Hi all,
I'd like to make a PA poweramp for a friend, and he's concerned with the budget. The toroid for a 2*200W amp would cost too much
I was thinking about an SMPS, is it possible for something like +/-60V 8A?
If someone can post (or link to) a schematic/article/tutorial, it would be nice
And, I've never worked with such stuff, are there any special precautions to take? Compared to linear supplies.
Thanks
Alex
I'd like to make a PA poweramp for a friend, and he's concerned with the budget. The toroid for a 2*200W amp would cost too much
I was thinking about an SMPS, is it possible for something like +/-60V 8A?
If someone can post (or link to) a schematic/article/tutorial, it would be nice
And, I've never worked with such stuff, are there any special precautions to take? Compared to linear supplies.
Thanks
Alex
Well, I can't offer any advice for line operated equipment (which any non-hybrid SMPS is until the transformer) but there's a real cool program that automatically computes things for you.
http://henry.fbe.fh-darmstadt.de/smps_e/
For 500W I'd recommend half or full bridge...
Tim
http://henry.fbe.fh-darmstadt.de/smps_e/
For 500W I'd recommend half or full bridge...
Tim
Don't go there! Every new electronic design has teething problems but large smps have large fangs. When problems occur, semiconductor devices will explode and eject hot fragments at high velocity. I've watched a team of engineers develop some large smps for power amps, they are complex things with high frequency, high voltage, rfi, and many evil noise problems which can only be solved by witchcraft.
I suggest to search in surplus equipment sellers for a used smps with suitable output and a nice safe housing. Also check mains and the air for conducted noise and rfi, under load, so that you do not create a broadcasting station in the amp
I suggest to search in surplus equipment sellers for a used smps with suitable output and a nice safe housing. Also check mains and the air for conducted noise and rfi, under load, so that you do not create a broadcasting station in the amp
You print this, throw it in the air and spot it floating down.
http://www.acronymfinder.com/af-query.asp?String=exact&Acronym=rif
But seriously, RFI can be detected with a probe... if you have a 50uA DC meter and a germanium diode you can scope things out.
Tim
http://www.acronymfinder.com/af-query.asp?String=exact&Acronym=rif
But seriously, RFI can be detected with a probe... if you have a 50uA DC meter and a germanium diode you can scope things out.
Tim
This is turning out to giving dangerous advices.I have looked a little into SMPS and I can say that this is nothing _I_ would do as a pro with limited knowledge because it's dangerous and also difficult. With this argumentation I think it's totally unsuitable for a novice. A 500-600 W mains operated SMPS is not for DIY!
If you want to start somewhere, start at < 48 V DC as power source.
Higher power SMPS
I agree with you, but think that alot need to be known about higher power smps.
What topology should be employed, especially for audio powered say 400watt audio amp.
One thing about the thread now is to discourage a newbie. That is not the intentions of the mentors of this forum.
Pls, if anybody fell that he or she cant offer advice, pls let go.
Also I will like to know where off-line smps design can be learn.
soundbag
I agree with you, but think that alot need to be known about higher power smps.
What topology should be employed, especially for audio powered say 400watt audio amp.
One thing about the thread now is to discourage a newbie. That is not the intentions of the mentors of this forum.
Pls, if anybody fell that he or she cant offer advice, pls let go.
Also I will like to know where off-line smps design can be learn.
soundbag
If you want to build a switching power supply, pick up a copy of the Abraham Pressman "Switching Power Supply Design" book. It's a big/thick/expensive book, but it's basically the "bible" of SMPS design, and it covers most theoretical things you'll need to know.
I design switching power supplies as part of my day job. I've been heavily mentored by several people but despite that, I still don't know what I'm doing half the time There's so much work involved; selecting/derating the proper components, heatsinking design, magnetics design, and designing a control loop that keeps the supply stable over almost all load conditions... it's not all that fun.
My suggestion is, if you're building an offline supply then go with the tried/tested/true/simple/reliable toroidal transformer for now - weight aside, you can't beat them.
