Supercapacitor question

[malium]

Hello,

Just a newbie question... I am building a d-class amp that will be powered from 3.7V battery pack (boosted to 5V). A bit current limited, when bass is punching, I am running a little tight on current. Could adding a supercapacitor after a battery source, before an ampiflier input help with peaks? I assume it would. But what size should it be?

Also, could supercapacitor in front of battery help with charging it faster in any way?

Many thanks!

[DutchGert]

Supercap's are not made for large currents, there made for equipment to run on very low currents for years.

A big ceramic (or smaller ones parallel) after the boost converter could help a lot. After the battery probably not because your amp uses the 5V from the boost converter and not straight from the battery.

About charging, no, just change the battery with a charger.

Perhaps u could post your schematic?

[Psi]

Supercap's are not made for large currents,

Tell that to the supercaps in my car that i used to replace the car battery.
They would like to have a word with you about their current handling ability 

It depends on the type of supercap.
Some are, as said above, just for battery backup and small currents.
While others are for storing and releasing a large amount of energy quickly (like regenerative braking in cars. Or boostcaps in large diesel trucks)

Placing a large supercap (of the high discharge type) on your 5V supply rail will give you much more available current for punchy bass.
You should check if it's going to cause problems having large energy storage after the 3.7 -> 5V boost reg. (i'm not sure what type of regulator you have)
You don't want the capacitor to backfeed and damage the regular or try and dump into the battery when the unit is off.

It's quite common in car audio systems to put a 0.5 - 1.0F capacitor across the battery to help boost bass if you have lots of subs.

[PA4TIM]

Those low ESR supecaps give a huge inruch current, for a car battery no problem but if you use it after a transformer it can be a problem.

[Psi]

Yeah, good point.

May need to add a low ohm resistor on the input side, basically adding some more series resistance.
That way it doesn't look like a short to the regulator but can still provides high current on the output side to the amp.


With 5V input this amp doesn't sound like its really powerful, im assuming its using a typical speaker like 2,4,8 ohm?
A supercap really is overkill (nothing wrong with overkill though if you want to do it   )

Some normal capacitors on the 5V rail maybe all it needs.

[malium]

Hello,

Thanks, I was under impression that low ESR, fast charge/discharge could help with fast peaks. It's about 1, 1.5 A draw, going to absolute maximum at 2.1 A when deep bass is played, and then I need something to help - otherwise I'll have to filter it out. So, bigger ceramic would be better or not?

I would put a blocking diode to avoid backfire, of course. Or not?
I suppose constant current regulator would have no problem with a supercap after? This is the cap I thought of using, any comments?

http://184.173.198.29/~slavric/file/repository/specsheetVEC2R5106MG1.pdf

[Psi]

Some regulators have issues with really low ohm loads and/or very large capacitors on the output.
So it really depends on your reg, but yeah, a current limit is a good place to start.

Blocking diodes should be avoided due to the vdrop.
Backfeed may not be a problem, it depends on your reg. (eg, mosfet body diode could be a problem)

[malium]

Some regulators have issues with really low ohm loads and/or very large capacitors on the output.
So it really depends on your reg.

Blocking diodes should be avoided due to the vdrop.
Backfeed may not be a problem, it depends on your reg. (eg, mosfet body diode could be a problem)

Many thanks! Went a bit out of my league there, slow down ... 

Ok, no blocking diode, logical - vdrop. Then, how to stop it from trying to go back? To count on the boost converter to stop it, yes, that should do it - if You ment that, I haven thought of that...

I suppose that for the rest, reg problems with low load, I just need to try

Any comment about the caps I linked at, are they appropriate?

Thanks again!

[Psi]

Those supercaps seem ok, but i would try a generic 4-8000uf 12V cap first.

Those supercaps are only 2.5V so you'd want 3 in series if you're going to put it on the 5V rail.

[malium]

Those supercaps seem ok, but i would try a generic 4-8000uf 12V cap first.

Those supercaps are only 2.5V so you'd want 3 in series if you're going to put it on the 5V rail.

 

[fcb]

I would be tempted to solve the problem at source, measure the amount of extra current you need and then look at increasing the power of the 3v7 to 5v converter.  You already have gobs of storage (the 3v7 pack).

[malium]

I would be tempted to solve the problem at source, measure the amount of extra current you need and then look at increasing the power of the 3v7 to 5v converter.  You already have gobs of storage (the 3v7 pack).

Well, I agree, but it's budget limited, and I would ideally use a 12v battery source, which is too expensive. So I opted for a finished power bank which can host 4x3,7V batteries, and provide about 1000mAH capacity, and 2.1A at the output. It has a boost converter built in...

