Sorry for going on-topic.
When i finish my exams next month i will try to post a pdf no longer than two pages telling how to do a psu in veroboard (the pads are useful in the psu) and not covering controversial safety issues other than "don't touch a live wire" and be sure everything is unplugged before making any modifications.
The idea is to give it the minimum requiered features to be acceptable (6600 uF/rail, four diodes, a snubber and a pair of noise-killer 0.22 uF at the secondaries*)
*These are imprescindible unless you want the amplifier to warn you each time somebody turns on a light in the building.
About safety, there are three things to worry about:
( 1 ) Improoper isolation
( 2 ) Failure in a non--diy component [transformer] provoking the presence of live mains where it shouldn't be.
( 3 ) Fire risk: I have repaired a lot of non-diy (commercial) stuff and i don't leave home/go to sleep with anything plugged on. The weirdest example of what could happen was the operating system ( a black & white one, but it is an operating system ) of my mobile phone hanging and provoking the battery not to stop loading when it reached full charge [guess the fire? I didn't build my mobile phone!] And i belive that the software causing that failure didn't have to pass any safety controls. Other curious examples of non-diy stuff becoming incredibly dangerous was the laptop of a friend whose screen broke. I found her happily chatting with it using an external monitor. From the laptop emerged a couple of thiny wires, a pink one and a white one. For those unfamiliar with laptops these wires come from the screen lighting inverter and carry 800-1200V no load and can supply 50 mA.
So there is no reason why a gainclone psu should not be at least as safe (i feel mines are much more safe) than the psu from a commercial piece of equipment.
If somebody writes something you don't like ignore it. If you feel you don't have the skills to build safe gear don't build it.
My opinion about Dan is he is an amplifier. I havn't seen him writting a single thing that wasn't already out there (*), so he isn't the problem. Maybe he has too much gain so safety or important things get clipped and equaled to less important-arguable-subjective opinions. The solution is obvious: He needs feedback.
(*) even in a thread talking about capacitors showing a "fletcher-munson" effect, the same (all forbidden words here) stuff lies in Joe Rasmussen audio pages since much before Dan got registered here.
When i finish my exams next month i will try to post a pdf no longer than two pages telling how to do a psu in veroboard (the pads are useful in the psu) and not covering controversial safety issues other than "don't touch a live wire" and be sure everything is unplugged before making any modifications.
The idea is to give it the minimum requiered features to be acceptable (6600 uF/rail, four diodes, a snubber and a pair of noise-killer 0.22 uF at the secondaries*)
*These are imprescindible unless you want the amplifier to warn you each time somebody turns on a light in the building.
About safety, there are three things to worry about:
( 1 ) Improoper isolation
( 2 ) Failure in a non--diy component [transformer] provoking the presence of live mains where it shouldn't be.
( 3 ) Fire risk: I have repaired a lot of non-diy (commercial) stuff and i don't leave home/go to sleep with anything plugged on. The weirdest example of what could happen was the operating system ( a black & white one, but it is an operating system ) of my mobile phone hanging and provoking the battery not to stop loading when it reached full charge [guess the fire? I didn't build my mobile phone!] And i belive that the software causing that failure didn't have to pass any safety controls. Other curious examples of non-diy stuff becoming incredibly dangerous was the laptop of a friend whose screen broke. I found her happily chatting with it using an external monitor. From the laptop emerged a couple of thiny wires, a pink one and a white one. For those unfamiliar with laptops these wires come from the screen lighting inverter and carry 800-1200V no load and can supply 50 mA.
So there is no reason why a gainclone psu should not be at least as safe (i feel mines are much more safe) than the psu from a commercial piece of equipment.
If somebody writes something you don't like ignore it. If you feel you don't have the skills to build safe gear don't build it.
My opinion about Dan is he is an amplifier. I havn't seen him writting a single thing that wasn't already out there (*), so he isn't the problem. Maybe he has too much gain so safety or important things get clipped and equaled to less important-arguable-subjective opinions. The solution is obvious: He needs feedback.
(*) even in a thread talking about capacitors showing a "fletcher-munson" effect, the same (all forbidden words here) stuff lies in Joe Rasmussen audio pages since much before Dan got registered here.
There is more people apart from us who have found writing psu tutorials to be really funny:
http://www.national.com/an/AN/AN-1849.pdf
Perhaps not for the youngest newcomers but really really interesting read.
http://www.national.com/an/AN/AN-1849.pdf
Perhaps not for the youngest newcomers but really really interesting read.
A. Pope said:
I nearly blew my fingers off with a squib made from guncotton.
I thought I'd permanantly damaged my hearing when I hit some potassium chlorate and brimstone mixture with a lump hammer.
Building electronic devices can be dangerous.
Serious accidents are fortunately rare.
I can't remember how many shocks I've survived. I once had RF from a 12 watt transmitter with an antenna tuner drill a little hole right through my ring fingernail. Oww!
It made me a bit more cautious with the 100W ones though.
