To Carel.
is not a good idea to disconnect "DTC" (connect to GND).
this pin at startup force at +5 V for soft start.
absolutely not good to change the percentage of PWM than 50% permanently.
transformer, RC and other components are not suitable for this operation.
at 28v (99%), including the PWM (494) is supercharged (is powered by a link on the transformer).
DTC without working, if something is wrong may explode.
is not a good idea to disconnect "DTC" (connect to GND).
this pin at startup force at +5 V for soft start.
absolutely not good to change the percentage of PWM than 50% permanently.
transformer, RC and other components are not suitable for this operation.
at 28v (99%), including the PWM (494) is supercharged (is powered by a link on the transformer).
DTC without working, if something is wrong may explode.
€ 30 will definitly provide you with an el-cheapo power supply... BTW, ATX PSUs are not all from china, the ATX standards are designed by Integrated Electronics Corporation - Intel {the same company who make the best PC chipsets and processors in the world}. They are from the USA. If you buy a linear solution for € 30, then you'll get what you've paid for. ATX units doesn't have a dirty operation and most certainly don't emit EMI {electro magnetic interference}. Yeah - no way, a linear transformer emit way more EMI. The whole idea of SMPS is to drive the transformer at very, very high frequencies {to reduce noise}, and to give out a more stable power output. They WON'T use PSUs in PC that emit EMI, it will definitly interfere with the components inside the PC, so this is not true :-(
Whatever the choice, we're lving in a democratic world {unless you're in China} - we can choose whatever we like. If I like SMPS, then I'm going to use it. If you like linear solutions, then use it.
Whatever the choice, we're lving in a democratic world {unless you're in China} - we can choose whatever we like. If I like SMPS, then I'm going to use it. If you like linear solutions, then use it.
Well, yes.... depending on the "AC load", if you use DC you'll get an electro-magnet, so you need AC if you want to measure imp, voltage drop etc. And leave the secondary open, initially anyway, if you're real sure 'bout what you're doing you can use a dummy-load, made of power-resistors, but again, depending on your goal.... You need to use resistors that isn't a "coil, cap and ohm"-thing, in other words, expensive once.....
But with the right IC and FETs and of cause, your trafo, that will supply the desired voltage and current, it is cheaper to build a 1 000W power supply. If you were to use linear design, your trafo will weight about 10kg, now you have the power directly from the mains supply, so 230V~@16A(in Sweden), and you can use a sugar-trafo with primary of 230VAC and ?VAC secondary, or you just "sample" the sinus wave from the mains-jacket, if you want +/- 30Volt you only "use" +30VAC and -30VAC. Optocoupler, some ferrite rod, 1 -2 mm diameter, shellacked copper wire....
Well I stop there, cuz it will get more and more complexed, but the main reason for using SMPS is that you can sample the mains supply, w/o bulky transformers, and you have a couple of kilo watt to play with. But don't start to build a SMPS that use the mains, use much lower voltage and be sure that you're separated from the mains, with a 1:1 safety trafo, or some other solution.. and when you're absolutely sure 'bout what you're doing, use the power in the "wall"...
Last edited:
Ar4: Sounds like a plan, and I'll help as much as can, if you "feed" me data(info)
There are millions different SMPS IC, I use an very old one called TL497 from Texas Instruments, but in small apps, and good for "beginners".. cuz they are sturdy and very easy to build with. And if you search the net you'll find many applications where TL497 is an excellent replacement IC. You can run it in negative mode as well, but it's an old circuit and there are much better once out there today. I happen to inherit a supply with over a thousand of this circuit, so.... (here's the circuit's datasheet url: TL497 pdf, TL497 description, TL497 datasheets, TL497 view ::: ALLDATASHEET :::) It can't get any simpler... and I've have had lots of fun with it, and I'm still using them(since it's a obsolete circuit, I haven't manage to sell them...), so "go on my son"
There are millions different SMPS IC, I use an very old one called TL497 from Texas Instruments, but in small apps, and good for "beginners".. cuz they are sturdy and very easy to build with. And if you search the net you'll find many applications where TL497 is an excellent replacement IC. You can run it in negative mode as well, but it's an old circuit and there are much better once out there today. I happen to inherit a supply with over a thousand of this circuit, so.... (here's the circuit's datasheet url: TL497 pdf, TL497 description, TL497 datasheets, TL497 view ::: ALLDATASHEET :::) It can't get any simpler... and I've have had lots of fun with it, and I'm still using them(since it's a obsolete circuit, I haven't manage to sell them...), so "go on my son"
Last edited:
joaquim, thank you for suggestions!
