How can I reduce the ripple in a dc circuit?
I tried making speakers but plugging them into my charger causes a buzzing sound.
The cheaper the charger the louder the sound.
I have this bunch of capacitors I can use.
The amp I used is a pam8403, if that helps.
Define 'charger' for me, because I'm envisioning some wall-wart that maybe has a single diode for rectification and maybe no filter cap at all.
You need a full-wave rectified supply with a decently sized filter cap for audio amplification applications. Preferably an LC filter in it to make it as smooth as possible. Don't ask me to draw you one you can google it.
A charger outputs a voltage higher than the battery, and will usually pulse the voltage because this lets the battery get to a higher capacity and is needed to check the battery voltage to prevent overcharging in some designs. This means the equipment you have connected to the battery is jumping back and forth between the battery voltage and slightly higher charger voltage.
Adding a lot of decoupling capacitors will help a little, but the better solution is an LDO voltage regulator between the battery and equipment. The voltage output of the regulator will be constant even with the varying input from the battery/charger. Use an LDO regulator because a switching regulators cause buzzing in audio equipment.
>>923846
Just a standard phone charger. Pic related.
I'm using a tp4046 to charge the batteries, a pair of 18650s.
>>923848
So it's not from the high ripple of a cheap charger?
>>923850
That thing you're calling a charger is just a transformer and rectifier that gives you a 5v DC output. Googling "tp4046" doesn't give anything, but I'm assuming that's your charge controller. That's much more likely to be the source of your noise, so any filtering on the 5v you send to it won't help.
>>923854
I understand that, but saying charger is easier.
Pic related is what I'm using for the batteries.
I do believe it's just a dirty current though, because it's less on my good quality chargers and nonexistent through the step up circuit I'm using to power the amp.
Definitely going to look into >>923848 though.
Power your amp with a battery or batteries. Then you will know for sure if it's a power supply problem. What voltage range are you supposed to be using?
>>923863
I did.
It's fine with my 18650s through a step up circuit.
Not so fine when plugged into a phone charger or any USB port.
It's much less on my original LG charger though, but intolerant when plugged into a computer or a third party charger.
The wave form that you are getting from your charger is causing harmonics that a single cap probably will not fix. Is your goal to run this amp of of USB power?
>>923917
I wish to run it off the battery but also use it while it's charging sometimes.
Also my dad wants me to make him some speakers for his computer, tiny 3w+3w, usb powered.
This buzzing sound would drown out most of the sound from the actual computer.
I've tried pressing a capacitor against the power input on the amp and it helped a tiny bit last night but not enough to actually be any good.
Do you have the PAM8403 on a circuit board? How are you connecting the USB cable?
>>924390
It's soldered on directly.
On my speaker it's connected to a step up circuit, powered by a pair of 18650s. The micro USB goes directly to the tp4056 to charge the batteries.
But I've also got an amp just soldered to a male usb. I was curious about the effects of the buzzing and have 10 amps because they're so cheap so it's not really a loss.
I'm fairly confident the buzzing sound is coming from ripple in the circuit.
How can I smooth that out?
the easy way is to start with <7 volts and use a 5 volt voltage regulator and a couple caps. You need someone that knows more about power supply design than me.
>>924448
You'll need an LC filter to smooth out high frequency ripple, typically a few ten uH and about 100uF. Go for a toroidal core type of design and use a low ESR capacitor. This will connect to the output of your noise source, or in your case the input to the circuit between the wall wart and amp.
Ebay has some ready made but there is a high chance of the capacitors being crap with high esr.
Alternatively try a capacitance multiplier http://sound.westhost.com/project15.htm but the downside to the multiplier is the energy lost as heat.
>>924512
>low esr capacitor
So I can't just get away with a regular capacitor?
Would a lower uf or higher be better?
And I should stick as close to my circuit voltage as possible?
>>924546
Are there any other markings on the back of that one? I might be able to find a datasheet, looks like it came from a PC power supply or motherboard.