It's not really a problem until certaint level of battery drainage, when bass draw a bit too much current, and I would just need to solve this. I think the supercup should do it, at amp's input, as a cheap solution...

[PA4TIM]

Check the datasheet of the boostconverter if it likes capacitive loads. In many boostconverters there is allready a diode between in and output.
The cap is rated 2.01 A and 2.5V for normal use so you need two in series and some  bleeders/loadspreading resistors over it to be sure voltage will stay under 2,5V per cap. Adding a resistor is just relocating the problem because you limit the whole psu at all time. You just need a 2.5 or 3A boostconverter.

You can calculate the inrush current (or simulate it) It is just the ESR that limits it, so the first few RC times the psu sees a shorts. Some boost converters then switch off until the short is removed and with a supercap that will be when it is charged so you can wait for ever. (not all boost converters do this)

Also the primair current (from the 3,7V source)  will be much higher. If the rendement from the converter is 80% (just a random number, see datasheet and rendment is often load depending) and current on the output is 2A then that is 2 x 5 =10W, then you need 12,5W on the input. That is 3.4A @ 3.7V. Many people forget this.

[malium]

Wow, thanks for Your patience and effort.

The thing is, I have a will factor, but I lack a bit on a knowledge side. For now, at least. Trying to learn as I go. Beside a day job and other obligations, little time, if any, is left to use for hobby

Now, I will ask, if somebody would be kind to help in more "plain english".

Check the datasheet of the boostconverter if it likes capacitive loads. In many boostconverters there is allready a diode between in and output.

What should I look for, how do You read the liking or not liking a capacitive load?


The cap is rated 2.01 A and 2.5A for normal use so you need two in series and some  bleeders/loadspreading resistors over it to be sure voltage will stay under 2,5V per cap. Adding a resistor is just relocating the problem because you limit the whole psu at all time. You just need a 2.5 or 3A boostconverter.

So, would the easy solution be to get single cap rated for more than 5V? Did not understand the part about bleeders/loadspreading resistors. I can get the ones I linked easier and cheaper, but if it is a pain to construct, I am searching further. Got the part about boost converter, it's logical.

Also the primair current (from the 3,7V source)  will be much higher. If the rendement from the converter is 80% (just a random number, see datasheet and rendment is often load depending) and current on the output is 2A then that is 2 x 5 =10W, then you need 12,5W on the input. That is 3.4A @ 3.7V. Many people forget this.

This is also a bit unclear to me. The power bank I bought (I am waiting it to arrive from China) has the boost converter included in it's circuit. It uses it to boost 3.7V from the battery to 5V at USB outputs. It states that it can deliver 1A at one USB output, 1,5A on the second, or 2.1A combined. So, this should be after the losses, correct?

[malium]

I would be tempted to solve the problem at source, measure the amount of extra current you need and then look at increasing the power of the 3v7 to 5v converter.  You already have gobs of storage (the 3v7 pack).

Wow, I totally misread this. Yes, can somebody tell me please, what is the possible real life current draw from 18650 3.7V battery? I bought this bank:

http://www.ebay.com/itm/370907097414?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1423.l2649

which states maximum of 2.1A output, but can this be modified to get more?

This is how insides look like:

http://www.ebay.com/itm/390667477382?ssPageName=STRK:MEBIDX:IT&_trksid=p3984.m1426.l2649

Sorry for the double post...

[PA4TIM]

If I were you I would forget about a supercap and just use the amp but do not cranck the volume to max. Without more info about your DC-DC converter and amplifier my crystal ball stays dark.

It goes to far to learn you here how to read a datasheet we do not know. That is also more or less basic electronic knowledge.

If caps are in series the total capacitance is only half the value and one cap must "feed" the other. Yiou do not want the top one charge to 3V and the other only 2V because those things protest with a very loud bang. These things are no toys and can be very, very dangerous.
The resistors make it possible to equal the charge but more important, they discharge the caps when power is of and prevent that they charge more as 5V if there is no load.

Your booster can deliver 2A but the batterys must deliver more. They must be able to deliver that. I now understand you bought some ready to use thing so probably it will be able to do so.

Why not make the amp for 12V and use a car or UPS/alarm system backup battery.  Or some other higher voltage battery pack (like used for RC cars etc) without a booster. Those things can deliver huge amounts of power.

Edit:
This is without batterys, the cells are typical 2450 mAh according a panasonic datasheet. And if I understand it well this is a charger for mobile phones with electronics that monitor the output and you can set the cgarge current. I is possible  they are not able to feed an amplifier.

[malium]

Yeah, I am sorry, I messed this thread pretty much. The thing is, I try to avoid asking too much on forums like this, rather search the internet myself or test it in real life. I tried asking on specific area forums several times in past, but usually got ditched in unpleasant manners, so I am kind a scared to ask a complicated questions, just bits that misses to complete the puzzle I am in.