I was playing with 555 as I was not very sure about myself I used 12V DC input, generated 27KHz frequency, used transistor as driver for another high power transistor and simply rewound trafo. So in result I got 2x31V AC. Circuit was a way to simple, but as I said, I was playing with it. So now I will upgrade it for a bit and then I will use some other, stable IC, probably the same TL497 as you suggested. But only for non-mains operations. For mains version I still will try to find ATX for modification. I think that I will add two LM338s for regulated output.
Btw most of music that is used to "move" my speakers are Swedish DM
I was playing with 555
as I was not very sure about myself I used 12V DC input, generated 27KHz frequency, used transistor as driver for another high power transistor and simply rewound trafo. So in result I got 2x31V AC. Circuit was a way to simple, but as I said, I was playing with it. So now I will upgrade it for a bit and then I will use some other, stable IC, probably the same TL497 as you suggested. But only for non-mains operations. For mains version I still will try to find ATX for modification. I think that I will add two LM338s for regulated output.Btw most of music that is used to "move" my speakers are Swedish DM
Cool, then you've probably heard 'bout my brothers band, Bring Your Own Knife(B.Y.O.K). Maybe not my flavour, but I'm a hugs "In Flames" fan, and B.Y.O.K. where compared with early In Flames, I think one critic wrote, "these guys will take over the torch, from In Flames, whom has left the DM scene long ago(I dislike their early records, but from "Soundtrack to your escape" and onwards, are really great, "Sense of Purpose" have they(IMHO!) found the right sound, though "Trigger", "Pin Ball Map" are great songs too..), 'The End' is the name of their(my bro's band) new Album, check them out at Bring Your Own Knife [LOOKING FOR SHOWS] on MySpace Music - Free Streaming MP3s, Pictures & Music Downloads! "Jake"'s my brother, his real name is Jacob, don't tell any one... haha... ). Back on track...
Starting with a 555 timer, is great learning experience. And if you gonna go ahead with your "modified ATX" as smps you basically only need to change the winding ratio, or change trafo, I'd use +5V, and rewind, or change the trafo, so it fit your requirements. If you come across a circuit diagram, your home free! Red is +5V. And bare in mind that you reduce the current as you increase the voltage... +5V is used on the mainbord as well as to HDD. Floppies...., so if your using an 20-pin ATX supply, you have higher Watt at +5V, than you have @+12V. If going to buy a new supply, you have +12V and at least 30A, since the CPU core has it own supply now a days(on a 24-pins ATX12V).
I have a quad core, AMD Phenom II x4, and my supply, according to the spec's, can deliver 42A(+12V, 2xyellow, 2xblack) on the "CPU" connector, so you have around 400W, if you'd use my supply...
One last thing, when you deal with digital circuits, the rippels are not critical, if you have 1=2.55 - 5.0V, and 0=0 - 2.5V - .5V for hysteris - well it speek for it's self, you can have 0.5V noise on the supply, the only really clean voltage if your building a MCU is at AVcc, so you need to filter the +/-35V with coils and decouple capacitors. And you need to wind the coils your self, so you get the right "Z"(imp.) and milli/micro-Henry, use a oscilloscope and sweep fq from 1hz to 100kHz, and use your 'scopes highest resolution... I can post a simple variable load/constant current source, 20A@100VDC or 1M -0 ohm... useful when building all kinds of supply and as dummy-load for audioamp's....
