Why not just connect the two trafo's, then you have your "CT", just remember to connect them in series, not parallel.... the coil or inductor, are you going to use as filters, for HF, or?
Go to Rod Elliots sound page(ESP), there can you find very basic and easy to use excel document, that calculates everything for you, you only have to input the wire area, core or no core, what fq or Henry, you want.
I think Texas Instrument have easy to use, filter programs, search the net, and use common sense when you choose the copper wire, .25 - 1 mm2 or go higher, depending on the current.... Good luck!
Go to Rod Elliots sound page(ESP), there can you find very basic and easy to use excel document, that calculates everything for you, you only have to input the wire area, core or no core, what fq or Henry, you want.
I think Texas Instrument have easy to use, filter programs, search the net, and use common sense when you choose the copper wire, .25 - 1 mm2 or go higher, depending on the current.... Good luck!
If you place the inductor (aka "choke") after the rectifier your get a DC voltage about equal to the AC voltage. Placing a cap first gives you a DC voltage of about 1.4 times the AC volts. So the choke input filter looses some voltage but solves the in-rush current problem nicely. What you are proposing is an "LC" filter before the regulator. Using two chokes would create an LCL or LCLC filter. The way to decide which to use is to first know how much ripple voltage you want going into the regulator. o decide on this look at the regulators drop out voltage and make sure none of the bottoms of the ripple will go under that voltage and then apply a "saftey factor" so as not to cut to close. An capastior input filter would give a larger voltage come not even close to dropout voltage but then the regulator will have to drop more voltage and run hotter.
Think about CLCRC, basically using a resistor in place of the second choke. Choose the resistor to drop the voltage to something the regulator can handle easy. a 25W can take some of the heat off the regulator.
Hi,
I give you my help, but frankly it is not clear.
PC already has the SMPS PWM (for regulation). Why do you want to put a linear regulator at output?
(This will dissipate a lot of power into heat).
it is easier to modify the transformer to obtain the regulated voltage (PWM) you need, I think.
However, if you refer to the filter output can wrap a cylinder ferrite 6mm diameter 30mm long wire with 1.24 mm (get about 10 uH), 220nF before and after the filter (this is p.greco filter).
Or buy one PSU true for audio, perfect for compact LM3886 amplifier, noise-free.
AP2, Thanks for suggestions. So why linear regulators? I want to make another PCB for rectifiers, and disconect smps trafo's secondary from original SMPS board. I dont want to mess up voltages with pwm regulation, so I rewound trafo's secondary for higher voltage at the same 33kHz freq.
I agree, thats a waste of power, but as I have lm388s and heatsinks.. And I want to remove fan from psu but add bigger heatsink to igbts and diodes.
I started my electronic practics with 40w soldering iron and old KT848 transistors so I dont want to buy new devices (and I cant spend my money too for it, I am just a student, without job). Usually I sample components and desolder from other devices to save money for better components at input path.
I hope now you understand why I chose pc psu.
I agree, thats a waste of power, but as I have lm388s and heatsinks.. And I want to remove fan from psu but add bigger heatsink to igbts and diodes.
I started my electronic practics with 40w soldering iron and old KT848 transistors
so I dont want to buy new devices (and I cant spend my money too for it, I am just a student, without job). Usually I sample components and desolder from other devices to save money for better components at input path.I hope now you understand why I chose pc psu.
AP2, Thanks for suggestions. So why linear regulators? I want to make another PCB for rectifiers, and disconect smps trafo's secondary from original SMPS board. I dont want to mess up voltages with pwm regulation, so I rewound trafo's secondary for higher voltage at the same 33kHz freq.
I agree, thats a waste of power, but as I have lm388s and heatsinks.. And I want to remove fan from psu but add bigger heatsink to igbts and diodes.
I started my electronic practics with 40w soldering iron and old KT848 transistors
I hope now you understand why I chose pc psu.
Hi,
Now is clare...this is as impossible mission but work.
I understand you student conditions (I remember of me).
filter is very good for remove HF residual (obvious one c-L-c for side) to gnd.
last question for you.
why not modify trafo (is simple) and complete it with my filter?
use at original pwm for +/-30V? it is simple job instead of your idea(I think).
Modify trafo is very simple,I can help you if you want
Regards
Hi,
C-L-C (10uH) 10A: filled all surface of cylinder (6x30)mm.with wire 1,12 mm
this block HF residual in very wide range of spectum. normally not necessary
applied formula.(this refers to suppressor filter (not for replace toroidal block inductance as original in the smps).