I tried to ask a "more simple" part of the question, although I do not see it simple, about the supercap, cause I have seen this being used in front of an amp somewhere, with good results (as I can read out). Just tried to get a confirmation. Here is the file I saw it in, if somebody wants to take a peek:

http://www.cap-xx.com/news/PumpUpTheVolume.pdf

I did not want to put the whole story or idea, scared I would end up with no answer, as I do not want to exploit Your time or hard earned knowledge, trying to rush over to the end by just using You and Your knowledge, nor leaving the impression that I want that.

I am a guy from many fields, have a strong desire to learn about electronics and it's components, but english is not my best weapon, and half of the terms You guys use is like a gibberish to me, I just can not follow it up. So, to shorten my sad story short, here is the chronology, if somebody wants to follow up and maybe help.

This are the amps I could use:

Ideally, I would use this one:

http://www.ebay.com/itm/281130043213?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649

This is the datasheet of the chip:

http://pdf.datasheetcatalog.com/datasheet/Tripath/mXyzxwwt.pdf

As You can see, it functions from 8.5V, so I was hoping to boost up the 5V battery source to get the desired Voltage, thus loosing some peek current which I meant to compensate with a supercap like they did in above mentioned file. When I see the gain and the limits of such circuit, I would limit the bass output to the level it does not go into clipping (if that is the term for speakers not getting enough current to play a certain note).

After many nights of searching, I opted for the battery pack that is mentioned above, but for the sake of follow up in this big chunk of a post, here it is:

http://www.ebay.com/itm/390667477382?ssPageName=STRK:MEBIDX:IT&_trksid=p3984.m1426.l2649

It is an housing, batteries are ordered separate. I picked this cause everything even remotely close to the needed voltage of ideal 12v costs few hundred dollars with shipping to my country, and that is not acceptable. I know it is not ideal, but I wanted to give it a try, as a last resort. This circuit passes 5V directly to amp (to boost converter, then amp) after it charges batteries (or it should, I asked), and shuts the charger part off, so I wanted to see if this could be used connected to a 5v power adapter when in house, not using batteries, outside using batteries - like in laptop. If somebody have a better solution, please make it, I spent months searching and digging with no luck. I want this circuit to be with two power supplies, diret inside, batteries outside. This power bank also has protections for the batteries, and it was cheap enough to try it. If it will not be good, I will pass it to somebody as a power bank, which it is.

Regarding the boost converter, I have a few, but also would buy an appropriate one, if the one I have is no good. I have this one for example:

http://www.ebay.com/itm/310717070508?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649

Datasheet:

http://www.xlsemi.com/datasheet/XL6009%20datasheet.pdf

So, I hope I did not leave something out, and sure hope I have not insulted anybody. Sorry for the trouble.

[PA4TIM]

No need to appologise, a forum is there to ask questions. If we only had people here with a lot of knowledge and no questions  it would be a dull thing and we would be only left with topics about cheap Chinese scopes and multimeters. 

If you can solder simple things yoursef you could make a boost convetor that can deliver enough current at 12V , That is not hard to do. Then you feed that straight from the batteries and use that phone charger only to  to charge them.http://www.pa4tim.nl/?p=4364 (a junkbox booster but I do not think it works from 3,7V but there are enough comercial ICs that do the job. I made a boost converter with a LT IC that makes 120V from a 1.5V battery.

In that case you do not need a supercap but probably you can use it like they do in the pdf.. You see they use resistors to balance the load. and two in series. And the value they use is not as extreme like you want to use. They use 2 times 550mF in series so a total of 0.275 F and you want to use 2 times 10F so a total of 5F . A lot more. When empty connected to 5V it will try to draw 40A and it takes 2 seconds to come within the range the boosterpack can deliver so that will die or, if you are lucky,  just starts to charge with 2A. If it decides to charge with 2A it wil take around 10-15 seconds before the cap is charged. But only if the amplifier is not drawing current the same time. That charge is drawn from your batterys so only switch the power on only if you need it.

The problem of the setup you now want, is you use a 3,7 to 5V charger/converter to make 5V and then go to 12V. the 8.5 uis the absolute minimum. So double loss.

Clipping is not a speaker that gets not enough current but an amplifier that gets to much input and " hits" it powerrails. It is forced to deliver more power as it should and goes in protection or heats up to much and dies. Using a supercap it is making it more easy to get clipping .http://en.wikipedia.org/wiki/Clipping_(audio)

[malium]

Many thanks on that, You are extremely kind.

Few more things, if I may.

