Battery and charging circuit advice

[mrkva]

Hello,

I have designed a very efficient pocket audio power amplifier. It runs of 12+ VDC and it consumes around 50-60mA in maximum volume.
Now I would like to add a battery + charging circuit. My hope was to make it 'chargable' via USB, but I am really not sure how would one do that. I found ICs designed for USB charging, but not for USB 12V charging, since (I assume) requires DC-DC converter or some voltage doubler or something.

So my questions are:
What kind of battery would you advice? It can weight up to 0,5-1kg and it needs to be 12V
What kind of charging system would you advice? Preferably something that can really do it with 5V/500mA of USB.

Thank you!!
Jonas

[Kremmen]

Wouldn't it be easier to specify a 1-cell LiPo battery that is easy to charge from USB, and power the amp from a simple boost converter that raises the battery voltage to 12V?
At the current level you quote that would not be diffcult at all.
As long as we are talking about Lithium tech, using single cells makes life decidedly easier since you don't have to worry about cell charge balancing  and there are numerous ready made USB charger solutions that you can copy (check Sparkfun open hardware solutions for example).

[mrkva]

Sounds interesting. What kind of voltage boost circuit would you recommend? I need to avoid the noisy converters, since it is audio application. Thank you!

[Kremmen]

I don't have any specific circuit in mind, but all boosters are SMPS circuits so there will be switching noise. At these low currents you could select one with the switching frequency far above the audio range and just ensure that your amp attenuates those frequencies sufficiently to avoid issues with stability and/or operating points. At something like 500 kHz to 1+MHz this should not be a real issue.

[mrkva]

I did measurements yesterday and my amp (also in spec) attenuates at 40kHz, so I think it might work!
Will look into it! Thanks

[Kremmen]

Do you mean 40 kHz is the corner frequency or 0dB gain frequency? Assuming the former and a typical 1st order attenuation of 20 dB/decade, a 500 kHz SMPS is borderline. Provided you have the instrumentation or simulation capability, it would be a good idea to verify gain and phase margin at the intended switching frequency.
Something like this might do the trick http://www.digikey.fi/product-detail/en/LM2623MM%2FNOPB/LM2623MM%2FNOPBCT-ND/483063. It is not by far the only one however and may not be the best one for this app, but just so you get the idea.

[Paul Price]

Use this circuit to convert a 3.7V (4.15V fully charged) li-ion 16850 cell to 12V for your circuit.

You can USB charge this 2200mA rated battery easily with my Super-Duper Charger Circuit below.


http://www.ti.com/product/LM27313-Q1

USB to 12V the easy way

LM27313
1.6 MHz Boost Converter With 30V Internal FET Switch in
SOT-23
General Description
The LM27313 switching regulator is a current-mode boost
converter with a fixed operating frequency of 1.6 MHz.
The use of the SOT-23 package, made possible by the minimal
losses of the 800 mA switch, and small inductors and
capacitors result in extremely high power density. The 30V
internal switch makes these solutions perfect for boosting to
voltages of 5V to 28V.
This part has a logic-level shutdown pin that can be used to
reduce quiescent current and extend battery life.
Protection is provided through cycle-by-cycle current limiting
and thermal shutdown. Internal compensation simplifies design
and reduces component count.
Features
? 30V DMOS FET switch
? 1.6 MHz switching frequency
? Low RDS(ON) DMOS FET
? Switch current up to 800 mA
? Wide input voltage range (2.7V–14V)
? Low shutdown current (<1 ?A)
? 5-Lead SOT-23 package
? Uses tiny capacitors and inductors
? Cycle-by-cycle current limiting
? Internally compensated
Applications
? White LED Current Source
? PDA’s and Palm-Top Computers
? Digital Cameras
? Portable Phones, Games and Media Players

[Kremmen]

For charging LiPo cells, i would suggest a dedicated solution, such as the http://www.digikey.com/product-detail/en/MCP73831T-2ACI%2FOT/MCP73831T-2ACI%2FOTCT-ND/1979802. Very simple with just a few auxiliary components and targeted at LiPo charging. You do know that overcharging LiPo batteries is risky? While again this is not the only workable solution, the advantage is that it removes any uncertainties and needless risk from the process.
If you want to be more scientific, take a look at this appnote: http://ww1.microchip.com/downloads/en/AppNotes/00947a.pdf

[Paul Price]

Everyone has their opinion, perhaps there should be a video made by Dave by on the subject of charging dangers or if there is really a myth that it is dangerous for people to charge Li ion batteries at small currents.

In my opinion and experience I think there have been myths created about battery charging.

Myth: Charging Li-ion batteries is always dangerous.

Truth:

With the very high capacity kilowatt type of  Lithium  cells for automobiles and passenger jets, there is a practical need to charge them at high currents, high altitudes, etc. Charging these kinds of batteries or even using them poses a danger due to the extreme high current involved. These batteries are of a different chemistry and very high current capacity.