Keep on rockin'
). Back on track...Starting with a 555 timer, is great learning experience. And if you gonna go ahead with your "modified ATX" as smps you basically only need to change the winding ratio, or change trafo, I'd use +5V, and rewind, or change the trafo, so it fit your requirements. If you come across a circuit diagram, your home free! Red is +5V. And bare in mind that you reduce the current as you increase the voltage... +5V is used on the mainbord as well as to HDD. Floppies...., so if your using an 20-pin ATX supply, you have higher Watt at +5V, than you have @+12V. If going to buy a new supply, you have +12V and at least 30A, since the CPU core has it own supply now a days(on a 24-pins ATX12V).
I have a quad core, AMD Phenom II x4, and my supply, according to the spec's, can deliver 42A(+12V, 2xyellow, 2xblack) on the "CPU" connector, so you have around 400W, if you'd use my supply...
One last thing, when you deal with digital circuits, the rippels are not critical, if you have 1=2.55 - 5.0V, and 0=0 - 2.5V - .5V for hysteris - well it speek for it's self, you can have 0.5V noise on the supply, the only really clean voltage if your building a MCU is at AVcc, so you need to filter the +/-35V with coils and decouple capacitors. And you need to wind the coils your self, so you get the right "Z"(imp.) and milli/micro-Henry, use a oscilloscope and sweep fq from 1hz to 100kHz, and use your 'scopes highest resolution... I can post a simple variable load/constant current source, 20A@100VDC or 1M -0 ohm... useful when building all kinds of supply and as dummy-load for audioamp's....
Keep on rockin'
In Flames is pretty cool. Dial 595 Escape, The Quiet Place and others \m/__(>.<)__\m/
I just checked B.Y.O.K, thanks for suggestion I like them.
Yeah, ATX modification with proper IC is simple. But when there is a lot of protections, specially UV protection it gets harder.
I had very fast modification, with new rectifier board and caps. Rewound trafo, added new anode ultrafast diode, used old cathode diode, 1880mkf caps per rail. It worked. But the trafo made some silent sound, so I decided to rewound it. So I did, but when powered on supply, there was a bang. Exploded primary side resistors and BJTs. I desoldered 560mkf caps, some LOW ESR [and LOW voltage too] caps + some other good stuff, for example radiators.
So now just to find ATX! I ordered ETD-44 core for some other experiments and a couple of TL497 and SG3525. And some NE555 too for some quick solutions.
Thanks for useful info joaquim. \m/
I just checked B.Y.O.K, thanks for suggestion
I like them. Yeah, ATX modification with proper IC is simple. But when there is a lot of protections, specially UV protection it gets harder.
I had very fast modification, with new rectifier board and caps. Rewound trafo, added new anode ultrafast diode, used old cathode diode, 1880mkf caps per rail. It worked. But the trafo made some silent sound, so I decided to rewound it. So I did, but when powered on supply, there was a bang. Exploded primary side resistors and BJTs. I desoldered 560mkf caps, some LOW ESR [and LOW voltage too] caps + some other good stuff, for example radiators.
So now just to find ATX! I ordered ETD-44 core for some other experiments and a couple of TL497 and SG3525. And some NE555 too
for some quick solutions.Thanks for useful info joaquim. \m/
Last edited:
You seem to be on solid ground, you've picked the right IC, and you have the right "attitude"... SG3525 datasheet from Motorola has some nice application notes, hint and diagrams. SG3525 pdf, SG3525 description, SG3525 datasheets, SG3525 view ::: ALLDATASHEET ::: At page 6 and 7 you have some very simple hookups for push-pull with FET's or BJT, or you can drive the transformer directly with the IC... I've forgotten this circuit completely, this one(SG3525) is extremely simple to build with, and you'll get great results. I believe that you're talking about Schottky diodes, "anode diode", like; https://www1.elfa.se/data1/wwwroot/assets/datasheets/vg601808_e.pdf. One thing to bare in mind when you wind your own trafo is to not make(as I did at first...), the primary's impedance too low. Then you probably blow your transistors(if you add some 'power' to them).