I wouldn't want to deprive anyone the sense of exhilaration it gives me to have experienced all these things and I don't seek to discourage other people from experimenting.
Sometimes, however, the questions asked on this forum betray a lack of knowledge or sufficient prior effort on the part of the questioners for a full answer to be accommodated in the scope of a short message.
For this reason replies may not be as to-the-point as questioners might wish. NOTHING, however, is better documented or more accessable on the web than electronics and computing.
If you want to know how to build a PSU you could do worse than read the Application Note previously linked.
w
I should probably just have taken Matthew's advice...
Matthew 15:14
"Let them alone: they be blind leaders of the blind. And if the blind lead the blind, both shall fall into the ditch."
I'm happy to see you have survived all those accidents...
I don't know if you are being ironic saying "you could do worse" Reading an application note does not hurt, and seeing what national feels needed (the amount of capacitance, the values for the snubbers, the absence of those rare circuit breakers that Daniel suggested) are a good starting point.
Knowing which cap value/voltage may be a good choice or why regulators aren't that good will not kill anybody.
The application note mentions the fact that it hasn't been tested for any safety regulation. I posted it here to give some values from a reliable source to prevent people from following anything they read here. You should have noticed the ones claiming that 1000uF is almost excessive. Belive that and you are doing worse.
For those who are still interested in building a psu after reading the whole thing, some advise about choosing the rectifier bridge.
I've just read a post where it says a bridge rated 35A can provide 70A of direct current. This is plain wrong, because capacitor loading happens in a spiky way.
The attached image shows the current that passes through a diode in a bridge rectifier that is powering a 50W/8Ohms gainclone.
Some clever people may thing 50W & 8Ohms means (I^2)R = 2.5A. Some cleverer may thing: Ipeak = sqrt(2)*Irms = 3.54A. If you are one of these then give a look at the plot.
EDIT: The red trace is the current the amplifier takes from the positive rail, green trace is the current that flows at the diode.
Maths are useless if they aren't driven by common sense.
EDIT2: Guess what happens to your masterpiece when one of these diodes becomes tired of it and shorts? I learned it the hard way.
I've just read a post where it says a bridge rated 35A can provide 70A of direct current. This is plain wrong, because capacitor loading happens in a spiky way.
The attached image shows the current that passes through a diode in a bridge rectifier that is powering a 50W/8Ohms gainclone.
Some clever people may thing 50W & 8Ohms means (I^2)R = 2.5A. Some cleverer may thing: Ipeak = sqrt(2)*Irms = 3.54A. If you are one of these then give a look at the plot.
EDIT: The red trace is the current the amplifier takes from the positive rail, green trace is the current that flows at the diode.
Maths are useless if they aren't driven by common sense.
EDIT2: Guess what happens to your masterpiece when one of these diodes becomes tired of it and shorts? I learned it the hard way.
Attachments
I guess what I'm trying to say is that it's sometimes a difficult balance to strike, when trying how to decide on how to reply to a question where safety hazards are involved.
I, for one, enjoy and feel duty bound to share what I know.
There are, however, some constraints on this. I put myself at hazard if I reveal trade secrets, for example.
I put myself at moral hazard if I recklessly encourage correspondents to embark on hazardous courses without some personal conviction that they truly appreciate the risks.
This is a pretty picky attitude given that this is an open forum and there is loads of risky material to be found if you eavsdrop on other peoples conversations, but given that many of the questioners have themselves not taken the time to eavsdrop on such conversations as is evidenced by the inanity of the questions I reserve the right not to hand everything to anybody on a plate.
w
I, for one, enjoy and feel duty bound to share what I know.
There are, however, some constraints on this. I put myself at hazard if I reveal trade secrets, for example.
I put myself at moral hazard if I recklessly encourage correspondents to embark on hazardous courses without some personal conviction that they truly appreciate the risks.
This is a pretty picky attitude given that this is an open forum and there is loads of risky material to be found if you eavsdrop on other peoples conversations, but given that many of the questioners have themselves not taken the time to eavsdrop on such conversations as is evidenced by the inanity of the questions I reserve the right not to hand everything to anybody on a plate.
w
It's not just safety issues that are difficult to explain in full. Where do you stop?
It reminds me of children when they are young when they go through that 'why?' phase. They ask you about something and you give them and answer and they reply "but why? And each explanation you elaborate on is met with another "why?".
Writing is a time-consuming process even when you can touch type, which I can't. And why write something down when it is already explained elsewhere. And (as has been said many times now) is available to all for free for the want a few clicks of the mouse.
I'm at my last year of the phisics degree and my opionion about maths is much more complicated that any single sentence. You can read "The unreasonable effectiveness of mathematics in natural sciences" and it's replies to get some background arround this 
.
But it wasn't my intention to go into philosophy, i was only trying to say that tweaking the part's values requieres some more understanding of how things work apart from some equations, and that applying blindly all those equations will probably end up in a worse result if not in an explosion. This has, of course, a good side: You don't need those equations to build a psu!
