electro-music.com   Dedicated to experimental electro-acoustic
and electronic music
 
    Front Page  |  Articles  |  Radio
 |  Media  |  Forum  |  Wiki  |  Links  |  Store
Forum with support of Syndicator RSS
 FAQFAQ   CalendarCalendar   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   LinksLinks
 RegisterRegister   ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in  Chat RoomChat Room 
Live streaming at radio.electro-music.com

  host / artist show at your time
  Twyndyllyngs Live tonight Chez Mosc
Please visit the chat
 Forum index » DIY Hardware and Software
Switching power supplies
Post new topic   Reply to topic Moderators: jksuperstar, Scott Stites, Uncle Krunkus
Page 2 of 2 [48 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
Goto page: Previous 1, 2
Author Message
halfshavedyaks



Joined: Jul 12, 2018
Posts: 65
Location: UK

PostPosted: Sun Jan 20, 2019 12:56 pm    Post subject: Reply with quote  Mark this post and the followings unread

Here's another idea.

start with a 12v supply like Grumble is using - or 2 single 12v supplies as at the beginning of this thread.

turn them up to approx 12.8-13v.

use a LM2940 regulator (with a VDO of 500mv)

LM2940 is only 1A but heatsinking is easier because the VDO is so low.

Presumably a pass transistor could be added to a LM2940.

there is even a negative version - LM2990 although those are a bit harder to find and more expensive they are out there.

this would be super efficient as there is minimal waste in the regulator. BONUS you can run your synth (or whatever) from 2x12v batteries and a solar system. Useful after the apocalypse or other off grid situations.

Are there downsides? I don't know.

Are there pitfalls running a pass transistor with a LM2940? I don't know.
Back to top
View user's profile Send private message
halfshavedyaks



Joined: Jul 12, 2018
Posts: 65
Location: UK

PostPosted: Mon Jan 21, 2019 10:35 am    Post subject: Reply with quote  Mark this post and the followings unread

I found a site with good prices on meanwell supplies in the UK.

example:

https://www.powersuppliesonline.co.uk/rs-50-12-50-4w-12v-4-2a-single-output-enclosed-power-supply.html

I was about to get a couple of the above supplies but I just scored a cheap job lot of used meanwell MDR-60-12.

The advantage of the MDR-60-12 is that the voltage is adjustable from 12-15v, allegedly.

This gives considerable flexibility in regulating the output or not.

I plan to try them with a passive filter and also with LDO regulators.
Back to top
View user's profile Send private message
Grumble



Joined: Nov 23, 2015
Posts: 992
Location: Netherlands
Audio files: 27

PostPosted: Mon Jan 21, 2019 1:30 pm    Post subject: Reply with quote  Mark this post and the followings unread

MDR-60-12 gives 120mV with the output terminated with 100nF and 47uF.
http://www.meanwell.nl/Images/pdf/datasheets/MDR-60-spec.pdf

_________________
my synth
Back to top
View user's profile Send private message Visit poster's website
halfshavedyaks



Joined: Jul 12, 2018
Posts: 65
Location: UK

PostPosted: Mon Jan 21, 2019 3:17 pm    Post subject: Reply with quote  Mark this post and the followings unread

Grumble wrote:
MDR-60-12 gives 120mV with the output terminated with 100nF and 47uF.
http://www.meanwell.nl/Images/pdf/datasheets/MDR-60-spec.pdf


that is typical of switching power supplies of that power. I expect to filter or regulate it in some way.
Back to top
View user's profile Send private message
halfshavedyaks



Joined: Jul 12, 2018
Posts: 65
Location: UK

PostPosted: Mon Jan 21, 2019 3:37 pm    Post subject: Reply with quote  Mark this post and the followings unread

I'd like to know what frequencies of ripple can be expected, none of the datasheets state that.

I note the bandwidth of the noise measurement is 20Mhz though. Chances are much of the ripple is well above audio frequencies. I'm not even sure I'll be able to measure it with my 20Mhz analog scope.

I'm a bit unclear why high frequency ripple is such a problem I would have thought it was easy to filter out. I also imagine that just caps won't do it though it will need an inductor and low ESR caps I would imagine.

I would be more concerned about high frequency EMF interference than ripple actually on the lines. But then I plan on the power supplies being on the floor in a separate (metal) box.

