monster Posted August 2, 2006 Report Share Posted August 2, 2006 I have a question here...i currently doing a power supply which is capable to supply max of 6V and 5A of output. the PSU is a digital PSU with microcontroller and LCD interface which means that the output can be adjusted to any desired value.the problem is the output ripple noise of the PSU is not really satisfying. Maybe can apply some filter circuit on the output terminal of the PSU. Am i rite? But my question is how to choose the right capacitor type and value? I try with a 470uf electrolytic capacitor but the problem is it causing slower stepping down the output voltage, let's say from 6V to 1V (in fact it shd be instantaneous). Any solution to that?All helps and advices are highly appreciated..thanks 1st. Quote Link to comment Share on other sites More sharing options...
audioguru Posted August 2, 2006 Report Share Posted August 2, 2006 Your PSU probably has a switching regulator. Switching regulators have ripple noise on their output due to the way they operate.That's why I like and use linear regulators that don't have switching ripple noise.Some PSU's have a switching pre-regulator for low heating, followed by a tracking linear regulator for low noise. Quote Link to comment Share on other sites More sharing options...
indulis Posted August 2, 2006 Report Share Posted August 2, 2006 Check out a Cuk converter if you believe all switch mode power supplies have noise and ripple. Quote Link to comment Share on other sites More sharing options...
audioguru Posted August 2, 2006 Report Share Posted August 2, 2006 Doesn't a Cuk Converter require a certain amount of load to avoid resonance in the output LC? Quote Link to comment Share on other sites More sharing options...
indulis Posted August 2, 2006 Report Share Posted August 2, 2006 Technically, all switch mode converters require a load (especially converters that run in the continuous conduction mode). Voltage mode converters are even worse in regards to min load levels because of the right halfplane zero in the transfer function.Specifically for a Cuk, I don't know. That converter topology has been covered by patents for many years, I'm not aware of anyone offering it as a off the shelf technology for standard offerings. For the most part, people just stay away. I did hear that Dr Cuk left Cal Tech and I'm not sure who holds the rights to that technology now.Just a guess, but I think with typical values L&C, the resonance frequency would be way outside the bandwidth of the converter, and in all likelihood there is enough R around to kill the Q.I guess not.... Quote Link to comment Share on other sites More sharing options...
ante Posted August 2, 2006 Report Share Posted August 2, 2006 Hi Guys,I found an article about Cuk converters, interesting project I think! [attachment deleted by admin] Quote Link to comment Share on other sites More sharing options...
monster Posted August 3, 2006 Author Report Share Posted August 3, 2006 thanks for the reply at 1st...well...my PSU actually nothing but juz consists of a transformer, brigde rectifier and +15/-15 voltage regulator for the op amp. so i guess there isn't any switching regulator in my PSU.as wat i stated early...the ripple noise can be reduced to <2mV by applying filter capacitor at the 2 output terminal of the PSU. the problem is i not sure the exact value of the capacitor and type to further reduces the ripple.another problem of applying filter cap at output caused voltage to drop during full load. I cant figure out wat causing the voltage drop.Furthermore, i dunno shd i use the LC filter, RC filter or etc methods.thanks and regards Quote Link to comment Share on other sites More sharing options...
audioguru Posted August 3, 2006 Report Share Posted August 3, 2006 Is the ripple at a low frequency or at a high frequency?The sensitivity to mains supply ripple for a cheap old opamp is a max of 150uV at the output for each volt of supply ripple. So maybe your ripple problem is caused by poor filtering of the opamp's input signal or lack of signal shielding.Adding a filter capacitor across the output of a supply doesn't cause the voltage to drop if the capacitor doesn't draw DC current. Your voltmeter might be resonding to the peak voltage of the ripple. Quote Link to comment Share on other sites More sharing options...
ante Posted August 3, 2006 Report Share Posted August 3, 2006 BTW; 6 Volts might not be considered ´High Voltage Stuff´ ;) Quote Link to comment Share on other sites More sharing options...
