Here's the problem in the ultrasonic and low hf range -- "unexplained offsets", and the "Seemingly Unexplainable Amplitude Error" -- THD which appears for no good reason. Every bipolar transistor can be turned into an RF detector, and they will. ADI and Linear discuss the problem in application notes.
You have right Marce but it is more dificult to made a SMPS with low EMC noise especially home made. The layout it is more critical and filtering need special capacitors.
The EMC and noise are more simple to reject when you work at 50-60Hz (100-120Hz). The noise problem with standard (linear) power supply become a real problem when you use very fast diodes (made for SMPS) and very big capacitors (over lets say 10mF-20mF on rail). Using very fast operationals will not help to keep switching noise outside from signal path. The switching noise will be amplified by fast operationals and will generate intermodulation distorsions.
The problem are even more visible if you work with video signals but, of course, all TV sets have SMPS. So it is possible but it is not a simple task.
I do a fair bit of SMPS layout for very critical designs, where noise would be even more of a problem, including audio and video communication systems, and contrary to what people think, when you look at SMPSs (or any PSU) they split down into simple sections.
Switching loops and analogue control signals, keep the switching loops and its related ground return separate from the main ground (connect at on point only) and you a lot of the way there.
it is one area of electronic designs where there is a lot design guides and layout notes, it is worth studying them, and where a guide is produced for a particular controller follow it, but always remember a demo board is designed to be used as a stand alone board, when placing the supply on a bigger board with circuitry on it you will have to look at loops and make sure they are isolated.
Switching loops and analogue control signals, keep the switching loops and its related ground return separate from the main ground (connect at on point only) and you a lot of the way there.
it is one area of electronic designs where there is a lot design guides and layout notes, it is worth studying them, and where a guide is produced for a particular controller follow it, but always remember a demo board is designed to be used as a stand alone board, when placing the supply on a bigger board with circuitry on it you will have to look at loops and make sure they are isolated.
I suppose generally, laptop SMPS are better designed & built than "cheaper" generic ones sold widely? I've built several 12VA (2x15v x 400mA) low power linear supplies & these cost me just under £15,00 (<$23,00) including the plastic psu case! The question is, would I be able to buy an off the shelf SMPS of comparable quality & performance? I doubt it!
My biggest hurdle so far has been sourcing a 30v laptop or similar quality SMPS!
Q:How much dynamic reduction in an audio preamp can one expect using a 24v psu instead of a 30v supply?
Last edited:
Allmost nothing if are made using OA. For a maximum 4-6V rms out (< 18Vp-p) You have a voltage reserve until cliping >6V.
But usually the output voltage shuld be under 6Vp-p, so no cliping.
I have dissected numerous "replacement" PSUs sold on ebay/Amazon and the quality is abysmal.
Mains wires as thin as a strand of hair, ready to snap off at the slightest provocation.
Huge voltage spikes on either of the output leads (100s of volts) made safe only with a hard connection to the mains earth!
Blue capacitor Y2 between primary and secondary of transformer, pushing untold voltages and noise spikes to the output - unless you ground it.
General white noise on the line increasing with load - as if the above problems were not enough. Unlike 100Hz ripple, this is multiple (high) frequencies, from a few KHz to 100s of KHz.
*******************
Even my genuine SONY VAIO adaptor, original manufacturer and two prong (and thus better quality), also pushes 10s of volts at the output - must be that damned capacitor.
******************
Akis, you're indeed a "minefield" of information!Lol! Much relieved, that I can skip dissecting these & getting a charred face! B.t.w how DOES one pry open these damn SMPS casings? Do these get sealed by elephants stepping on them?
Are these worth the time & effort modifying, replacing new caps,wires etc or should I forget about SMPS & just stick to my linear psu?
Are these worth the time & effort modifying, replacing new caps,wires etc or should I forget about SMPS & just stick to my linear psu?