>>924546
The higher the ESR the less effective it will be. It also need to be used in conjunction with the inductor.
For the price you might as well try one of these pre-built units. http://www.ebay.com/itm/RTF-LC-Filter-LC-FILTER-3AMP-2-4S-/121652015549
>>923850
>So it's not from the high ripple of a cheap charger?
One way to find out is to try a different power supply, make sure it can supply at least a few amps to rule out any current limiting kicking in.
Those USB chargers can typically do around 1 amp max before the voltage regulation starts to give up.
>>924561
>My family owns a pc shop
Prefect, look for capacitors on the standby rail that will have seen very little load in their lifetime.
ESR is the equivalent series resistance, it is the effective parasitic series resistance of the capacitor and is an undesirable trait. They tend to cost a little more than regular duty capacitors which is why cheap Chinese switch mode power supply's tend to have high ripple.
Panasonic FM or FC series would be an example of a low ESR capacitor.
As for the inductor, can you find any pic related in some junked PC equipment? The output stage of an ATX power supply is a good source for these and around the CPU socket on motherboards.
When used together they form very effective filters for high frequency ripple.
>>924567
Perfect*
Also when salvaging part stay away from capacitors with bulging or leaking tops, they are probably the cause of why the item has been taken out of service.
>>924567
Most power supplies I get are low end generic ones that almost always have horrible ripple suppression. Pic related.
I'm not too sure what an inductor looks like.
Is it just a resistor?
>>924572
Shit I posted the wrong picture. Here are inductors to look for.
>>924572
>>924570
>>924567
Here's a pic from the side if it helps.
>>924573
These?
>>924574
Use the inductor circled in red with any one of the capacitors circled in blue. Inductor goes first in series with the positive wire and the capacitor goes after it with the positive lead connected to the inductor and load, negative lead to ground.
>>924576
Thank you so much anon.
You've helped a lot.
I really appreciate it.
Just final question, in this circuit would the voltage and capacitance affect the effectiveness much?
>>924575
>>924574
Here I made a crappy paint-cad drawing showing how to connect them to your circuit and form an LC filter.
>>924579
Saved.
That's really easy to understand.
I've been saving a couple of TV speakers for a portable boombox, I can't wait to try this out.
>>924580
12 ohm 5w.
Should sound amazing.
>>924578
>in this circuit would the voltage and capacitance affect the effectiveness much?
Of the LC filter or the circuit operation in general? If the former then the larger capacitor would help up to a point then the returns would start to diminish. 1000uF is already way past the point of diminishing returns.
The capacitor voltage rating is just the maximum rating it work on without going pop.
That inductor is probably around 47uH which is the sweet spot for this, you can always remove turns if you get any weird effects.
>12 ohm
I hate to be the bringer of bad news but the pam8403 is designed to drive 4 ohm speakers, 12 ohm will result in very bad performance and may even introduce oscillation as its a class D amp.
>>924581
twelve oam is a rather unusual oamage. cool beans though
>>924585
Would it be possible to get away with a small one like from a power bank?
I have a dead power bank I could take apart.
And the amp I'm using is the pam8610.
Rated for 10w or 12w 4ohm, i don't remember. Should be fine, and if it's not I'll look for another amp.
>>924587
Looks like its around 10uH, so maybe a tad on the small side but overall it should be ok.
>pam8610
Are you powering that from 5v?
>>924586
That's what I thought too.
I'm curious how the amp will handle it.
If it doesn't play nice I'll use it to drive 3+3 3w speakers to increase the resistance and find another amp for these.
>>924585
I can start a new thread after receiving the amp if you're interested, I wouldn't mind sharing the results, I'm curious about how it'll turn out too.
>>924592
>are you powering that from 5v
Nope. I want my amp to work, haha.
I'll put it at max voltage, because of the speakers. I have step up circuits rated at up to 35v iirc, if they can't handle the current I'll run them in parallel.