Has anyone used a switching supply and actually definitely had a problem with the ripple?
Back to top
View user's profile Send private message
Grumble



Joined: Nov 23, 2015
Posts: 992
Location: Netherlands
Audio files: 27

PostPosted: Tue Jan 22, 2019 12:16 am    Post subject: Reply with quote  Mark this post and the followings unread

Posted Image, might have been reduced in size. Click Image to view fullscreen.

Looking at this figure might make it clear why high frequency signals are so hard to get rid off.
Capacitors are not just pure capacitance, but also have resistive and inductive components.
Especially the inductive part plays a big role when frequencies are in the MhZ region.
The noise from a switching power supply is mostly high frequency pulses, so the power spectrum of these pulses contain extremely high frequencies and to make things worse if you use a processor in your system, like an Arduino, the clock frequency of this processor might interfere with these switching power supply noise resulting in an audible product of this interference.
So whenever using a processor on a module, always use capacitors with a low capacitance with good high frequency behavior ánd use inductances like ferrite beads close to the power entrance of the module.

_________________
my synth
Back to top
View user's profile Send private message Visit poster's website
halfshavedyaks



Joined: Jul 12, 2018
Posts: 65
Location: UK

PostPosted: Tue Jan 22, 2019 2:46 am    Post subject: Reply with quote  Mark this post and the followings unread

that's a great graphic. It confirms what I already had in my head but it is much clearer and more specific.

Parallel caps at 1% of the value of the previous one are a trick I have used before.

Inductance in caps is a problem for sending HF ripple to ground. But inductance in series with the load should block HF ripple quite effectively. It just needs to be inductors with thick wire and low DC resistance to avoid sag.

When you look at the noise/ripple from your SMPS do you see identifiable frequencies? If so what frequency?

If using a regulator after the SMPS I suspect that regulators vary a lot in how well they reject HF ripple. It is very different from 100hz transformer ripple.
Back to top
View user's profile Send private message
Grumble



Joined: Nov 23, 2015
Posts: 992
Location: Netherlands
Audio files: 27

PostPosted: Tue Jan 22, 2019 5:20 am    Post subject: Reply with quote  Mark this post and the followings unread

Quote:
When you look at the noise/ripple from your SMPS do you see identifiable frequencies? If so what frequency?
I don't know, have to measure it. My scope just calculates the rms value of the noise.
Quote:
If using a regulator after the SMPS I suspect that regulators vary a lot in how well they reject HF ripple. It is very different from 100hz transformer ripple.
It is best to remove the HF noise as much as possible before the regulators and be sure to have enough headroom, that way you don't have to worry about that.

Ripple Rejection RR: f = 120Hz, VO = 15V to 25V typ. 71.0 dB
Output Resistance rO: f = 1KHz 18 mΩ
for a typical LM7812

_________________
my synth
Back to top
View user's profile Send private message Visit poster's website
Grumble



Joined: Nov 23, 2015
Posts: 992
Location: Netherlands
Audio files: 27

PostPosted: Tue Jan 22, 2019 5:24 am    Post subject: Reply with quote  Mark this post and the followings unread

Quote:
Inductance in caps is a problem for sending HF ripple to ground. But inductance in series with the load should block HF ripple quite effectively. It just needs to be inductors with thick wire and low DC resistance to avoid sag.
It is even possible by choosing the wrong values for components used to filter the power lines that you make things worse by introducing ringing in your power supply line(s), but how to avoid this is not clear to me.
_________________
my synth
Back to top
View user's profile Send private message Visit poster's website
halfshavedyaks



Joined: Jul 12, 2018
Posts: 65
Location: UK

PostPosted: Tue Jan 22, 2019 5:33 am    Post subject: Reply with quote  Mark this post and the followings unread

[quote="Grumble"]
Quote:
It is best to remove the HF noise as much as possible before the regulators and be sure to have enough headroom, that way you don't have to worry about that.


exactly. AND if the filter can be made with low enough DC resistance a regulator may not be needed.