A little tale of SMPSs, bought a container load from China, samples worked perfectly the items sent when tested were abysmal, couldn't CE them they were so bad. Cheep SMPSs from e-bay and similar are going to be just that cheep.
Considering the PSU is the heart of any system, what percentage of total system cost would you say is reasonable?
Me 25-50%.
Considering the PSU is the heart of any system, what percentage of total system cost would you say is reasonable?
Me 25-50%.
Some years ago I bought a bucket load of Murata BWR-15/330-D12A isolated switchers for dirt cheap; they sell for around $70 each new. 12V input with +/-15V @ 330mA output.
I've used these to power pretty much all of the line level projects I've done and never had an issue with them. I'll usually throw a CLC on the output just because I like to think it does something to lower their 70mVp-p ripple. The best part is how small they are.
I've used these to power pretty much all of the line level projects I've done and never had an issue with them. I'll usually throw a CLC on the output just because I like to think it does something to lower their 70mVp-p ripple. The best part is how small they are.
I wouldn't compromise on the p.s.u within it's limits, as this is the heart of any electronic system for it's efficiency & safe operation;but having said that it need not be an overkill with zillion caps, filters & inductors. A good balance between cost & the quality of the end product one is aiming for should be the yard stick I.M.H.O
Flat head screwdriver, insert into the seam that runs in the middle of the case, the case is really two halves glued together, try to prise it open at a single point, work your way around the seam trying to twist the blade until one point gives, once it gives yank the two halves apart with your hands if you are strong and wearing gloves, or insert a wider the blade into the new opening and twist some more along the seam. In addition there always is adhesive tape or gum holding the internal plastic/metal case - so even when the external case is open it still requires a lot of pulling and really slowly to get the insides out.
You're excused for not knowing that obscure "OA".
opamp is universally known.
I too and probably many others, did not know what OA meant.
Can someone, or many together, design a SMPS component (with PCB layout of course) that we (the non experts) can then use as a building block for our projects?
One of my requirements would be :
input : 110-240 VAC NOT EARTHED
output : from 5V DC to 40V DC at various amps, depending on the module, 5A would be top requirement, but we could have 0.5A, 1A, 2A and 3A modules.
adj : a pin accepting a voltage that adjusts the DC output of the SMPS
current limit : only to prevent device from destruction (eg shorted output)
thermal limit : to prevent device from destruction
reverse voltage at the output protection : would be nice to have (eg diode at the output)
Once a range of such modules were available, with PCB, we would be able to incorporate them into our designs.
For example I have bought a bunch of LM2596 DC-DC converter modules on ebay, very cheap less than a pound each, and I use them wherever there is a need to convert a usually higher DC to a lower DC. They have a trimmer to adjust the output voltage, easily replaceable with a transistor and a resistor to become variable output.
I have also made a switched capacitor voltage doubler module, so that I can get double voltage where I need.
I can mount those ready made PCBs onto other PCBs in a modular design. So like I said above, if anyone would be willing to design a really good quality SMPS it would be great.
One of my requirements would be :
input : 110-240 VAC NOT EARTHED
output : from 5V DC to 40V DC at various amps, depending on the module, 5A would be top requirement, but we could have 0.5A, 1A, 2A and 3A modules.
adj : a pin accepting a voltage that adjusts the DC output of the SMPS
current limit : only to prevent device from destruction (eg shorted output)
thermal limit : to prevent device from destruction
reverse voltage at the output protection : would be nice to have (eg diode at the output)
Once a range of such modules were available, with PCB, we would be able to incorporate them into our designs.
For example I have bought a bunch of LM2596 DC-DC converter modules on ebay, very cheap less than a pound each, and I use them wherever there is a need to convert a usually higher DC to a lower DC. They have a trimmer to adjust the output voltage, easily replaceable with a transistor and a resistor to become variable output.
I have also made a switched capacitor voltage doubler module, so that I can get double voltage where I need.
I can mount those ready made PCBs onto other PCBs in a modular design. So like I said above, if anyone would be willing to design a really good quality SMPS it would be great.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.