forum and ask for feedback?

Waiting for the next episode, "will it fry?" 

[poptones]

900W out of the amp, so I want a supply with more than 1kW.

You're making a 900W amp and you're trying to skimp on the power supply?

It isn't an amplifier with a bipolar design, so it will not do that at all.

If it isn't "bipolar" then you'll also need output coupling caps.

So, your goal is to build a 900W amp that uses a single supply rail, has cheap coupling transformers on the input and uses coupling caps on the output...

here. That, a car battery, and a charger will get you there for 200 bucks. And it won't kill you.

[DarkPrince]

Hello O.P.

I am guessing you don't have an educational background in EE/EET... I wouldn't recommend a design of this scale but I can only recommend not to do so. May I suggest as a last hope, just play around with a circuit simulator (such as the free, LTspice or the many other spice simulators). It could be life saving.


On a site note, simulating 1mF post-rectification is amusing; A heart-warming 50kW power surge.. gotta love it.

[AndyC_772]

To anybody, who wants to say, that this cannot work, explain why.

OK, I'll bite - but only because I'm genuinely concerned that you're going to start a fire, electrocute yourself, or both.

The non-inverting input of the op-amp is grounded. That'll break it straight away, though I presume it's just a schematic error.

The comparator is powered through a network which has a source impedance of about 10k, yet it's expected to drive a 100R load. Any output current has to come from the comparator's power supply, so as soon as the comparator output rises about zero it'll pull its positive supply voltage down almost to ground. The comparator is, therefore, turned off.

With the comparator off, so is Q2. This means your output terminals are: +325V, +325V connected through 10,000uF (which is 0.3 Ohms @ 100Hz), and +325V connected through 5000uF (0.6 Ohms @ 100Hz). You should label these pins "lethal", "lethal" and "lethal", in that order.

Please, take DarkPrince's advice and try your design out in simulation first - it'll save a lot of grief, perhaps quite literally.

LTSpice is free from Linear Technology (www.linear.com) and it's a really superb tool that I can't recommend highly enough. Download it, draw up your circuit and have a play.

[David_AVD]

I think "OMFG!" is the correct response to this design.  Surely a Darwin Award in the making here. 

[Flunze]

With the comparator off, so is Q2. This means your output terminals are: +325V, +325V connected through 10,000uF (which is 0.3 Ohms @ 100Hz), and +325V connected through 5000uF (0.6 Ohms @ 100Hz). You should label these pins "lethal", "lethal" and "lethal", in that order.

Nope. How should any current flow through the mosfet, if it is turned off?




The components for the whole Amplifier cost me just 50€. The Transformer is over 2 times heavier in price (139€). Well, I will buy a Toroid converting 230VAC to 115VAC and modify the secondary side. >> It's not, that I have any other stuff to pay.  <<

[Tepe]

Unfortunately a simple 1kW Toroid transformer costs 139€. And the Secondary is not even yet rewound. If it was cheaper, I would have bought one, because Toroids are most used for Amplifiers.

Use an SMPS like many commercial  kW amps do. It's also better for your electricity bill.

[Flunze]

[idiotsection]

edit: this is what happened when a trolling bot havent felt what an electric shock is.

Shut up, I am pretty sure, I know what electric shocks are. I had about 30 big ones.

(he explained everything and missed R90 and R91)

Are you dumb? There is no R90 and R91 in Picture 2. And what did I just say?

Please let's just think about Picture 2. I will explain it completely.

YES! That's what she said!
[/idiotsection]

Use an SMPS like many commercial  kW amps do. It's also better for your electricity bill.

2 Facts about that: There is no SMPS with that Voltage Outputs on the planet. A SMPS for 1 KW would cost 200€, way more expensive than a toroid transformer. But thanks for a finally constructive comment.

Test your circuit out using a low voltage AC source like a wallwart scale your components accordingly. When you correct all the errors and get it to switch relably then scale up.If your goal is a Hysteretic type regulator here's a hint you need hysteresis.Then record your scaled up test for prosperity.

What mosfet's do you plan on using?

I will not do that, because I am to pissed of by the other freaks just complaining about air. But thanks for a realy, realy, constructive comment! That is, what makes a forum good!
I would have used IRFP460, because I also use them in the Amp and "mass" order them for that. ^^


But hey, I will use a toroid transformer, this is way more practical for a forum with a dumbass:smart ratio of about 9:1, everybody will be happy with that.

[T4P]

Unacceptable abuse of another member removed by moderator. Another outburst like this will earn the user a 1 week ban !

[Flunze]

Because 40% of all the Amps, I saw have toroids, 30% have a normal transformer and 30% a SMPS. Why do you say, I am not smart? You are not living in germany, so how can you know, what the industry sells here? Don't call the chinese smarter than me as insult. Chinese CAN produce high quality, if the industry gives the order and money to do that. The only reason, why most chinese products are bad, is because the industry does not want to pay much money at all.


A toroid also is the cheapest way and can be rewound easy, without disassembling the whole transformer core.