For trafo I mean simple modify (not rewind).
Also...for your modify said beefore, toroidal block is not necessary because you use fixed pwm,then only suppressor filter is necessary.
Regards
Last edited:
Well AP2 seems to know what he's talking about, and it was kind of obvious that you needed a power resistor to "correct" the current...
RC-filter is most commonly used in active filter designs, and I use 0R15//0R15@10W each in my linear PSU designs, I build(cause I love the warm sound) mostly class-A amp's, and to get a clean DC-source, I use a "capacitance multiplier", they are great for use with class-A stage, and can be use with AB-stage as well, but then you need a bigger cap at the output, and it can go in reverse, and "boom", bye bye cap multiplier... A-stage draw most current when not in use and the current drop as the watt increases... AB-stag draws some current, "quiescent current", and it varies from design to design... but 20ma to 100ma, put a fan or bigger heat sink on you AB amp and change the quiescent current from .02A -> 1A and you'll end up with a class-A stage...
And LM3386 is a pure B-stage, which equals zero quiescent current "capacitance multiplier" are out...which leave you with three choices, and SMPS might be the right one, if you get rid of all the "noise". SMPS are compact, high powered, efficient, unregulated is never a good idea and regulated PSU for B-stage are some times tricky to build.... so listen to AP2.... I think he can help you, better then I can... good luck to you, and "may the force be with you.." haha...
RC-filter is most commonly used in active filter designs, and I use 0R15//0R15@10W each in my linear PSU designs, I build(cause I love the warm sound) mostly class-A amp's, and to get a clean DC-source, I use a "capacitance multiplier", they are great for use with class-A stage, and can be use with AB-stage as well, but then you need a bigger cap at the output, and it can go in reverse, and "boom", bye bye cap multiplier... A-stage draw most current when not in use and the current drop as the watt increases... AB-stag draws some current, "quiescent current", and it varies from design to design... but 20ma to 100ma, put a fan or bigger heat sink on you AB amp and change the quiescent current from .02A -> 1A and you'll end up with a class-A stage...
And LM3386 is a pure B-stage, which equals zero quiescent current "capacitance multiplier" are out...which leave you with three choices, and SMPS might be the right one, if you get rid of all the "noise". SMPS are compact, high powered, efficient, unregulated is never a good idea and regulated PSU for B-stage are some times tricky to build.... so listen to AP2.... I think he can help you, better then I can... good luck to you, and "may the force be with you.." haha...
Well AP2 seems to know what he's talking about, and it was kind of obvious that you needed a power resistor to "correct" the current...
RC-filter is most commonly used in active filter designs, and I use 0R15//0R15@10W each in my linear PSU designs, I build(cause I love the warm sound) mostly class-A amp's, and to get a clean DC-source, I use a "capacitance multiplier", they are great for use with class-A stage, and can be use with AB-stage as well, but then you need a bigger cap at the output, and it can go in reverse, and "boom", bye bye cap multiplier... A-stage draw most current when not in use and the current drop as the watt increases... AB-stag draws some current, "quiescent current", and it varies from design to design... but 20ma to 100ma, put a fan or bigger heat sink on you AB amp and change the quiescent current from .02A -> 1A and you'll end up with a class-A stage...
And LM3386 is a pure B-stage, which equals zero quiescent current "capacitance multiplier" are out...which leave you with three choices, and SMPS might be the right one, if you get rid of all the "noise". SMPS are compact, high powered, efficient, unregulated is never a good idea and regulated PSU for B-stage are some times tricky to build.... so listen to AP2.... I think he can help you, better then I can... good luck to you, and "may the force be with you.." haha...
Hi,
I do not understand what you mean.
Perhaps the suppression filter that I recommended to "AR4"?
It is good for clean DC-Out ,because it is SMPS no linear psu.
now some of my PSU?
I do not use capacitors multiplier.
Sorry, I honestly do not understand what you mean.
Regards
Last edited:
Sorry, I ought to be more obvious, in what I mean... I only meant that you(AP2) knew what you where talking about, LC filter is the right way to suppress ripples from a SMPS unit.
I rumbled on about my own 'stuff', off topic, just throwing out some thoughts and ideas.
And I tend to build analogue solutions even though a MCU and "AVR Studio" and a few lines of C(and the proper MCU like ATMega32 w/ 8 ADC@16MHz) would do the same thing, and more precise....