I'd highly recommend starting with a car audio amp SMPS design. These are a lot safer to work with, and there's lots of designs available for them that you can copy. These will teach you most things you'll need to know about SMPS design; eventually you can apply this to an offline design.
I design switching power supplies as part of my day job. I've been heavily mentored by several people but despite that, I still don't know what I'm doing half the time
There's so much work involved; selecting/derating the proper components, heatsinking design, magnetics design, and designing a control loop that keeps the supply stable over almost all load conditions... it's not all that fun.My suggestion is, if you're building an offline supply then go with the tried/tested/true/simple/reliable toroidal transformer for now - weight aside, you can't beat them.
I'd highly recommend starting with a car audio amp SMPS design. These are a lot safer to work with, and there's lots of designs available for them that you can copy. These will teach you most things you'll need to know about SMPS design; eventually you can apply this to an offline design.
gmarsh said:
I agree whole-heartedly -- Pressman's book is sometimes sold at a discount on Amazon or Yahoo.
If you want a place to start (you've got to start somewhere) get Velleman's PWM motor-controller -- this SG3525 based-controller has enough flexibility that you can build simple buck, boost, inverting controllers, look at the waveforms and noise -- use the controller as a gate-controller.
An offline switcher, ie, one which draws current right from the mains, is not the place to start. On the other hand, seeing how a low power boost or buck converter actually works is worth the time and effort.
If you get the Velleman -- have an a.m. radio on while you play with it -- you'll see some of the real world problems.
high power SMPS
Thank you,
My problem is that I cant afford the Pressman book for now.
I am not interested in battery powered smps but the
calculations for the off line imput inductor from the 230v mains on my workbench for rod elliots project 89.
I am not new to electronics but to smps. I have built a couple of audio amps of moderate watts for djs. I always work from the mains always. Also I have started the project already but really
like to work from the mains.
Pls, kindly let me have the imput inductor calculations just before
the switchers.
Thank you.
SOUNDBAG
-------------------------
"measure everything,
dont trust anyman"
Thank you,
My problem is that I cant afford the Pressman book for now.
I am not interested in battery powered smps but the
calculations for the off line imput inductor from the 230v mains on my workbench for rod elliots project 89.
I am not new to electronics but to smps. I have built a couple of audio amps of moderate watts for djs. I always work from the mains always. Also I have started the project already but really
like to work from the mains.
Pls, kindly let me have the imput inductor calculations just before
the switchers.
Thank you.
SOUNDBAG
-------------------------
"measure everything,
dont trust anyman"
Re: high power SMPS
And additionally, once you finish this supply by ordering all the parts in low quantites, making a PCB or three, blowing a bunch of FETs during development, etc... I can assure you that the cost of the supply will be quite higher than that of the Pressman book.
But if you insist on a switching power supply, here's how I'd approach it:
- "car amplifier" topology; a half bridge (push-pull) configuration, using a pair of fast MOSFETs or IGBTs as the switches.
- switching frequency of 40-50KHz (IGBTs) or 80-100KHz (MOSFETs)
- ETD49 core, 3F3 ungapped, with two primary side windings of equal turns. # of turns depends on core used, switching frequency and peak load.
- Center tapped output winding, rectified by a bunch of schottkys.
- another ETD49 as the output inductor. You could wind both "inductors" on one core, or use two inductors. This will be a gapped 3F3 core, and I *strongly* recommend getting the Pressman book because it has an excellent section on magnetics design.
Now, how to control it... I'd use a UC3846 current-mode controller sitting on the primary side of the supply.
- UC3846 is "kickstarted" by a high value/high voltage resistor, which provides a weak startup supply for the chip. When it starts up and undervoltage shutdown kicks out, an auxiliary winding off the main transformer continues to power the chip.
- UC3846 drives FETs/IGBTs directly.
- UC3846 senses primary side current using an off-the-shelf 1:200 current sense transformer.
- Optocoupler used to sense secondary side voltage.
Use an isolation transformer when you build and test this supply, otherwise you won't be able to probe anything in the control circuitry with a scope.