What would be the most efficient way for me to limit and/or regulate the current I allow the amp to draw from the batteries?

Is there a simplish way to use two power supplies, like in laptop? Is there a finished pcb that would allow that, and what would be it's name?

Would it be definitly smart to use some type of capacitor between battery and the amp input?

Thanks!

[PA4TIM]

The simplest way to limit current is not turning the volumeknob to the max.  Like you do not want to run a car engine allways at max refs. This is also save because now if you play to loud without knowing you can not overheat the amp or loudspeakers. And it takes longer before the battery is empty.

The best way is to make the powersupply so that it can deliver enough current. The cheapest and most easy way is to take a 12V lead-accid battery, They are cheap and easy to charge and with 7Ah you can play several hours : http://nl.farnell.com/yuasa/7-12/battery-lead-acid-12v-7ah/dp/2083825

or use a  RC batterypack (but they need expensive chargers), or use the battery from a portable drill (choose it somewhere between 9 and 15V , they can deliver power and are compact).

You can connect the amp straight away to 12V, no problems with unfiltered boost converters that spoil your music and best of all, you can play it longer as 30 minutes. Batterys do not like if you decharge them so much the voltage drops to low, so add a simple voltmeter or comparator and a LED to warn you if the battery needs to be charged. And no need for supercaps, lead accid batterys can deliver more then enough current for such a small amp.

You can build the battery, amp, speakers in a box. Then add a connector for a battery charger and use it with the charger connected in doors and on the battery outdoors.

In our country they call it a crate-radio of case/chest-radio Here is a site that tells you how to do it:  http://www.accusafe.nl/how-tos/how-to-kistradio-kratradio-bouwen/
You only need to make a connector on the amplifier housing and plug in a wire that you connect to the battery box for outside use

[malium]

Hey, thanks again PA4TIM!

The problem is that I am not doing this for my own use, I am building this for someone who is... well, I would like to avoid this persons use of the device in any way than just simple on-off, plug and unplug. Taking the battery out for charging, battery monitoring and such actions should and needs to be avoided by any means.

I have the amp running at home with SLA battery, but again, it needs to be portable, very portable, and easy to charge (plug and play) - not exactly SLA-s qualities. I wanted to use LiFePo4, but it is too expensive including needed protection pcb's,  RC's battery pack tough to maintain and also expensive.

I have to get use of You now, when You are that kind

Please bear with me and confirm my math.

1. Let say the amp is 88% efficient at 8 Ohm load, 10W per channel capable (stereo). Assuming 12V supply. Then, it should be that it uses 1,6A if 100% efficient, which it is not, thus asking for 1,4A as a maximum needed. Correct?

2. For the sake of argument, let's say I can provide a 12V, but 1A supply to it. I now have 0,4 A missing for it to play along even at the loudest setting.

3. By testing, I concluded that the only scenario this comes to show is only at deep bass notes, which is logical.

4. Now, my possible solutions are to add a capacitor in front of the amp input, hoping it would have enough storage to fill in when lacking, which is not often, as the that deep bass is not that much present, and it leaves enough space between to allow the capacitor to get ready for another round. Am I correct by now? I was hoping that supercapacitor would be ideal, cause of fast charge/discharge ability, thus asked here for pointers. If somebody can say how big it should be to play a part here, or confirm I can just use a regular capacitor, please go ahead.

5. Next solution is to limit the output, but then I would like to solve it inside of the circuit, so I do not allow the amp to amplify to the levels that are problematic. I would not want that it is even possible to rise output to problematic levels, so I need a solution within. I do not like this solution a tiny bit

6. Other solution is to try and filter out the deepest bass notes, so the amp never tries to pull whole 1,4A. I do not like this either.

7. The last solution I can think of, is that I go with 9V and 1,5A, which is enough and possible with the exact same circuit, but then my lack of knowledge comes to the scene big time. I know this will limit the output, which is less important and I can deal with that on other fields, but how about the battery? This battery packs has some kind of a constant current boosters right? So they supply the rated output of 9V till the drain of the battery? It does not come lower than 9V?
Cause the amp will run at 9V, but it will not at 8V...

Any comment is welcomed and payed with a virtual beer. Or an apple, if You prefer virtual fruits.

[rsjsouza]

Just to keep the fun on this thread... Check the Youtube video about supercaps

[Zero999]

It's quite common in car audio systems to put a 0.5 - 1.0F capacitor across the battery to help boost bass if you have lots of subs.

Hmm, I've heard about that and it sounds like audiophoolery.

What's the impedance of a car battery? How much current is required to start the car?

To get any more than 36W per 4 Ohm load you need a boost converter which should deliver a relatively constant voltage so any super capacitor would be best placed across the output of the power supply.