Let's push a bit the state of art. Shouldn't a 0R2-0R5 resistor between the rectifier and the main filter caps be a standard in all psu's? From what i see in spice it makes life much easier for the rectifiers apart from giving a cleaner FFT, the gainclone will not sound that quieter if feed 32V instead of 33, and the resistor is very cheap.
EDIT: I'm almost never ironic, but i don't know how things sound like as i'm not english.
. But it wasn't my intention to go into philosophy, i was only trying to say that tweaking the part's values requieres some more understanding of how things work apart from some equations, and that applying blindly all those equations will probably end up in a worse result if not in an explosion. This has, of course, a good side: You don't need those equations to build a psu!
Let's push a bit the state of art. Shouldn't a 0R2-0R5 resistor between the rectifier and the main filter caps be a standard in all psu's? From what i see in spice it makes life much easier for the rectifiers apart from giving a cleaner FFT, the gainclone will not sound that quieter if feed 32V instead of 33, and the resistor is very cheap.
EDIT: I'm almost never ironic, but i don't know how things sound like as i'm not english.
Yes indeed. And then there are the miscommunication problems of language barriers, terminologies, and even the typo that can have a profound effect.
It's the "ego" thing that causes most problems AFAIC. It's a barrier to being honest with yourself. THAT'S not going to help you learn and it certainly isn't helpful to constructive conversation.
Daniel was trying to do a good thing by starting these threads and I don't believe anyone really wanted to stop him. But the title itself put him in a tough spot for his level of expertise.
I've thought, as I watched his journey here, if he would only ask questions, instead of offering opinions, he would have progressed faster and further. And even thought I should play that part myself to help the thread live up to its name, by documenting a simple build, and letting the questions and answers show the many options that could be implemented.
I feel I owe something to this community, because I've built dozens of these amps with absolutely no prior electronics knowledge. In fact, I still don't know how or why these things work. I do know home wiring however, and how to follow directions.
I don't have the time either, to take on such a project. But there's really no need. It's been done and very well. Nuuk won't blow his own horn properly, but I will again. He's put years into his very generous sharing of information and I'll bet over ten thousand amps have been built because of his work.
The same generosity has been shown by Peter Daniel and many others. The amp boards that he and Brian Bell designed in public, and "group sold" are a gold standard for implementing not just help for the beginner, but also a quality product marketed in an honest way. Not to mention his remarkably artistic constructions.
There have been many ego problems displayed as these threads grew over the years. Read them too, as there's a lesson there too.
It seems to me the biggest problems for the beginner is where to start and collecting the parts. It's really not that hard. Pick a direction and take a step.....and oh yeah....ask questions and listen to the answers.
Thank you Onasis, that's generous praise indeed.
Funny thing is although I realised that I was helping some guys along the 'road', I never thought about how many GC's may be in existence due to Decibel Dungeon.
We should mention Thorsten Loesch as well. I believe that he was the originator of the IGC circuit. And let's not forget Gregg who ran the (very friendly) Chipamp.com forums. You have rightly mentioned PD whose work inspired me to 'have a go' originally.
And of course this very forum, probably the main vehicle for the spreading of the GC 'gospel'!
Funny thing is although I realised that I was helping some guys along the 'road', I never thought about how many GC's may be in existence due to Decibel Dungeon.
We should mention Thorsten Loesch as well. I believe that he was the originator of the IGC circuit. And let's not forget Gregg who ran the (very friendly) Chipamp.com forums. You have rightly mentioned PD whose work inspired me to 'have a go' originally.
And of course this very forum, probably the main vehicle for the spreading of the GC 'gospel'!
And not to forget Fedde Bouwman whose original website inspired me. He was actually one of the first people to promote GC on this Forum.
http://www.fedde.nu/audio/chill-amp.html
http://www.fedde.nu/audio/chill-amp.html
Nice solution
I know that many of you (except Nuuk who encourages that at Decibel Dungeon) will argue with me, but a great solution is to use a laptop smps. These are not extremely expensive, absolutely safe (better if you get a short-circuit proof one) and not as noisy as many people thing (use long wires and give them a turn to a ferrite toroid to get some extra noise immunity).
The 470uF-1000 uF caps on the board will care about the high frequencies.
I have done that and found absolutely no problem! No dangerous wiring at all!
(shouting): USE ONLY PSU'S THAT ARE NOT SAFETY EARTHED!!!!!!!!!!!!!!!!
These are not a problem to find in spain.
I know that many of you (except Nuuk who encourages that at Decibel Dungeon) will argue with me, but a great solution is to use a laptop smps. These are not extremely expensive, absolutely safe (better if you get a short-circuit proof one) and not as noisy as many people thing (use long wires and give them a turn to a ferrite toroid to get some extra noise immunity).
The 470uF-1000 uF caps on the board will care about the high frequencies.
I have done that and found absolutely no problem! No dangerous wiring at all!
(shouting): USE ONLY PSU'S THAT ARE NOT SAFETY EARTHED!!!!!!!!!!!!!!!! These are not a problem to find in spain.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.