Richard C: Right On!
I just made an experiment (see below) to see how the MOSFET's should be temp. compensated. You need around 2.5 V on the VBE multiplier, and some 5.5 V of temp. stable Zener voltage on top of that. I used a varialbe Zener diode to find these values, why not use steady resistors in the VBE multiplier, and a variable Zener diode in the real amp?
The Variable Zener Diode is: TL431A from TI. (cheap).
You don't need the 'Emitter' resistors! Just short them together as i did.
I just made an experiment (see below) to see how the MOSFET's should be temp. compensated. You need around 2.5 V on the VBE multiplier, and some 5.5 V of temp. stable Zener voltage on top of that. I used a varialbe Zener diode to find these values, why not use steady resistors in the VBE multiplier, and a variable Zener diode in the real amp?
The Variable Zener Diode is: TL431A from TI. (cheap).
You don't need the 'Emitter' resistors! Just short them together as i did.
Attachments
Something for guy here maybe?
http://www.diyaudio.com/forums/showthread.php?s=&threadid=6700&highlight=
Lars, I gather that you are going to use this transformer for some project of yours. Can you tell us something about it?
BTW: Did this "not ultra low cost transformer" have something to do with this thread?
http://www.diyaudio.com/forums/showthread.php?s=&threadid=6700&highlight=
Lars, I gather that you are going to use this transformer for some project of yours. Can you tell us something about it?
BTW: Did this "not ultra low cost transformer" have something to do with this thread?
Sure PerAnders, we discussed the use of Switched Mode PSU's on page 5.
It is not so expensive. This one is factory wound for the UL listing, but since it only contains about 25 windings in all, any DIY'er can easily make it at home. The ferrite cost the formidable sum of 1$, and 2m of electrical wire cost .. nothing. (almost). So a 1kW transformer for around 1$ .. not bad i think. Of course you need a bit of support circuit, but it can also be put together cheaply.
It is not so expensive. This one is factory wound for the UL listing, but since it only contains about 25 windings in all, any DIY'er can easily make it at home. The ferrite cost the formidable sum of 1$, and 2m of electrical wire cost .. nothing. (almost). So a 1kW transformer for around 1$ .. not bad i think. Of course you need a bit of support circuit, but it can also be put together cheaply.
.... if he/she has accurate instructions... very important if the switching speed is high.Lars Clausen said:It is not so expensive. This one is factory wound for the UL listing, but since it only contains about 25 windings in all, any DIY'er can easily make it at home.
The problem with SMPS is the rather high cost for the support parts. A 1 kW PS is not very cheap but light...
Peranders: Here are the schematics of a design i made a couple of years back. It actually played (and well too) for several months as a bird-nest circuit, before we took it out of the demo room, so no one would get hurt. .
A better image can be seen here: http://www.lcaudio.dk/smps1dia.jpg
Of course it can be tweaked with a xtal oscillator, and maybe a real Gatedriver IC, but in think the performance of this low cost circuit is quite good. And the price is low on all parts.
WARNING! Build at your own risk, HIGH VOLTAGES !!!
A better image can be seen here: http://www.lcaudio.dk/smps1dia.jpg
Of course it can be tweaked with a xtal oscillator, and maybe a real Gatedriver IC, but in think the performance of this low cost circuit is quite good. And the price is low on all parts.
WARNING! Build at your own risk, HIGH VOLTAGES !!!
Attachments
I was more into a design which met CE approval... Isn't power factor correction necessary here and much more filtering?
Here is something to put your teeth in.
Check the ready design. I have this PS at work. Very nice working PS which I got for free!
http://www.linear.com/pdf/lt1509.pdf
Here is something to put your teeth in.
Check the ready design. I have this PS at work. Very nice working PS which I got for free!
http://www.linear.com/pdf/lt1509.pdf
Attachments
You have to read a 500 page book to get all the details on the CE approval.
Strictly you actually need the PFC, but very few manufacturors follow this rule, so i think you can get approval even without PFC. Not quite sure though.