Cabwood Posted August 3, 2006 Report Share Posted August 3, 2006 The biggest source of ripple is from an inadequate reservoir capacitor between the bridge recifier and the regulator. Make that bigger, and you reduce the ripple at the regulator's input, and the regulator's ripple rejection isn't stretched so far.Assuming you are drawing 1A from the supply's output, A 1000uF capacitor will experience a ripple of roughly 8V, which must be rejected by the regulator - hard work by any account. Double the capacitance, and you halve the ripple. Under those same conditions, 10000uF will therefore reduce the ripple down to less than 1V, which will result in less ripple at the regulator's output.Keep in mind that the drop-out voltage of the regulator will determine the minimum input voltage that it can properly regulate. The LM317 for example needs an input which is at least 2.5V higher than it's output. So, to get it to reliably output 6V it must have an input voltage which never drops below 8.5V. If the input ripple (across the reservoir capacitor) is too large, that voltage may drop below 8.5V, and since the regulator may then fail to regulate, it will appear at the output as 100Hz (UK) or 120Hz (US) ripple.If this is not the problem, then how about passing the existing output through another regulator (two stage regulation, I suppose you would call it). Since the ripple at the input of the second regulator is so small, the output should be pretty well flat. The problem with this is again the drop-out voltage of the regulators. You'd need to ensure that each regulator's input is at least a couple of volts higher than its output, typically.It doesn't make sense to put large capacitors at the output of a linear regulator, for the reason you have observed, and because most circuits don't sink current in huge spikes. When circuits do do that it's better to position the capactior electrically close to the current sinking component itself. Quote Link to comment Share on other sites More sharing options...
monster Posted August 4, 2006 Author Report Share Posted August 4, 2006 to ante sorry if i posted the thread in a wrong section :Pwell actually the 6V 5A is just a single power supply output, in facts, my PSU consists of +25V 1A and -25V 1A of output as well. Since now in the troubleshooting stage, i dun intend to mess up all the part as it complicated the situation. Further, the +25V/-25V supply is separated with the 6V supply.After the transformer is the bridge rectifier, so there isn't any regulator for the AC to DC voltage. the only regulator circuit i have is the LM7815 and LM7915 for the op amp. the output is control by 2 power transistor based on the control signal from the MCU. to audioguru, i cant get the freq of the signal from the scope as the scope indicating several MHz. Is that possible? I think something wrong with the reading or scope setup. Even the ripple noise rms of the scope is different from the meter reading. the op amp i used is LF412 from NS or TI. Quote Link to comment Share on other sites More sharing options...
audioguru Posted August 4, 2006 Report Share Posted August 4, 2006 Since your 6V supply doesn't have a voltage regulator then its ripple is at 120Hz in North America.Your 15V regulators or opamp are oscillating at several MHz or the opamp is picking up a local SW radio transmission at several MHz. Quote Link to comment Share on other sites More sharing options...
monster Posted August 9, 2006 Author Report Share Posted August 9, 2006 Okay i'll try to troubleshoot further and will notify if got any improvement over the ripple noise...thanks for the guidances.. :) Quote Link to comment Share on other sites More sharing options...
monster Posted August 10, 2006 Author Report Share Posted August 10, 2006 hi..i noticed that my power source (24-0-24 rating transformer connected to full wave bridge rectifier and filter out by 2x4700uF, 1000uF, 0.1uF and 5.6K resistor all in parallel) having a ripple noise of around 4mV-8mV. Any idea to further reduce the ripple noise? and hopfully can reduce my overall PSU ripple.Any formula to calculate the ideal cap value?thks and regards. Quote Link to comment Share on other sites More sharing options...
audioguru Posted August 10, 2006 Report Share Posted August 10, 2006 4mV to 8mV of ripple on an unregulated supply is very, very low. You must have measured it without a load. Load it with 1A and the ripple voltage will be more than 1Vp-p.You said your transformer is rated at 24V-0V-24V? Then its unregulated DC voltages should be about 32VDC. Quote Link to comment Share on other sites More sharing options...
monster Posted August 10, 2006 Author Report Share Posted August 10, 2006 yes...the measurement was done without load..with load it definately higher(with 25ohms at full load). the desired max output is 25V 1A.the 32VDC is supply to power transistor circuitry to output the desired voltage. Quote Link to comment Share on other sites More sharing options...
audioguru Posted August 22, 2006 Report Share Posted August 22, 2006 Attach a schematic of the circuit and show parts values. Mark where you measured the ripple voltage.Does the transformer get hot when all DC voltages are loaded? Quote Link to comment Share on other sites More sharing options...
Voytek Posted August 22, 2006 Report Share Posted August 22, 2006 I have a question here...i currently doing a power supply which is capable to supply max of 6V and 5A of output. the PSU is a digital PSU with microcontroller and LCD interface which means that the output can be adjusted to any desired value.the problem is the output ripple noise of the PSU is not really satisfying. Maybe can apply some filter circuit on the output terminal of the PSU. Am i rite? But my question is how to choose the right capacitor type and value? I try with a 470uf electrolytic capacitor but the problem is it causing slower stepping down the output voltage, let's say from 6V to 1V (in fact it shd be instantaneous). Any solution to that?All helps and advices are highly appreciated..thanks 1st. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.