>>924595
I don't think it's going to work at all on 5v anon as that chip needs 7v to even get its internal oscillator section running. Maybe try using a 12v wallwart to power it instead?
>>924597
>have step up circuits rated at up to 35v iirc
All from that one 5v wall wart? What is the max current rating stated as on the label? Watts = volts x amps so a 5v wall wart rated for 2 amps can put out 10 watts max excluded losses in the boost converter, but I reckon it can do about 1 amp before the output voltage starts to drop.
>>924595
>I can start a new thread after receiving the amp if you're interested, I wouldn't mind sharing the results, I'm curious about how it'll turn out too.
Sure, I'm interested on seeing how it goes.
>>924602
I have a box of these at home, they're rated at 2A but I've had them running at 3A fine.
Voltage drops from around 5.3 to around 5.15v at 3A, and it gets very warm.
If that doesn't work I'll just connect the batteries in series to bump up the voltage. I'd rather not do that though because a step up circuit would provide a more consistent voltage.
Also don't forget the wall wart will only be charging the batteries.
It's the batteries that will be powering the speakers.
If I use 5 batteries in parallel and the speakers need 12w (12v 1A) then each battery will be providing barely over 2w. It'll just be a matter of the step up circuit, how much current it can provide, and how efficient it is.
I have 2 already and it's not expensive to other another one or two if it's not powerful enough.
>>924595
>If it doesn't play nice I'll use it to drive 3+3 3w speakers
Two 3 ohm speakers in series per channel? If so then that would be a better option, it would act like one 6 ohm speaker and this amp would have no problems driving that at low voltages.
Generally the higher the speaker impedance the higher the voltage you need to provide enough swing for a given output power. That's why these low voltage amplifier chips need speakers that are 4-8 ohm impedance, otherwise they just can't drive them well enough to make them very loud.
>>924614
>>924602
This is the step up circuit I'm using.
I tried using it before to charge my tablet, but it didn't want to charge with anything less than 18w. You can see the solder.
Hence why I'll probably use 2 in parallel to power the amp.
>>924614
>It's the batteries that will be powering the speakers.
Ah that sounds more like it, I was thinking you were trying to power a 12v amplifier chip with a puny 5v USB adapter.
What voltage will your battery pack put out?
>>924619
That gives me an idea.
I'll actually give the amp a little more than the rated voltage to see if there's any improvement.
I'll try it at 15v, and then again at maybe 17v to see if I hear the difference.
I am aware it might damage the amp or flat out kill it, but now I'm curious.
>>924620
I've got some of those, both the input and output capacitors turned out to have very high ESR making it run unstable at high loads and give around half a volt of ripple voltage.
To remedy this salvage some of those ATX power supply capacitors and put one on the input and one on the output of the boost board, make sure they are very close to the ones already there.
But make sure the capacitors are rated for the voltage you intend to use them for, rule of thumb is to select capacitors rated for at-least 1.3x the voltage.
>>924622
3.7v would be preferable to maintain a higher mAh, but I think I'll have to go up to 3.7v.
I'll be using 18650s. Another advantage of the pc shop my parents have.
>dead laptop batteries
They're currently being used as a book stopper, so I'm glad I'm finding a proper use for them.
Also had a friend close down his ecig shop recently. He gave me a few 18650s, 18350s, and 26650s. Also a carry bag (vapeking).
>>924627
A higher voltage would make the capacitor less effective, correct?
>>924624
According to the datasheet it'll handle upto 16.5v before going pop but it recommends not going any higher than 15v for good safety margins.
Watch out when stress testing amplifiers as when it fails it could end up shorting the power supply straight through your speakers,this could burn out the voice coil in the process.