You can get almost all the info you need from the datasheet, your PWM/SM-IC and the FET you chose. The datasheet for the SG3525 have some proposal FET's, and then you can look them up(the FET's) and get some useful info, like Vgs, Rds, IRF9540/540 is a nice cheap Fet, with low Rds, 0.3/0.2 ohm and Turn-On delay time is 60ns(<-- this value tells you the max fq for the FET you'll use, and it's based on the transistors In/Output Capacitance, which is for this FET 700pF/450pF)... It's an old "workhorse", and you can probably find cheaper and faster FET's... But I have inherit a stock of this complementary FET's too, so why not use them...
God luck, with your project.. I've been inspired by all this, so I'm going to(try anyway ) design a high power SMPS test-DC source, and I got in my head that it would be nice if you could use it in reverse, as a load.. I'm gonna start a new thread about this project, I have the idea in my head, but I need to get it on paper, and "in to" OrCad for some theoretic testing.... A DC Source/Constant Current Source/Variable "power resistor" al á Joaqim, it can be done, and it will . We are not blessed with high temp and lots of sun shine in Sweden, so it can be a "rainy day" project, those we have lots of, this time of the year...
You can get almost all the info you need from the datasheet, your PWM/SM-IC and the FET you chose. The datasheet for the SG3525 have some proposal FET's, and then you can look them up(the FET's) and get some useful info, like Vgs, Rds, IRF9540/540 is a nice cheap Fet, with low Rds, 0.3/0.2 ohm and Turn-On delay time is 60ns(<-- this value tells you the max fq for the FET you'll use, and it's based on the transistors In/Output Capacitance, which is for this FET 700pF/450pF)... It's an old "workhorse", and you can probably find cheaper and faster FET's... But I have inherit a stock of this complementary FET's too, so why not use them...
God luck, with your project.. I've been inspired by all this, so I'm going to(try anyway
) design a high power SMPS test-DC source, and I got in my head that it would be nice if you could use it in reverse, as a load.. I'm gonna start a new thread about this project, I have the idea in my head, but I need to get it on paper, and "in to" OrCad for some theoretic testing.... A DC Source/Constant Current Source/Variable "power resistor" al á Joaqim, it can be done, and it will . We are not blessed with high temp and lots of sun shine in Sweden, so it can be a "rainy day" project, those we have lots of, this time of the year...
Last edited:
Before I forget... I mentioned "Rds" and "On-Time Delay", this is importen, be course you want to switch your SMPS as fast as possible. Around 250k - 500kHz, why? Higher fq = smaller core = less winding... Say that the core saturate @ 10A, 100kHz, increase the switching fq, you can pull more current from the same transformer core. And one last thing, the impedance the transistors "feels" increase with increasing fq. So these are the main things;
Higher frequency = smaller core@the same current
core size/type(form-pressed ferrite powder)
Rds, Ids, I/O capacitance(On-Time delay), for FET's
High frustration tolerance, and lots of fun... and try and err, go ahead with crazy ideas, either it works or it want... it can be expensive though ..haha.. Don't use mains, before your 200% sure about what you're doing, if possible don't use mains couple at all... so your idea with using an ATX psu is safer... but a warning here as well, you have a transformer that double the mains voltage, so around 450 - 500V~, and then they switch this voltage at about 500kHz, and transform it down to +/-5, +/-12, or whatever voltage(DC-DC as well)... so you have lethal voltage on the ATX-mainboard too, but that transformer is almost always marked "Danger High Voltage"..... that's it for now, the main thing is that you're having fun, and use common sense...