[quote="Grumble"]
Quote:

Ripple Rejection RR: f = 120Hz, VO = 15V to 25V typ. 71.0 dB


this ripple rejection figure is for 120Hz mains ripple from a transformer. I would guess a 7812 might be much worse at rejecting HF switch mode ripple.
Back to top
View user's profile Send private message
halfshavedyaks



Joined: Jul 12, 2018
Posts: 65
Location: UK

PostPosted: Tue Jan 22, 2019 5:35 am    Post subject: Reply with quote  Mark this post and the followings unread

Grumble wrote:
It is even possible by choosing the wrong values for components used to filter the power lines that you make things worse by introducing ringing in your power supply line(s), but how to avoid this is not clear to me.


yes I am trying to work that out. I suspect trial and error is easier than making the calculations. It is important to know the frequency you are trying to filter out.

this looks like a useful article, I am trying to understand some of it now.

https://www.murata-ps.com/data/apnotes/mpan-01.pdf
Back to top
View user's profile Send private message
halfshavedyaks



Joined: Jul 12, 2018
Posts: 65
Location: UK

PostPosted: Tue Jan 22, 2019 5:52 am    Post subject: Reply with quote  Mark this post and the followings unread

I think as long as the self resonant frequency of the inductor is well above the frequency you need to filter out then it should be OK.

I was looking at these inductors:

https://www.murata-ps.com/data/magnetics/kmp_1400.pdf

I'm assuming that the switching freq of the PSU is in the range 50-200khz. so an inductor with a SRF of 4Mhz or more should be fine. for example 1410454C
Back to top
View user's profile Send private message
Grumble



Joined: Nov 23, 2015
Posts: 992
Location: Netherlands
Audio files: 27

PostPosted: Tue Jan 22, 2019 11:30 am    Post subject: Reply with quote  Mark this post and the followings unread

Quote:
When you look at the noise/ripple from your SMPS do you see identifiable frequencies?

I tried with a 100MhZ digital scope and a 60MhZ analog one, but there was no trigerable signal for both scopes. Just a band of noise.

_________________
my synth
Back to top
View user's profile Send private message Visit poster's website
halfshavedyaks



Joined: Jul 12, 2018
Posts: 65
Location: UK

PostPosted: Fri Jan 25, 2019 10:14 am    Post subject: Reply with quote  Mark this post and the followings unread

The PSUs arrived and I have tested.

with a small load (120R to 40R) and no filter I see spikes 17uS apart which is the main switching frequency of 60Khz.

I also see a lower level wave at about 470Khz.

the spikes level is about 30mv peak to peak

with a simple cap-inductor-cap filter the spikes are down to 12-15mv PTP and the 470khz wave is gone.

so far I haven't come to any good conclusions about what value caps and inductors work best. A wide range had similar results.


the big thing I noticed though is that as soon as I turn on the computer in the same room I get a band of switching noise at a very high frequency and about 20mv that completely drowns out the meanwell PSU noise on the scope and makes it invisible.

the computer has a quality seasonic PSU and is connected to a UPS. I guess isolating the computer more is actually more important than filtering the meanwell supply!

I haven't tried the new PSUs with the synth yet, I have to build some distribution boards first.
Back to top
View user's profile Send private message
Grumble



Joined: Nov 23, 2015
Posts: 992
Location: Netherlands
Audio files: 27

PostPosted: Fri Jan 25, 2019 11:58 am    Post subject: Reply with quote  Mark this post and the followings unread

That sounds promising!
I guess it is equally importand to have noise suppression on each module.

_________________
my synth
Back to top
View user's profile Send private message Visit poster's website
halfshavedyaks



Joined: Jul 12, 2018
Posts: 65
Location: UK

PostPosted: Fri Jan 25, 2019 12:32 pm    Post subject: Reply with quote  Mark this post and the followings unread

Grumble wrote:
That sounds promising!
I guess it is equally importand to have noise suppression on each module.


Yes! I think I will build distribution boards with ferrite beads and small capacitors on every header.

the issue is that the spikes are very high "frequency", it is not a 60Khz sine wave!

so lots of small inductors and ferrites and lots of small caps are the way to go.

fortunately it is easy to make the DC resistance negligible for low inductance values like ferrite beads.

