Question about using small "potted" board mount SMPS with low power audio circuits like preamps or active crossovers.
I've been playing with this circuit: 12dB / Octave Linkwitz Riley Crossover, and I quite like the result. For power it needs +/- 12v at order of magnitude 100mA current.
I am thinking of using a supply like Digikey p/n 945-2097-5-ND or the like because I can fit the entire lot into a small box.
I would escape the supply rails with (say) parallel 100n film + 1n ceramic caps and maybe series inductors if need be. I think this would keep the voltage clean.
Question is whether I can expect problems with radiated EMI or noise from the SMPS?
Thanks in advance!
BTW - attached pics of the power amp I'm using, some of you might know it well!
I've been playing with this circuit: 12dB / Octave Linkwitz Riley Crossover, and I quite like the result. For power it needs +/- 12v at order of magnitude 100mA current.
I am thinking of using a supply like Digikey p/n 945-2097-5-ND or the like because I can fit the entire lot into a small box.
I would escape the supply rails with (say) parallel 100n film + 1n ceramic caps and maybe series inductors if need be. I think this would keep the voltage clean.
Question is whether I can expect problems with radiated EMI or noise from the SMPS?
Thanks in advance!
BTW - attached pics of the power amp I'm using, some of you might know it well!
Attachments
Hi,
I like to use such power modules ... never had any issues regading noise or EMC.
I´d just look for modules with possibly higher clock frequency (>100kHz) and lower ripple.
200mVpp ripple is quite high and 67kHz rather low standard.
For the given specs the price tag of the Recom appears rather high.
Made very good experience with AC-DC and DC-DC converters from XPPower ... see the ECL15 UD01 or UD02.
A bit of post filtering is necessary and sufficient, but using active filtering with capacitance multipliers/gyrators to ´clean´ the supply lines is recommended.
The modules supply for already constant output voltage, which just asks for a ´fast´ filtering.
I find discrete cap multipliers (look for Keantoken´s cap multiplier) better in this regard than 3-pin regulators.
jauu
Calvin
I like to use such power modules ... never had any issues regading noise or EMC.
I´d just look for modules with possibly higher clock frequency (>100kHz) and lower ripple.
200mVpp ripple is quite high and 67kHz rather low standard.
For the given specs the price tag of the Recom appears rather high.
Made very good experience with AC-DC and DC-DC converters from XPPower ... see the ECL15 UD01 or UD02.
A bit of post filtering is necessary and sufficient, but using active filtering with capacitance multipliers/gyrators to ´clean´ the supply lines is recommended.
The modules supply for already constant output voltage, which just asks for a ´fast´ filtering.
I find discrete cap multipliers (look for Keantoken´s cap multiplier) better in this regard than 3-pin regulators.
jauu
Calvin
If you're not careful you should expect problems. With a little caution they work splendidly.
Best way to tell, absent a spectrum analyzer, is to have a small transistor radio handy -- if you can hear the interference on a blank part of the spectrum when you start up, there's an issue.
Consider putting ferrite bead or a small choke on the input to the converter in any event.
I've used the Murata DC-DC converters with no major issues -- as long as I put a small choke on the input.
you have op-amp supplies already, look at the power amps schematic and repeat that method. You probably won't need 100mA, each op-amp needs about 4-8 mA, for example a 2 way X-over with buffers is only 25 mA. Once you start building these active thingies you want a power logic sequence for starting the power amps last and shutting down it first to prevent speaker damage. Looking at your images it makes sense to build this into the power amp chassis.
SMPS are all different beasts , some are alot noisier than others, the thing they don't tell you is the common mode noise they introduce. unless you want to revisit the power supply over and over and would steer clear of switchers for linear audio preamps. edit > depending on your comfort level , you could build your own custom transformer by adding wires onto that toroid .
SMPS are all different beasts , some are alot noisier than others, the thing they don't tell you is the common mode noise they introduce. unless you want to revisit the power supply over and over and would steer clear of switchers for linear audio preamps. edit > depending on your comfort level , you could build your own custom transformer by adding wires onto that toroid .
Last edited:
and this helps the OP how?
most ppl can build a pretty good DC linear supply but not many can specify or tame an off-line SMPS. for example in audio / common mode noise can be easily converted, down-sampled in band with SE audio systems. > if you don't have the tools and knowledge a SMPS can bite you in the a**. Most good pro audio amps prevent RF ingress, why would one purposely add a AM transmitter inside a small signal box?
Last edited:
It helps the OP by saying that linear supplies are often a source of LF noise that is often in the audio frequency range, sorry didn't know I had to ask permission before posting a comment....
Common mode ferrite....or choke.
And how many SMPSs are used in pro audio gear.....
SMPSs are not as bad as made out, just pointing out that linear supplies can have noise in the audio band so its horses for courses and HF noise is easier to filter than LF noise.