I'm still *strongly* recommending the use of a conventional toroid power supply for this project; it's simpler, it's cheaper and there's many fewer ways to break it.soundbag said:
And additionally, once you finish this supply by ordering all the parts in low quantites, making a PCB or three, blowing a bunch of FETs during development, etc... I can assure you that the cost of the supply will be quite higher than that of the Pressman book.
But if you insist on a switching power supply, here's how I'd approach it:
- "car amplifier" topology; a half bridge (push-pull) configuration, using a pair of fast MOSFETs or IGBTs as the switches.
- switching frequency of 40-50KHz (IGBTs) or 80-100KHz (MOSFETs)
- ETD49 core, 3F3 ungapped, with two primary side windings of equal turns. # of turns depends on core used, switching frequency and peak load.
- Center tapped output winding, rectified by a bunch of schottkys.
- another ETD49 as the output inductor. You could wind both "inductors" on one core, or use two inductors. This will be a gapped 3F3 core, and I *strongly* recommend getting the Pressman book because it has an excellent section on magnetics design.
Now, how to control it... I'd use a UC3846 current-mode controller sitting on the primary side of the supply.
- UC3846 is "kickstarted" by a high value/high voltage resistor, which provides a weak startup supply for the chip. When it starts up and undervoltage shutdown kicks out, an auxiliary winding off the main transformer continues to power the chip.
- UC3846 drives FETs/IGBTs directly.
- UC3846 senses primary side current using an off-the-shelf 1:200 current sense transformer.
- Optocoupler used to sense secondary side voltage.
Use an isolation transformer when you build and test this supply, otherwise you won't be able to probe anything in the control circuitry with a scope.
Re: Re: high power SMPS
Five years of engineering school and that's all the advice you can give him?
There are, of course, other books ... but the correspondent would be well advised to read up on the apnotes which manufacturers like National Semi, TI, Maxim, Infineon, SGS et. al. provide -- and these are free.
It goes without saying that the precise and exact answer which is sought is given by one of the aforementioned companies on their website.
gmarsh said:
Five years of engineering school and that's all the advice you can give him?
There are, of course, other books ... but the correspondent would be well advised to read up on the apnotes which manufacturers like National Semi, TI, Maxim, Infineon, SGS et. al. provide -- and these are free.
It goes without saying that the precise and exact answer which is sought is given by one of the aforementioned companies on their website.
SMPS Design for Audio
Hi everyone,
Sorry I was gone for sooooo long. OK, If you're so bent on a 400W Supply, try taking apart any one of a number of Computer AT (ATX preferred) power supplies. Preferrably above 400W.
Don't use the supply itself, but use it as a guide on how to model your version. I agree that 40-50 kHz is a good starting place, but I would just stick with the MOSFETs, as they are easier to use, and easier to drive.
The audio manufacturer Soundstream actually made an SMPS-powered audio amp a few years ago: the DA-2 was rated at 200W x 2 RMS into 8 Ohms, and 400W x2 RMS into 4 Ohms.
I actually downloaded some pics of one opened up that was for sale on eBay last year. If I can find them, I will post them.
As for the books, Pressman's book is a good one, but also check out George Chryssis' book "High Frequency Switching Power Supplies". Latest edition was 1989, but it's still very relevant.
The other good book to get (if it's still available) is Marty Brown's book (I can't remember the title), part of the EDN series. All of the semi's in that book were Motorolas, but since ONsemi (spun off from Motorola) no longer makes MOSFETs above 250V, you'll have to substitute IR MOSFETS for the Motorola 500V units.
I also agree that the UC3846 is a good chip (I use the UC1846, mostly because I like Mil-Spec components ), but check out ONsemi's MC33025 (also a Mil-Spec Chip). I have used this chip before with great results. It is a PWM chip that can be used either voltage-mode or current-mode. With a start-up current is only 500uA, it works well in either DC-DC or off-line applications. Its oscillator is good up to 2MHz, so you can get some pretty fast designs going. 
Anyway, these three books, Pressman, Chryssis and Brown, make up THE SMPS bible, and will give you over 90% of what you will need to get going. Ask anyone who designs SMPSs either as a hobby, or for a living, and they will agree. Don't forget the manufacturers' datasheets and app notes. As jackinnj stated, these are VERY useful, and free, since you can download most of them from the respective websites.
best of luck, and keep us posted.
