What particular characteristic of a capacitor should I look for when selecting one for filter duty in a SMPS ?
There are capacitors that are Low ESR, Low Impedance,Inductance, HF (High frequency), etc..
My guess is that HF types (for high freq ripple) would be best suited but I am not completely sure.
{edit}: this is a low power smps (cd player) and the final output voltages are regulated via regulators.
There are capacitors that are Low ESR, Low Impedance,Inductance, HF (High frequency), etc..
My guess is that HF types (for high freq ripple) would be best suited but I am not completely sure.
{edit}: this is a low power smps (cd player) and the final output voltages are regulated via regulators.
This is not a straight easy answer.... The output filter cap choice is dependent on many factors....
Many SMPS circuits, even new ones are poor designs... many designers have never advanced from 1980's SMPS technology(Unitrode era) and still continue to produce garbage and with very borderline loop stability... I still see large voltage mode switchers and i also still see low frequency switchers, anything under 300kHz....
It is many of these cases where you need the output cap to have large ESR..For example where the ESR ZERO will cancel the 2nd order pole and produce a -1 slope (-20dB/Decade).... Replace the output cap with a low ESR ceramic and you will create a screaching oscillator!!!!
Most modern switchers would like to have low ESR ceramics on the output...but they do not come in large values given the voltage range....SOmetimes you need the electrolytics for the size and a few ceramics for bypass, depending on the output load...
If you can provide more detailed infor on the SMPS you are working with as well as the power requirements..
Chris
Many SMPS circuits, even new ones are poor designs... many designers have never advanced from 1980's SMPS technology(Unitrode era) and still continue to produce garbage and with very borderline loop stability... I still see large voltage mode switchers and i also still see low frequency switchers, anything under 300kHz....
It is many of these cases where you need the output cap to have large ESR..For example where the ESR ZERO will cancel the 2nd order pole and produce a -1 slope (-20dB/Decade).... Replace the output cap with a low ESR ceramic and you will create a screaching oscillator!!!!
Most modern switchers would like to have low ESR ceramics on the output...but they do not come in large values given the voltage range....SOmetimes you need the electrolytics for the size and a few ceramics for bypass, depending on the output load...
If you can provide more detailed infor on the SMPS you are working with as well as the power requirements..
Chris
well I am talking about the psu of the pioneer dv-563 player.
let me see if I can scan just the psu schematic page and post it tonight.
I know people have already done some mods and I could just copy them but I want to run it by the "Power Supply" forum and get some expert opinions.
let me see if I can scan just the psu schematic page and post it tonight.
I know people have already done some mods and I could just copy them but I want to run it by the "Power Supply" forum and get some expert opinions.
You should not modify a multi-output flyback (the topology employed in most DVDs, VCRs, TVs, etc...) Capacitor parasitistics (ESR) are often part of the desired frequency response of the output filter in these circuits, as it has been already pointed out.
You may consider additional common-mode filtering or output pi-filtering at best.
You may consider additional common-mode filtering or output pi-filtering at best.
ok here's the schematic. Whats the best approach to make better of what we have here ?
Let me know if you have questions or need more info.
Let me know if you have questions or need more info.
An externally hosted image should be here but it was not working when we last tested it.
I am unable to read the tiny print on the outputs...maybe you can clarify this for me...
It appears as Eva says...No current limit...
It also appears that they used series pass regulation on the secondary side for a few of the output voltages... In these cases, you can add extra caps to those outputs.... But you would need to get out some equipment and measure the noise floor in the output voltages to get an idea where you need to make your improvements....
This is typical crappy power supply from the far east....This circuit was lifted from another design and the output reg was most likely horrible, so they attached some secondary series regs to "band-aid" the problem, which makes switcher efficiency take a nose dive....
It is a flyback....Probably a dis-continous mode ...which means they are probably not dealing with a (RHPZ) Right-Hand-Plane-ZERO.....adding extra output Caps on the not regulated outputs would push the pole lower and reduce loop gain at cross-over...
You could de-couple the output with a large L then use a cap afterwards....need to verify the results...if you don't have fancy gear..you step response it and look at the over-shoot to get a rough idea if your safe...
Chris
It appears as Eva says...No current limit...
It also appears that they used series pass regulation on the secondary side for a few of the output voltages... In these cases, you can add extra caps to those outputs.... But you would need to get out some equipment and measure the noise floor in the output voltages to get an idea where you need to make your improvements....
This is typical crappy power supply from the far east....This circuit was lifted from another design and the output reg was most likely horrible, so they attached some secondary series regs to "band-aid" the problem, which makes switcher efficiency take a nose dive....
It is a flyback....Probably a dis-continous mode ...which means they are probably not dealing with a (RHPZ) Right-Hand-Plane-ZERO.....adding extra output Caps on the not regulated outputs would push the pole lower and reduce loop gain at cross-over...
You could de-couple the output with a large L then use a cap afterwards....need to verify the results...if you don't have fancy gear..you step response it and look at the over-shoot to get a rough idea if your safe...
Chris
To my disappointment I replace more cheap standard electrolytics in SMPS than I like. Despite those clever engineers that design them they often use normal caps that blow from too high frequency/heat buildup or whatever. All too often I get the impression they just use a value they happen to have in stock.
Panasonic FM is a good choice.
Panasonic FM is a good choice.
sure, here's the legend in this order starting from the top:cerrem said:
SW +12V
E +6V (this one then goes to 3 other regulators. Two 3V and one 5V regulator)
E +6V
P_CONT
EV +3.3V
SW +1.8V
GND
GND
GND
GND
-28V
FLDC+
FLDC-
Interesting note about the FM series. Seems like they are the lower impedance and higher ripple current cousins of the FC series.jean-paul said:
cerrem said:
Chris,
You make some interresting comments here....
Sorry if this is off topic
I often use N87 core material at 300kHz, but have been looking for something more effecient.... the problem is that working temperature is above 65 degrees C and I have been unable to find something better. The 300kHz freq is fixed (don't ask) can you propose another core material?
Thanks for any input.
\Jens
If your using EPCOS material.... Take a look at the N49 material...
Core losses are at it's lowest from 60 to 90 C.....
Chris
Core losses are at it's lowest from 60 to 90 C.....
Chris
there is no way to tell what is the switching frequency of this smps, other than actually measure it, right ? 
percy said:there is no way to tell what is the switching frequency of this smps, other than actually measure it, right ?
If you have the values of the circuit components on the driver circuit... Lower right side of schematic, you then can calculate the self res. freq...
Chris
I have good reason to believe its 200Khz.
All the caps are Rubycon ZL (except a couple of exceptions, see below), so they are all Low ESR. They just cheaped out on the size of the capacitors. They could use some help there.
The 'input' cap is an unknown brand and I think way undersized at 82uf.
And another one is the line filter cap which right now is 0.1uf. Any suggestions for what to use for these ?
All the caps are Rubycon ZL (except a couple of exceptions, see below), so they are all Low ESR. They just cheaped out on the size of the capacitors. They could use some help there.
The 'input' cap is an unknown brand and I think way undersized at 82uf.
And another one is the line filter cap which right now is 0.1uf. Any suggestions for what to use for these ?
- Status
- Not open for further replies.