And a very well designed filter, so you "wash away" all noise, if you don't have an oscilloscope, set your DMM to AC 2V, and use power resistors or similar as load to your SMPS, and the formulas are, P=U*I, or P=U^2/R, (so P=35^2/R(i.e.500)) u'll know the power consumption and if u read 0.768VAC you get an idea how clean and linear your supply is, and the easiest way to suppress ripples is bigger cap's.... for linear technology... but it's not a great idea to just put 10 000µF at the output, if you happen to get a short circuit or the current in the capacitors, for some reason want to go in reverse, you will blow up your equipment...
Higher frequency = smaller core@the same current
core size/type(form-pressed ferrite powder)
Rds, Ids, I/O capacitance(On-Time delay), for FET's
High frustration tolerance
, and lots of fun... and try and err, go ahead with crazy ideas, either it works or it want... it can be expensive though ..haha.. Don't use mains, before your 200% sure about what you're doing, if possible don't use mains couple at all... so your idea with using an ATX psu is safer... but a warning here as well, you have a transformer that double the mains voltage, so around 450 - 500V~, and then they switch this voltage at about 500kHz, and transform it down to +/-5, +/-12, or whatever voltage(DC-DC as well)... so you have lethal voltage on the ATX-mainboard too, but that transformer is almost always marked "Danger High Voltage"..... that's it for now, the main thing is that you're having fun, and use common sense... And a very well designed filter, so you "wash away" all noise, if you don't have an oscilloscope, set your DMM to AC 2V, and use power resistors or similar as load to your SMPS, and the formulas are, P=U*I, or P=U^2/R, (so P=35^2/R(i.e.500)) u'll know the power consumption and if u read 0.768VAC you get an idea how clean and linear your supply is, and the easiest way to suppress ripples is bigger cap's.... for linear technology... but it's not a great idea to just put 10 000µF at the output, if you happen to get a short circuit or the current in the capacitors, for some reason want to go in reverse, you will blow up your equipment...
Last edited:
Well isn't it why we all do it
And as I wrote in an earlier reply;
You have the right IC, ideas, and will.. so you're on solid ground. IMHO I think SG3525 is a lovely little IC. So simple in its design, and so flexible, I went through my "Power supply IC" catalogue cabinet, and found four SG3525.. so those are going to be the "heart" of my SMPS variable DC-supply, w/ extra feature...
And as I wrote in an earlier reply;
You have the right IC, ideas, and will.. so you're on solid ground. IMHO I think SG3525 is a lovely little IC. So simple in its design, and so flexible, I went through my "Power supply IC" catalogue cabinet, and found four SG3525.. so those are going to be the "heart" of my SMPS variable DC-supply, w/ extra feature...
Off the shelf SMPS for $80
Just use two of these center tapped for bi-polar.
.
36V DC 9.7A 350W Regulated Switching Power Supply New - eBay (item 390213631038 end time Jul-26-10 11:20:05 PDT)
.
Don't fall into the anti hype by those that have never done a listening comparison. A modern switch mode power supply will sound much better than any big linear supply. More transparent and dynamic.
Thanks in advance.
I need a suitable smps for my LM3886(x2) based audio amplifier. I'm new to smps I can't find any smps circuit that suits. please anybody help..
Just use two of these center tapped for bi-polar.
.
36V DC 9.7A 350W Regulated Switching Power Supply New - eBay (item 390213631038 end time Jul-26-10 11:20:05 PDT)
.
Don't fall into the anti hype by those that have never done a listening comparison. A modern switch mode power supply will sound much better than any big linear supply. More transparent and dynamic.
well doesn't that defeat the purpose.. isn't this forum called, "Do It Yourself Audio"....