'73,
Steve
Hi everyone,
Sorry I was gone for sooooo long. OK, If you're so bent on a 400W Supply, try taking apart any one of a number of Computer AT (ATX preferred) power supplies. Preferrably above 400W.
Don't use the supply itself, but use it as a guide on how to model your version. I agree that 40-50 kHz is a good starting place, but I would just stick with the MOSFETs, as they are easier to use, and easier to drive.
The audio manufacturer Soundstream actually made an SMPS-powered audio amp a few years ago: the DA-2 was rated at 200W x 2 RMS into 8 Ohms, and 400W x2 RMS into 4 Ohms.
I actually downloaded some pics of one opened up that was for sale on eBay last year. If I can find them, I will post them.
As for the books, Pressman's book is a good one, but also check out George Chryssis' book "High Frequency Switching Power Supplies". Latest edition was 1989, but it's still very relevant.
The other good book to get (if it's still available) is Marty Brown's book (I can't remember the title), part of the EDN series. All of the semi's in that book were Motorolas, but since ONsemi (spun off from Motorola) no longer makes MOSFETs above 250V, you'll have to substitute IR MOSFETS for the Motorola 500V units.
I also agree that the UC3846 is a good chip (I use the UC1846, mostly because I like Mil-Spec components
), but check out ONsemi's MC33025 (also a Mil-Spec Chip). I have used this chip before with great results. It is a PWM chip that can be used either voltage-mode or current-mode. With a start-up current is only 500uA, it works well in either DC-DC or off-line applications. Its oscillator is good up to 2MHz, so you can get some pretty fast designs going. Anyway, these three books, Pressman, Chryssis and Brown, make up THE SMPS bible, and will give you over 90% of what you will need to get going. Ask anyone who designs SMPSs either as a hobby, or for a living, and they will agree. Don't forget the manufacturers' datasheets and app notes. As jackinnj stated, these are VERY useful, and free, since you can download most of them from the respective websites.
best of luck, and keep us posted.
'73,
Steve
Re: Re: Re: high power SMPS
My "advice" hinted at the overall construction of the supply and component selection involved, to give soundbag an overall idea of what stuff has to be designed for these supplies. There's a lot more to calculate on a power supply than an "input inductor", and most offline SMPSes that I know of don't use input inductors for anything other than PFC.
Good recommendations though (and thanks too, N-Channel).
Five years of engineering school gave me one course that hinted at buck regulators and didn't do any practical design work. It is engineering school after alljackinnj said:Five years of engineering school and that's all the advice you can give him?
There are, of course, other books ... but the correspondent would be well advised to read up on the apnotes which manufacturers like National Semi, TI, Maxim, Infineon, SGS et. al. provide -- and these are free.
It goes without saying that the precise and exact answer which is sought is given by one of the aforementioned companies on their website.
My "advice" hinted at the overall construction of the supply and component selection involved, to give soundbag an overall idea of what stuff has to be designed for these supplies. There's a lot more to calculate on a power supply than an "input inductor", and most offline SMPSes that I know of don't use input inductors for anything other than PFC.
Good recommendations though (and thanks too, N-Channel).
Re: SMPS Design for Audio
Linn uses SMPS --
http://www.linn.co.uk/spec_sound/spec_sound.cfm
http://www.linn.co.uk/spec_sound/pr...NavBar=playback&CFID=1575013&CFTOKEN=70640412
of course, they have had to recall the LK240 , AV5105 and AV5150
http://www.cpsc.gov/cpscpub/prerel/prhtml04/04162.html
N-Channel said:
Linn uses SMPS --
http://www.linn.co.uk/spec_sound/spec_sound.cfm
http://www.linn.co.uk/spec_sound/pr...NavBar=playback&CFID=1575013&CFTOKEN=70640412
of course, they have had to recall the LK240 , AV5105 and AV5150
http://www.cpsc.gov/cpscpub/prerel/prhtml04/04162.html
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.