Just read my last post, where I wrote down the prices, than you see, that a toroid makes perfect sense.

[Bored@Work]

Why do you say, I am not smart?

He is probably saying this, because everything you did show and wrote here, and how you behave doesn't look smart at all.

You are not living in germany,

Living in Germany does not automatically make you smart. You just demonstrated that.

[EEVblog]

NOTE:
I have removed the schematic and PCB images on the grounds of safety.
I'm all for publishing unsafe designs as learning examples or whatever when clearly labelled and explained etc, but it is too easy for someone to stumble upon this thread in the future and think this is a good or safe circuit to use. It is not.

Yes, I'm playing "information police" here which I absolutely hate to do, but I think it is necessary.

Dave.

[Flunze]

You are not living in germany,

Living in Germany does not automatically make you smart. You just demonstrated that.

First of, it had nothing to do with being smart, but the fact, that a toroid is used very often. Secound, I behave like that, because people call my thoughts shit, without explaining it - most likely they don't explain it, because they do not understand it.

Lets assume you were able to switch your mosfets what do you think will happen with no hysteresis and a weak drive circuit and large input and output capacitance. I'll tell you your mosfets will go “POP” or your conductors or traces would cook,or your rectifiers its a crap shoot what would go first.

Popping the mosfet throught that high current... Finally a comment, that realy explains, why this can not work... The rectifier will not hold that current, also a good reason. Well, it would not work that simple at all. Understood! The only thing, that would not pop, would be the PCB, because I got a high current solution for that. xD

Yes, I'm playing "information police" here which I absolutely hate to do, but I think it is necessary.

Dave.

Nobody is gonna blame you for that. AcHmed99 explained, why it is not possible, it works, so this is ok.

~~~~~~~~~

Well, like sayd before: I will buy a toroid transformer and rewind the secondary side. This is not cheap at all, but the cheapest way, to do it.

[T4P]

I have probably designed far more power supply sections for amps then you do, you think i won't understand? Different topologies but all leading to the same thing : safety. You have violated multiple laws of safety from the first post and still disregarding safety up till the moment you decided to buy a toroid, what was wrong with that?

[Flunze]

Whoot? My comment is removed, or did I forget to sent it?

Well, I have nothing against safety, but it is useless, if an amplifier outputs lethal Voltage. And call 90V not lethal. It is not as dangerous as 325V i.e., but still dangerous and unsafe to touch.


If ATX Supplies would be very cheap, I would put them in series, with disconnecting the earth from all, but one. This would work with no problems at all, but is not cheaper at all, it might be the most expensive version. Well, at least you get a regulated voltage. ^^
But is it save, to pull so much power out, without using fans? If you use fans for all, you would have such a loud sound, that the amp is useless. I used a very old ATX supply for my small Subwoofer, but it is just pulling out about 15Watts max, so there will not be a dangerous temperature rise. Pulling out all the power or at least 50W per supply could be dangerous without proper "heat extraction".


So, a toroid is not cheap, but the cheapest solution for my problem. Also: You will be happy about the "safety", because it is mains insulated. Why do SMPS for that high power and "high" voltage outputs cost so much? way over 200€ is not cheap or even close to that.

[GeoffS]

Well, I have nothing against safety, but it is useless, if an amplifier outputs lethal Voltage. And call 90V not lethal. It is not as dangerous as 325V i.e., but still dangerous and unsafe to touch.
.

Statements like these are the reason that your posts are edited/deleted. Go and read up about the effects of electricity on the human body before commenting on 'safety'.

[David_AVD]

Well, I have nothing against safety, but it is useless, if an amplifier outputs lethal Voltage. And call 90V not lethal. It is not as dangerous as 325V i.e., but still dangerous and unsafe to touch.

The 90V is referenced to the mains input.  That's what most people are trying to warn you about.  Even 90V can be deadly.

[Monkeh]

Why do SMPS for that high power and "high" voltage outputs cost so much? way over 200€ is not cheap or even close to that.

Because they put actual engineering and proper components into them? And then they test them, and give you nice manuals, and spend a massive amount of money on compliance testing, and have custom housings, and shipping, and advertising..

Nah. None of that can cost anything.

[Tepe]

Just wait till you calculate the necessary cooling system for the amp. It will not be cheap.

[GeoffS]

better calculate your grave's cooling system. @flunze: we dont workout from pcb (picture2) first, we work out from schematics first (picture1)... awaiting troll...

Both images have been removed.

[Flunze]

Just wait till you calculate the necessary cooling system for the amp. It will not be cheap.

I got a few pounds of Aluminum heatsinks all screwed togeteher. You can easily dissipate a few hundreds watts with them and they would only get 20Kelvon hotter. That is not a problem at all! I better take more cooling, than not enought.



It's about the same mass as in this picture, just a bit bigger and that was for an 1000W Amplifier.



Unnecessary comment removed

[Tepe]

I got a few pounds of Aluminum heatsinks all screwed togeteher.

What class amp are you building? A, B, D, ...? It might be fun to do a back-of-the-envelope calculation of the dissipation.