Now I don't want to comment anymore because it will not help you... But one method you could use is an external SMPS to provide a system voltage then localised LDO power islands for the circuitry, this will allow better separation of power and channels if required will limit noise and also make noise more localised... And finally allow you to use a Linear (big bulky and hummy) supply or a SMPS fed by a two pin power connector, you can also shield the main supply in a separate box no magnetic or E field EMC, add ferrites to the output etc....
Common mode ferrite....or choke.
And how many SMPSs are used in pro audio gear.....
SMPSs are not as bad as made out, just pointing out that linear supplies can have noise in the audio band so its horses for courses and HF noise is easier to filter than LF noise.
Now I don't want to comment anymore because it will not help you... But one method you could use is an external SMPS to provide a system voltage then localised LDO power islands for the circuitry, this will allow better separation of power and channels if required will limit noise and also make noise more localised... And finally allow you to use a Linear (big bulky and hummy) supply or a SMPS fed by a two pin power connector, you can also shield the main supply in a separate box no magnetic or E field EMC, add ferrites to the output etc....
of course this is a public forum, but your ambiguous post just adds doubt and confusion. ( and you haven't cleared it up yet )
In my experience HF often is more difficult to filter out than LF, b/c you need RF techniques on filter coils and ceramic bypasses. LF is a lot easier, look at an op-amps PSSR for some clues. LM317 ripple and noise compared to the SMPS 200mV broadband ripple and hash hmmm...
SMPS come in all sorts of flavors, of course some are better than others, but again you need experience and one gains some experience by making mistakes and then fixing them. If the OP wants to travel down that road it's his choice, I'm just posting caution.
for a little taste just on quantifying the effect not on the fixes! see here https://youtu.be/BFLZm4LbzQU
In my experience HF often is more difficult to filter out than LF, b/c you need RF techniques on filter coils and ceramic bypasses. LF is a lot easier, look at an op-amps PSSR for some clues. LM317 ripple and noise compared to the SMPS 200mV broadband ripple and hash hmmm...
SMPS come in all sorts of flavors, of course some are better than others, but again you need experience and one gains some experience by making mistakes and then fixing them. If the OP wants to travel down that road it's his choice, I'm just posting caution.
for a little taste just on quantifying the effect not on the fixes! see here https://youtu.be/BFLZm4LbzQU
Last edited:
yup linear lab PS with common mode, can you imagine the tail chasing with 2 SMPS beating together.
why are you upset/ spend some time in the lab pulling your hair out with these things.
OK dont look at CM noise effects yet, simply compare the rpple and noise specs of a SMPS to basic linear regulators output (BTW the linear excels at LF) and then decide on the xtra filtering needed looking at the op-amps PSRR. SMPS can have up 2 or 3 different ripple components from line related to multiple switch related harmonics. closed loop / open loop, synchronous/ unsychronous and everything in between.
why are you upset/ spend some time in the lab pulling your hair out with these things.
OK dont look at CM noise effects yet, simply compare the rpple and noise specs of a SMPS to basic linear regulators output (BTW the linear excels at LF) and then decide on the xtra filtering needed looking at the op-amps PSRR. SMPS can have up 2 or 3 different ripple components from line related to multiple switch related harmonics. closed loop / open loop, synchronous/ unsychronous and everything in between.
Last edited:
I have done only with layouts, small ones in headsets, efficiency was the requirement here, to just last few weeks doing a BIG one, copper ribbon transformer, 6oz copper BIG... All power supplies are/can be problematic. SMPSs do get bad press in some areas, but it can be unfounded these days.
For such small current requirements as said I would be tempted with local LDOs and a master supply to the unit. It is common these days to have distributed power islands not just on digital but analogue as well, smaller local current draw thus smaller and more efficient local filtering, quite often 317/337 reg's as well for larger analogue sections.
Wasn't really that upset just being a drama queen, and spotted the common mode was done on a linear which tickled me, would have enjoyed the video more but the sound on my laptop has failed!!! and I cant find why.
For such small current requirements as said I would be tempted with local LDOs and a master supply to the unit. It is common these days to have distributed power islands not just on digital but analogue as well, smaller local current draw thus smaller and more efficient local filtering, quite often 317/337 reg's as well for larger analogue sections.
Wasn't really that upset just being a drama queen, and spotted the common mode was done on a linear which tickled me, would have enjoyed the video more but the sound on my laptop has failed!!! and I cant find why.
well I have designed, built, and tested SMPS I wouldn't recommend them for analog stuff unless the pro's clearly outweighed the cons. specs don't tell you too much until you get them in the system and start testing. you can expect pretty much rubbish at 200mV p-p! like Jack I have used some pretty good ones from murata, DC/DC designed for low noise VCO varactor TV tuners.