And next CE approval is not enough, you also need UL approval and 3 other approvals, to cover the entire world market. Each cost a load of money to obtain, why we never released a Switched Mode PSU. And should you change anything, you have to start back at square one. We are talking at least 100.000$ in approval costs for one design.
Strictly you actually need the PFC, but very few manufacturors follow this rule, so i think you can get approval even without PFC. Not quite sure though.
And next CE approval is not enough, you also need UL approval and 3 other approvals, to cover the entire world market. Each cost a load of money to obtain, why we never released a Switched Mode PSU. And should you change anything, you have to start back at square one. We are talking at least 100.000$ in approval costs for one design.
I think both you and I can make the conclusion that it's much work to design a very good SMPSU
But Michael Bladelius, Sweden has designed a monster amp with 7 kW SMPSU, Grendel called.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=12903&highlight=
But Michael Bladelius, Sweden has designed a monster amp with 7 kW SMPSU, Grendel called.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=12903&highlight=
If the harmonics are to high you will not get it through a CE certification no matter what. I think if it is a class a amp pulling 300W. Without PFC and inputfilter you will never be able to meet there criteria.
On the other hand if it is a class AB design pulling like 10 - 60W i actually think it will walk right through the test. even if the SMPS is a 500W peak design.
So with a class H design, i would believe the design is ok. But only if the tests are as loose as they some times is.
Like only driving a the amp into 8Ohm at a low power level which is identical to an average listning level.
But on the other hand.. It is not cheap, if the test fails. It like throwing 20000DKR out the window!
On the other hand if it is a class AB design pulling like 10 - 60W i actually think it will walk right through the test. even if the SMPS is a 500W peak design.
So with a class H design, i would believe the design is ok. But only if the tests are as loose as they some times is.
Like only driving a the amp into 8Ohm at a low power level which is identical to an average listning level.
But on the other hand.. It is not cheap, if the test fails. It like throwing 20000DKR out the window!
I edited this thread and moved out of context parts to Texas http://www.diyaudio.com/forums/showthread.php?s=&threadid=23634
Please keep discussion on topic.
Please keep discussion on topic.
The thread is opened again and I’d like to salute this step of moderators. The name of the new born Texas thread is however very maliciously named. Its name certainly doesn’t have anything with its content. But since these are my first posts that finished in Texas (I am posting here about year and a half), I am honored and I’ll not make a question about this. I just don’t like Peter’s last sentence (threat) about the “bashers” in the last sin binning notification. Seems like this is definitely advertising free space in which can be dangerous to counteract to the advertising.
Richard C, Re: your statement in the WCTC thread: I’ll not post anything more in this thread, so you are safe about this. Among the other reasons I’ll not post because the suggested solution doesn’t have much with the way to make a cheap amplifier. Tip for the ultra low cost amp: deal with the supply, it is ridiculous to try to make something better than LM3875 for $6 (or less). Btw, you’ll not use it anyway in the amp, but to lower the total cost of the system significantly, avoid spending EUR200+ for the clock.
Good luck making cheap amplifiers,
Pedja
Richard C, Re: your statement in the WCTC thread: I’ll not post anything more in this thread, so you are safe about this. Among the other reasons I’ll not post because the suggested solution doesn’t have much with the way to make a cheap amplifier. Tip for the ultra low cost amp: deal with the supply, it is ridiculous to try to make something better than LM3875 for $6 (or less). Btw, you’ll not use it anyway in the amp, but to lower the total cost of the system significantly, avoid spending EUR200+ for the clock.
Good luck making cheap amplifiers,
Pedja
No idea for the good name of it, it was just a comment about the intentions included in the exsisting name... it is in Texas anyway.Peter Daniel said:I also don't like the title of a Texas thread, so if you have any suggestions, I'll gladly change it (to anything you want, it's in Texas anyway)
Pedja
How about this http://www.diyaudio.com/forums/showthread.php?s=&threadid=23634
The best titles are usually the lines taken directly from members' posts
The best titles are usually the lines taken directly from members' posts
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Solid State
- Ultra Low Cost Amp Design