And SMPS are great for audio design, as linear designs are too. Why don't you use batteries... 90% of the sound quality, is depending on the speakers, and if you have an absolutely clean DC supply, SMPS, Linear or Batteries, you can't mean that you "can hear" the difference!! It's impossible, a bad SMPS, as it's most common used in digital applications, doesn't it matter if you have some high fq noise, so it's harder, more difficult, to build a good Switch Mode Power Supply than it's to build a non-regulated linear one(that's the challenge, and why I'm doing it, for the challenge)... just use bigger decouple capacitors... there's a simple formula for calculating the "size" of the decouple capacitors, to suppress ripples at a specified wattage...
With your logic, I should build a SMPS battery charger, and run my amp's from batteries. I've actually build a preamp, with NiCd batteries, for more then 20 years ago, and the purpose was to get zero noise from the supply, which was in it's own housing, and it was build so when one(of two) battery pack was empty, the other took over, and a neat little diode told me that I was running on Sec. and primary pak was being charged....
And truth be told, I think my super simple ECC82 line-amp sounds much better, even though it might not have the "transparency and dynamic" to it, as the battery(or SMPS)..... Jesus!
Very possible
It's obviously very possible to hear the differences between different power supplies as it is to hear the differences between diodes, caps, opamps, cables, whatever. Large Batteries do sound the best but are less convenient.
well doesn't that defeat the purpose.. isn't this forum called, "Do It Yourself Audio"....
And SMPS are great for audio design, as linear designs are too. Why don't you use batteries... 90% of the sound quality, is depending on the speakers, and if you have an absolutely clean DC supply, SMPS, Linear or Batteries, you can't mean that you "can hear" the difference!! It's impossible
It's obviously very possible to hear the differences between different power supplies as it is to hear the differences between diodes, caps, opamps, cables, whatever. Large Batteries do sound the best but are less convenient.
Well cables, diodes, caps, resistors etcetera are passive components, with a cap and a coil, you can make a 6dB HP-filter i.e.
You can never make me believe that you can hear any different, between different supplies to the same amp/speaker hookup, if the DC-sources are identical, say you use a 'scope and a spectrum analyser, if DC comes from a battery, linear or SMPS supply.
But there are people how believes that CO2 increase the temperature as well, even though it was a Swede that speculated in the mid 70's on a BBC program about climate, that carbon dioxide might have some(emphasize 'might' and 'have some') impact on the climate....
And honest to God, you have so many other factors that makes "real" impact on the sound, and yes you don't want "noise" of any kind, lurking its way into the amplifier.. And read a bit about different classes in audio, and D-amps are hardest to build, and T...(I mean, they are pre designed IC for god's sake, with MCU's and other stuff... I love to build and design audio amps, speakers, guitar effects and more, so I don't want to buy a circuit for $1000, and they are SMD mounted! Analogue audio-amp's till I die...)
But believe what you want, Sweden(and USA) are free speech countries...
Try to listen to a MP3-track, without the DAC..... All you put in is analogue and all that comes out are...11100101, or?
You can never make me believe that you can hear any different, between different supplies to the same amp/speaker hookup, if the DC-sources are identical, say you use a 'scope and a spectrum analyser, if DC comes from a battery, linear or SMPS supply.
But there are people how believes that CO2 increase the temperature as well, even though it was a Swede that speculated in the mid 70's on a BBC program about climate, that carbon dioxide might have some(emphasize 'might' and 'have some') impact on the climate....
And honest to God, you have so many other factors that makes "real" impact on the sound, and yes you don't want "noise" of any kind, lurking its way into the amplifier.. And read a bit about different classes in audio, and D-amps are hardest to build, and T...(I mean, they are pre designed IC for god's sake, with MCU's and other stuff... I love to build and design audio amps, speakers, guitar effects and more, so I don't want to buy a circuit for $1000, and they are SMD mounted! Analogue audio-amp's till I die...)
But believe what you want, Sweden(and USA) are free speech countries...
Try to listen to a MP3-track, without the DAC..... All you put in is analogue and all that comes out are...11100101, or?