I also made a 22uH inductor with just 4 turns of mains flex through a 25mm ferrite toroid, I will try that too on the main PSU output.
Back to top
View user's profile Send private message
lorencz



Joined: Oct 04, 2015
Posts: 18
Location: Detroit
Audio files: 2

PostPosted: Thu Feb 21, 2019 7:02 am    Post subject: Reply with quote  Mark this post and the followings unread

So, a bit of an update: I've been using two of those cheap, off-brand +/-15v switching supplies in a large case I built about a month ago. It's four rows of 84hp, so current draw was a big concern for me, which is why I opted for switching supplies rather than transformers. It seems like when you get to a certain threshold for current, switching supplies start to make a lot of sense economically, since beefy transformers are pricey. Anyway, I've got one LM7812/7912 regulator board hooked up to each supply. It's the same type of board I would use with a transformer, I just bypassed the diodes. (I had the boards lying around from old cases, which is why I didn't just make a new regulator board.) So, current is limited by the regulators to about 1amp per rail, per supply. The boards have four big 3300uF caps on them, plus the smaller caps for the regulators. The power then goes to my distribution boards which I designed with spaced out 10uF caps and 100n caps.

Anyway, to cut to the chase, I've had zero problems with this setup, and I actually suspect I could get away with two 12v supplies in series, without regulation. The biggest contributor of noise to my signal path is actually the busted USB audio interface I was using. 🙂

At the end up the day, for small systems, I'd probably just go with a traditional transformer, for simplicity, but switching supplies have been a dream for my current setup.

The other place I'd consider using switching supplies is a small skiff where space is at a premium, but I'd probably use something smaller than the knock-off meanwell supplies.
Back to top
View user's profile Send private message Visit poster's website
halfshavedyaks



Joined: Jul 12, 2018
Posts: 65
Location: UK

PostPosted: Thu Feb 21, 2019 8:59 am    Post subject: Reply with quote  Mark this post and the followings unread

A quick update from me too.

My rack seems to run fine from my 2 5A meanwell PSUs no discernable ill-effects. No regulation, just some passive filtering.

BUT I haven't recalibrated my oscillators or anything like that yet and I still haven't built a box for the PSUs or generally optimised the setup so it's a bit early to say for sure.

My impression is that filtering is much more relevant than regulation. Most regulators won't eliminate PSU switching spikes and the PSU outputs are already well regulated in other respects.

My experiments with putting the scope on various parts of my music system suggest that other parts of my setup unrelated to the synth (computer, UPS, clock, usb power supplies) are much worse sources of PSU switching noise than the PSUs for the synth.

So I have bought a lot of ferrites suitable for mains power cables and whatever else is connected to switching power supplies in the room and I'm gradually deploying them as convenient. Results unknown so far I haven't finished and haven't had time to check for results.

My feeling so far is that unless you are obsessive about eliminating switching tech from every aspect of your studio (and indeed your whole house) then running a synth from a couple of these is not a problem, or at least is lost in the general noise.
Back to top
View user's profile Send private message
PHOBoS



Joined: Jan 14, 2010
Posts: 4440
Location: Moon Base
Audio files: 625

PostPosted: Thu Feb 21, 2019 9:37 am    Post subject: Reply with quote  Mark this post and the followings unread

good to hear it is working so far Very Happy

could it be that switched supplies are more susceptible to mains noise from other devices ?

_________________
"My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider
Back to top
View user's profile Send private message Visit poster's website AIM Address Yahoo Messenger MSN Messenger
halfshavedyaks



Joined: Jul 12, 2018
Posts: 65
Location: UK

PostPosted: Thu Feb 21, 2019 10:38 am    Post subject: Reply with quote  Mark this post and the followings unread

PHOBoS wrote:
good to hear it is working so far Very Happy

could it be that switched supplies are more susceptible to mains noise from other devices ?


The noise from other devices is visible even with the synth switched off. Though I suppose that doesn't prove much.

Also there's no basis for comparison - I haven't checked for noise from other devices when the synth is connected to a linear PSU, I suspect it is much the same.