Pros use them for sound reinforcement, but not in studios so much, but more so when the interfaces go all digital.
Pros use them for sound reinforcement, but not in studios so much, but more so when the interfaces go all digital.
I have worked on many electronics systems including very sensitive analogue and we have designed and used SMPSs with no problems (or no more problems than normal) and one audio based system had more problems with a linear supply, including pro gear.
I am sorry but I do have to disagree with you here, not only are they smaller, cheaper but with all the legislation on power consumption they are often the only choice. I do believe Mr Pass and other consumer audio based companies are using them.
I see to many of them being used (every job) to share your view I am afraid. Though as I have said in many systems power is distributed to local convertors or filtered and isolated power islands.
We will have to agree to disagree on this....
I am sorry but I do have to disagree with you here, not only are they smaller, cheaper but with all the legislation on power consumption they are often the only choice. I do believe Mr Pass and other consumer audio based companies are using them.
I see to many of them being used (every job) to share your view I am afraid. Though as I have said in many systems power is distributed to local convertors or filtered and isolated power islands.
We will have to agree to disagree on this....
please point out a N Pass commercial audio design that use SMPS?
Marce your background is PCB design, so your beliefs are from a distance eg 2nd hand opinions. tactical comms headphones are far from sensitive analog requirements IMO. besides a military electronics SMPS is a little different than picking out a potted offline switcher from a catalog.
the OP should look at more options before randomly choosing a device, he has already selected a linear transformer over a SMPS for the amp portion so it doesn't make sense suddenly switch over for a few op-amps. he should do the tradeoffs. , Maybe you should build and TEST one or two Hi-Fi audio projects before suggesting others use them willy nilly. why don't you pick out a good one out for the OP, then you can help him when he has some issues in testing?
Marce your background is PCB design, so your beliefs are from a distance eg 2nd hand opinions. tactical comms headphones are far from sensitive analog requirements IMO. besides a military electronics SMPS is a little different than picking out a potted offline switcher from a catalog.
the OP should look at more options before randomly choosing a device, he has already selected a linear transformer over a SMPS for the amp portion so it doesn't make sense suddenly switch over for a few op-amps. he should do the tradeoffs. , Maybe you should build and TEST one or two Hi-Fi audio projects before suggesting others use them willy nilly. why don't you pick out a good one out for the OP, then you can help him when he has some issues in testing?
Its getting rather boring this assumption of what I do, whet I know and how far divorced from stuff I am.... Do you know how many years I spent testing and building electronics test gear, what I did at collage etc... No
And the same sh** build some audio stuff etc.
.Sorry but you don't know how involved I get in projects sorry....
Neither do I just do Mil stuff, we do a whole lot of stuff including sensitive analogue.
After the last insult like this all I can say is it is a pathetic way to carry on a discussion, presume the other person has no clue about what is going on and thus insinuate his comment are invalid and have no worth. Happening more and more on here and rather unexpected from you.
Just because I disagree....Tear me down...
I have suggested some things that are relevant and worth considering you have decried me at every point because you don't like SMPS supplies.
Nelson Pass has mentioned in other threads that he uses SMPSs.
I work on site as part of a team so with both headphones and the next gen vehicle comms stuff I worked on I was in the middle of it all, we worked as part of a team, including acoustic engineers so maybe I might have seen and done more than you presume, and done under proper testing conditions...
And the same sh** build some audio stuff etc.
.Sorry but you don't know how involved I get in projects sorry....
Neither do I just do Mil stuff, we do a whole lot of stuff including sensitive analogue.
After the last insult like this all I can say is it is a pathetic way to carry on a discussion, presume the other person has no clue about what is going on and thus insinuate his comment are invalid and have no worth. Happening more and more on here and rather unexpected from you.
Just because I disagree....Tear me down...
I have suggested some things that are relevant and worth considering you have decried me at every point because you don't like SMPS supplies.
Nelson Pass has mentioned in other threads that he uses SMPSs.
I work on site as part of a team so with both headphones and the next gen vehicle comms stuff I worked on I was in the middle of it all, we worked as part of a team, including acoustic engineers so maybe I might have seen and done more than you presume, and done under proper testing conditions...
Last edited:
My last comment as I don't want to play with people who are so condescending, but maybe if you read this....
http://www.firstwatt.com/pdf/art_amp_camp_1.pdf
http://www.firstwatt.com/pdf/art_amp_camp_1.pdf
Hi, thanks for the comments. I will be putting the crossover into a separate box. So I want to find something small to use for it's PSU, that's all. I can certainly build something with linear regulators and a transformer, just thought these smps devices are much smaller.
I like your suggestion about adding a winding to the toroid in the power amp. That's some out of the box in-the-box thinking!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.