Listening
You should try some listening tests on a revealing system. You will be surprised what makes a difference to the sound. Pretty much everything.
.
.
You should try some listening tests on a revealing system. You will be surprised what makes a difference to the sound. Pretty much everything.
.
.
Well cables, diodes, caps, resistors etcetera are passive components, with a cap and a coil, you can make a 6dB HP-filter i.e.
You can never make me believe that you can hear any different, between different supplies to the same amp/speaker hookup, if the DC-sources are identical, say you use a 'scope and a spectrum analyser, if DC comes from a battery, linear or SMPS supply.
But there are people how believes that CO2 increase the temperature as well, even though it was a Swede that speculated in the mid 70's on a BBC program about climate, that carbon dioxide might have some(emphasize 'might' and 'have some') impact on the climate....
And honest to God, you have so many other factors that makes "real" impact on the sound, and yes you don't want "noise" of any kind, lurking its way into the amplifier.. And read a bit about different classes in audio, and D-amps are hardest to build, and T...(I mean, they are pre designed IC for god's sake, with MCU's and other stuff... I love to build and design audio amps, speakers, guitar effects and more, so I don't want to buy a circuit for $1000, and they are SMD mounted! Analogue audio-amp's till I die...)
But believe what you want, Sweden(and USA) are free speech countries...
Try to listen to a MP3-track, without the DAC..... All you put in is analogue and all that comes out are...11100101, or?
Well yes, pretty much everything, and of course does a poor designed linear/SMPS... have an impact... but the choice of output transistors, and narrow tolerance MKP caps, slew rate in an op-amp, bandwidth, bla, bla, bla... I vote that why drop this discussion.... it becomes silly and pointless... just one last thing, switching -whatever, will radiate...
I've actually done every imaginable tests with all kinds of amp classes, power supplies, speakers.... It's my job, or part of it anyway, we have an development department, where I test new ideas in "real" life(not just pspice and 'bench hookups). And the demanding-costumer often wants $2200 tube-amp, SE with power-triode output tubs, push-pull vs. single ended have a very different sound, and there's reason why... As long as the DC-voltage is "the same", all values are exactly the same, no difference what so ever, one is linear, one smps and batteries, say lead acid. Can you mathematically explain why SMPS vs. Linear, will make the same amplifier(circuit board), sound different, better, worse...... direct current is flat, and if you produce this with SMPS, LINEAR, BATTERY, and it is flat direct current trough-out the Watt-range of the same amp, plz show me the math behind it....
I vote that why drop this particular discussion.... it becomes silly and pointless... just one last thing, switching -whatever, will radiate... and will interfere even if you're switch@1MHz... so shield every thing if you want to use the same housing...
I vote that why drop this particular discussion.... it becomes silly and pointless... just one last thing, switching -whatever, will radiate... and will interfere even if you're switch@1MHz... so shield every thing if you want to use the same housing...
Can anyone give me suggestion for output filter inductors?
I started to make a new secondary design for PC PSU.
So trafo have dual secondary, not CT. Each secondary is connected to two ultrafast diodes, one for cathode, one for anode. So now I have (in concept) 2x 35v and 2x-35V. I want to use two LM338 To regulate output for -30v.
But I am not sure about output inductores. They should be before first caps.
Here are my idea: Rectifier->Inductor->Caps->Regulator->->Joining and - for GND->some more caps
Should I use two,two dual or four inductors?
I started to make a new secondary design for PC PSU.
So trafo have dual secondary, not CT. Each secondary is connected to two ultrafast diodes, one for cathode, one for anode. So now I have (in concept) 2x 35v and 2x-35V. I want to use two LM338 To regulate output for -30v.
But I am not sure about output inductores. They should be before first caps.
Here are my idea: Rectifier->Inductor->Caps->Regulator->->Joining and - for GND->some more caps
Should I use two,two dual or four inductors?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.