None of this noise is directly audible, at worst it will have some sort of subtle secondary effects that are, but it also may not.
Back to top
View user's profile Send private message
Grumble



Joined: Nov 23, 2015
Posts: 992
Location: Netherlands
Audio files: 27

PostPosted: Wed Feb 27, 2019 3:29 am    Post subject: Reply with quote  Mark this post and the followings unread

Quote:
Decoupling caps (or 'filter caps' as some producers incorrectly market them*) are the most beneficial when actually next to the IC's they're meant to be affecting, they are local energy storage . If a module needs decoupling caps the designer would have already placed them on the board, next to an IC, and wouldn't be relying on a cap 12 inches away at the other end of a ribbon cable.

The other reason a cap may be needed between a PSU and load is to help setup a regulator correctly. unless the regulator is on the busboard there is no need for their output caps to be on the busboard.

Adding extra caps to a busboard doesn't increase it's usefulness, it just adds to the amount of inrush current a synth is going to draw on startup, and may cause your power supply to trip.

*in order to be a filter, a cap has to be combined with a resister, to form an RC network.

This is something I read on another forum after I noticed there were no caps on a busboards.

_________________
my synth
Back to top
View user's profile Send private message Visit poster's website
maltasynth65



Joined: Jul 20, 2016
Posts: 9
Location: MALTA

PostPosted: Thu Feb 28, 2019 8:19 am    Post subject: switchers and all that...... Reply with quote  Mark this post and the followings unread

hey there......just to straighten out things...... Smile single rail switchers CAN be wired together to provide a bipolar psu. In fact, in other forums, I have encountered posts moaning and groaning about the interactivity between rails on tr- and dual-rail single switchers. Also, these kind provide very different current capacities on their rails, and are known to have current demand related voltage fluctuations. Single-rail switchers do not have these issues and are totally independent from each other. They also provide plenty more power per rail, and their cost is never-you-mind similar to the duals. One very important factor which usually is totally ignored is the fact that a decent psu is only part of the equation, and that there are at least two other factors which feature greatly when it comes to building a trouble-free, stable power distribution system for your monsters. The first is the fact that there are module designs and there are module designs. Some can be (are) dependent upon supply stability for their advertised performance, and any significant power supply variations would reflect badly upon their stability. There is nothing much to do if such is the case, except to either replace the module with a similar but more power-supply agnostic module, or else make for a better power supply system. The other factor is the distribution hardware delivering power from the switchers to the modules themselves. A solid, good-specc'ed, well-laid out network coupled with the observance of simple rules can go a long way towards a trouble-free modular supply.
Back to top
View user's profile Send private message
rst



Joined: Apr 26, 2019
Posts: 3
Location: Dublin, Ireland

PostPosted: Wed Jun 05, 2019 9:14 am    Post subject: Reply with quote  Mark this post and the followings unread

halfshavedyaks wrote:
I found a site with good prices on meanwell supplies in the UK.

example:

https://www.powersuppliesonline.co.uk/rs-50-12-50-4w-12v-4-2a-single-output-enclosed-power-supply.html



Reichelt too has reasonable prices for Mean Well:

https://www.reichelt.com/ie/en/switching-power-supplies-50-4-w-12-v-4-2-a-mw-lrs-50-12-p202961.html?&trstct=pol_5

(€9.95 delivery though)

Additional 5V output (but only 0.5A on -12V) in this one:

https://www.reichelt.com/ie/en/switching-power-supply-triple-output-64-w-5-12-12-v-mw-rt-65b-p147885.html?&trstct=pol_0

While browsing Mouser came across this external Mean Well GP50A13D-R1B version that might be interesting for those not wanting to play with mains wiring:

https://www.mouser.ie/ProductDetail/MEAN-WELL/GP50A13D-R1B?qs=sGAEpiMZZMs1jjUfAXmXyuvswF%2F8u6HwWO1S%252B%2FchfRrNw0x6efNuhQ%3D%3D

(€20 delivery - free for orders over €50)
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic Moderators: jksuperstar, Scott Stites, Uncle Krunkus
Page 2 of 2 [48 Posts]
View unread posts
View new posts in the last week
Goto page: Previous 1, 2
Mark the topic unread :: View previous topic :: View next topic
 Forum index » DIY Hardware and Software
Jump to:  

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum
e-m mkii

Please support our site. If you click through and buy from
our affiliate partners, we earn a small commission.


Forum with support of Syndicator RSS
Powered by phpBB © 2001, 2005 phpBB Group
Copyright © 2003 through 2009 by electro-music.com - Conditions Of Use