Datasheet http://www.diodes.com/_files/datasheets/PAM8610.pdf
>>924630
Running a capacitor on a voltage anything higher than what its rated for will make it explode violently. Skip to 0:20 seconds https://youtu.be/TbvNLoZMDtg?t=20s
>>924632
I meant a capacitor rated for a higher voltage than used being used for.
>that explosion
I'm worried.
And in that case I'll test it at 16v.
Also if the voice coil burns out I'll just repurpose them as passive radiators. Not a complete loss.
>>924635
>I meant a capacitor rated for a higher voltage than used being used for.
Gotcha, capacitors can be safely used with voltages lower than they are rated for. The voltage rating printed on the side of them is just the absolute maximum they can take before going pop.
For example a capacitor rated for 16v can happily be operated at any voltage under than.
>I'm worried.
As long as you don't exceed the voltage ratings on the side of them and don't install them backwards then everything will be fine and dandy.
I see literally thousands of capacitors in my field and have yet to see one explode when used within their limits.
Attached is a photo showing the markings on a capacitor, uF = microfarads and that is the total capacitance value. The larger the uF the more charge it can store.
>>924641
>under than.
under that*
god I need my eyes testing.
So this thread can be summed up as follows;
>Power supply voltage too low or not able to supply enough current
>Speaker impedance mismatch
>Ripple voltage from the power supply or boost converter is too high
>>924644 >>924641
I'm assuming you're replying from a phone. It's easy to make spelling mistakes there. I'd recommend a swipe keyboard, TouchPal is my personal preference.
I'm new to all this diy stuff, it's not as easy as I'd hoped for many things but it's still very fun and I enjoy making things I'll use.
>>924654
I'm on a laptop m8, but I have to admit I had the nice lady's of /gif/ open for the duration of this thread so typing was quite a task.
>I'm new to all this diy stuff, it's not as easy as I'd hoped for many things but it's still very fun and I enjoy making things I'll use.
It can be quite confusing at first, especially since there are so many new things to learn when it comes to electronics. I recommend watching the eevblog on youtube as the 1000's of electronics related videos uploaded there certainly helped me when I first started.
If you're really set on getting into this wonderful hobby then I recommend getting a good bench power supply with voltage and current adjustments, for example http://www.amazon.com/Tekpower-TP1803D-Linear-Digital-Variable/dp/B00EUH18DC
It doesn't have to be that exact one but having a solid power supply with adjustable voltage and current readouts really makes things simpler and allows you to focus on the circuit you are making and not have to worry if its the wall wart causing the problem or if the circuit needs work.
Good luck with the amplifier project!
>>924661
>/gif/
Ah, that explains a lot :')
And thanks for the advice.
Going to test the lc filter on a smaller pam 8403 speaker before using it on my boombox just to be sure I know what I'm doing haha.
>>924664
Will the amp be powered with those boost converters?
>>924665
Yes.
Also know where to find a reliable boost convertor? These eBay ones never work as specified.
>step up 3v to 5v
>rated 2a output
During my test.
>1A output, gets so hot it left a blister on my finger where I touched it
>10% voltage sag
Is dx any good for these stuff?
>>924667
In that case you should first try it with the LC filter between the boost converter and amp like in this pic, the optional extra capacitors can be from that junked atx supply. They don't need to be very big, 47uF to 220uF would be plenty, they'll help counteract any boost converter instability caused by high capacitor ESR.
>>924667
The rating of the Chinese converters should be taken with a pinch of salt. They are typically based on the LM2577 controller which has a switch voltage drop of about 1v (max), so 5v @1 amp of output current is 1.6 amps @ 3v input = 1.6 watts average power dissipation in the switch (built into the controller), the peak currents are about double this still.
You could always add a little heatsink if need be, a scrap piece of metal makes for a nice cheap heatsink.
>Is dx any good for these stuff?
I'm not familiar with what this is sorry.
>>924667
>Is dx any good for these stuff?
Oh dealextreme? I think there pretty much selling the same stuff as ebay sellers, but with perhaps more product support available.