I just meant that you seemed to have a good grasp when it comes to SMPS, and filters to get a clean DC "signal"... I'm sorry if I made you confused... Your input was right on the money.... like the Power Resistor i.e.
Now I'm doing it again!!! Rumbling on....off topic!!!
But, Yes I meant the suppression filter you recommended "AR4".
The "other stuff", was just some (un-)useful data...
I'll try to be more clear in the future! And to the point!
Regards - Joaquim
I rumbled on about my own 'stuff', off topic, just throwing out some thoughts and ideas.
And I tend to build analogue solutions even though a MCU and "AVR Studio" and a few lines of C(and the proper MCU like ATMega32 w/ 8 ADC@16MHz) would do the same thing, and more precise....
I just meant that you seemed to have a good grasp when it comes to SMPS, and filters to get a clean DC "signal"... I'm sorry if I made you confused... Your input was right on the money.... like the Power Resistor i.e.
Now I'm doing it again!!! Rumbling on....off topic!!!
But, Yes I meant the suppression filter you recommended "AR4".
The "other stuff", was just some (un-)useful data...
I'll try to be more clear in the future! And to the point!
Regards - Joaquim
Last edited:
Just keep "offtoping"
I am happy to hear some advices that have been tested in your own expierence. But I am still thinking about linear regulation. It makes a load to smps and isn't that good?
Question about caps.. Will non-low-esr caps go for boom? I have about 20 Vishay caps rated @100V & 470uF. I cant find datasheet for them (no series nr). Once I had added them to SMPS after low-esr ones... Worked fine, but that was just for a few hours a day.
I am happy to hear some advices that have been tested in your own expierence. But I am still thinking about linear regulation. It makes a load to smps and isn't that good?
Question about caps.. Will non-low-esr caps go for boom? I have about 20 Vishay caps rated @100V & 470uF. I cant find datasheet for them (no series nr). Once I had added them to SMPS after low-esr ones... Worked fine, but that was just for a few hours a day.
Just keep "offtoping"
I am happy to hear some advices that have been tested in your own expierence. But I am still thinking about linear regulation. It makes a load to smps and isn't that good?
Question about caps.. Will non-low-esr caps go for boom? I have about 20 Vishay caps rated @100V & 470uF. I cant find datasheet for them (no series nr). Once I had added them to SMPS after low-esr ones... Worked fine, but that was just for a few hours a day.
I think it's a good idea, if you use a linear reg. - like you said- it acts like a load for the SMPS unit. I f I got it right your "chain" looks like this(approx.);
SMPS--CLCRC(i.e. as LP-filter)--Lin.Reg---RCR(last res. to gnd)
The last "filter" after your regulation don't need to be "well done", 'cuz if you use big capacitors just before you pull out the regulated voltage and amp's, you increase the risk of blowing up your linear regulator, if it becomes in reverse or short circuit...
So always use twice the current power diode to protect your equipment... Now I'm back on "capacitance multiplier" again, since it's a linear way to regulate and increase the capacitance(for less the €10 you can build a "cap. multi." that is equal to a 1F(!!) capacitor).
I wrote that one could use them with an AB-stages, but then you need to have a "battery" at the output, if the amp switch to B operation or just jump up to twice the power... So you need to have "juice" to feed it with, but a rule of thumb is to not exceed 4700µF, so you're not able to drive a 100W AB-amp with a simple cap. multiplier. To the point... look at some "capacitance multiplier", it might give you ideas. I would probably use a adjustable regulator, like LM317T, and beef up the power with MJ15003 or "the old workhorse" 2N3055(max. 3A 40V per unit, MJ15003 can handle 5A@40V), or a power op-amp like LM12CLK(or plain LM12), it can handle 80W, 50V, and comes in a TO-2 casing with four pins.... LM12 have some very good application hints in the datasheet, and a hookup for volt/amp adj. The drawback is the price, they cost in Europe €30 - €40 each, in China €5 including shipping....
LM317 have a 3A sort, that also comes in a TO-2 casing... and good application hint and diagrams... Active components suppress ripples better the a LC/RC filters does.. LM317T have >80dB noise reduction i.e. But if you use a BJT and feed it with ripples... well you'll get bigger ripples.... trail and error, it can only go wrong
LM723, is worthy a look....
Last edited:
That is good but TOP250 very unpredictable http://www.irbislab.ru/modules.php?name=Content&pa=showpage&pid=23
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.