This looks like a more beefier version of your current converter and works down to 3.5v too http://www.dx.com/p/lm2587-high-power-dc-dc-3-5-30v-boost-converter-module-green-155170
It'll still get pretty warm but the added solder on the bottom will help with thermal dissipation, these IC's can continue to work fine at temperatures well above our pain thresholds which is about 40C IIRC. Its not uncommon to find things like this happily chugging along at 80c+, although lower temperatures are preferred for obvious reasons as you recently found out.
>>924671
I cut a heatsink for it, didn't help.
It wasn't just hot enough to hurt, it left a blister on my finger. I thought I could run 2 in parallel to decrease the load on each one so it doesn't get as hot. It melted through the hot glue holding them together.
I'm just saving them for a project that doesn't need much power, the pam8403 amp seems to peak around 800-900mA with heavy bass, I'll most likely use it for them. But I'm never buying one of these again.
I'll look for some relatively low uF capacitors then, most of mine are around 1000uF, if not higher.
I appreciate the drawings, they help visualise what your explaining
>>924692
>I thought I could run 2 in parallel to decrease the load on each one so it doesn't get as hot.
It works in theory but is tricky in practice, you can end up with one of the converters taking more of the load.
Using a volt meter independently set the output voltages beforehand so they are exactly the same, this should make them balance the load pretty evenly. Both units should reach similar temperatures when connected together, if ones getting hotter then its doing most of the work.
>>924676
http://www.dx.com/p/produino-dc-3-5-24v-to-dc-4-0-30v-voltage-step-up-boost-module-red-282902
I was leaning more towards something slightly cheaper like this.
>>924695
That's pretty much the same as what you have now just with a different PCB and layout.
Ebay to the rescue http://www.ebay.com/itm/LM2587-DC-DC-Booster-Converter-Step-Up-Voltage-Regulator-3-5-30V-to-4-0-30V-M96-/291550045047
Might take a while to arrive but for the price I guess its to be expected.
>>924701
>They don't just split the load 50/50?
Not if the set output voltages are different, if one is set slightly higher then it will take most of the load current.
>>924701
Good evening mate ( :
More capacitors will help to only to a certain point (which is probably around a few 100uF), having high capacitance capacitors certainly won't do any harm but again the point of diminishing returns comes into play. It will however lower the overall ESR.
Capacitors are often put in parallel in switching power converters when the ESR of one capacitor is too high, that's the reason why that ATX PC supply has so many capacitors of the same value by the output wires. Its to reduce the ripple voltage (in conjunction with the inductor)
This is what ripple looks like, a perfect DC voltage source would be a completely straight line and ripple is the amount of deviation from this. The best we can do is filter it to acceptable levels with capacitors and inductors, imo for most circuits its good enough to aim for under 100mV of ripple voltage.
>>924711
Should have said that the pink waveform is the output ripple voltage from an AC point of view (part that is changing), the DC level is ignored here.
The yellow waveform can be ignored but its the switching waveform across the controllers internal transistor switch, its job is to pulse width modulate the current through the boost converters on board inductor so the output voltage can remain at whatever its been set to. Again it can be ignored here.
>>924711
>not if the output voltages are different
Damn, how much of a difference? Because with the voltage being adjustable it's extremely difficult to get them exactly the same. Plus I'm using a $5 eBay multimeter which is great for static voltages, but slightly less accurate in this situation. Iirc it was as close as I could manage.
>diminishing returns
Should have anticipated that
An atx power supply is supposed to stay under 150mv of ripple. Usually the generic ones can't handle this.
>>924734
there's a simple and obvious fix: set the output voltages to 5.7V and add a diode in series with each + output.
>>924635
Update to this, I'm thinking about just skipping the batteries and making a desktop speaker now that I've found a 15v laptop charger rated at 5A.
Pic related.
>>924572
i can fap to that layout, desu
it's so fucking well ordered my OCD doesn't even blink