These batteries are not the same as the those used in home computers and these are always physically safe to charge at low currents (C/4 or less). The only danger is with over charging or with charging at currents sufficient to create significant heat.

There have been several stories of multi-cell Li ion battery packs overheating and even starting laptops on fire.

This happened only when they were fast-charged at currents approaching that which can create significant heat  and with a defective charging circuit design. If a power supply charging circuit  stresses and overcharges and heats up  the battery, then there cam be a chance of danger.

If danger means explosion and physical injury to personnel, danger implies attempt to charge at at high currents with a defective power supply, in a confined area, especially with many cells, such as in the case of a Li ion battery stuffed into the heat insulated belly of laptop or some other poorly ventilated space, and only then where charging currents are high enough to create a high temperature rise. This can happen, at a a high charging rate, at high currents, at some level approaching a level greater or being to equal to or approaching > .5  mAH capacity of the battery in a heat-insulated area.

In the case of the 2200mAH batteries like the 3.7V used in laptops, I have tried deliberately shorting them out. Under these conditions this causes the batteries quickly overheat and emit gas and fluids and the batteries were permanently damaged, but never any violent explosion. This type of abuse or accident is not the same as charging a battery at low currents.

I have been designing chargers with regulated voltage and current and implemented  chargers into many devices using Li-ion batteries for many years from milliwatt timers to handheld vacuum cleaners and motor driven equipment. When charged at low rates with individual cells or even  small battery packs  at C/4 or less, there never has been a problem with physical"danger".

Charging Li-ion batteries at some small fraction of capacity at a regulated end of charge voltage, such as charging a 2200mAH battery at .3Amp is perfectly safe.

Overcharging at small currents and charging to a level >4.2 volts will quickly or proportionally damage the battery according to the time  (and proportionally higher levels >4.2 voltage) that the batteries are subjected to. This is not charging this is battery this is Li ion battery abuse, and  this abuse, especially in the case with small  capacity overcharge and small over voltage abuse can only damage the battery's usable life time, so this in itself does not cause  physical danger.

The circuit I've suggested limits current and voltage to safe levels, does the same charging as any competent dedicated battery charger ic.  The current is firstly limited by the USB port itself (to some fraction of an ampere) that is used for charging and secondly by the design of the charging circuit itself. Having a "dedicated" battery charging ic would not change the inherent protection provided by the USB port current limiting circuit that makes charging safe.

[amyk]

In the case of the 2200mAH batteries like the 3.7V used in laptops, I have tried deliberately shorting them out. Under these conditions this causes the batteries quickly overheat and emit gas and fluids and the batteries were permanently damaged, but never any violent explosion. This type of abuse or accident is not the same as charging a battery at low currents.

Cylindrical lions have vents precisely to prevent explosion. The electrolyte is flammable, and that combined with the heat, can cause a "vent with flame" event. Under the same condition, pouch lipos will swell and burst, then flame.

(Of course this is if the battery is charged; shorting a truly dead (0V) cell won't do anything.)

[Paul Price]

Amyk, this is what I said in my reply:

If danger means explosion and physical injury to personnel, danger implies attempt to charge at at high currents with a defective power supply, in a confined area, especially with many cells, such as in the case of a Li ion battery stuffed into the heat insulated belly of laptop or some other poorly ventilated space, and only then where charging currents are high enough to create a high temperature rise.

[ealex]

...
The circuit I've suggested limits current and voltage to safe levels, does the same charging as any competent dedicated battery charger ic.  The current is firstly limited by the USB port itself (to some fraction of an ampere) that is used for charging and secondly by the design of the charging circuit itself. Having a "dedicated" battery charging ic would not change the inherent protection provided by the USB port current limiting circuit that makes charging safe.

You can't rely on the USB port current limiting.

I've seen several behavior modes:
- completely cut the power to the entire USB hub chip - all port disabled until a system power down
- cut the power just to that port until the overload is removed
- no limit whatsoever - the usb cable melted in my hand ( the +5V was directly connected to the MB without any fuse, and the cable resistance was high enough not to trip the MB's current limit protection )

Using a dedicated IC is the best way, in my opinion ( small 8 pin SOT with just a resistor to set the charge current and a thermistor input).
I've used them in a small run of hand held terminals charged from a USB port and the only problem i had was related to poor quality cables: the voltage drop at the maximum charge current was over 2 volts.

[Paul Price]

ealex:

If you can't rely on USB current limiting, what good would a dedicated charger IC be to help that situation.

 My charger circuit limits voltage and current and anyway to safe levels, just easier to build with hand parts.

 The only effect I have had with overloading the current output on an USB port was the PC circuit limiting the current or popping up with a message that I was attempting to draw too much current. If I manage to short out the USB power or draw too much current then it has always been in my experience is that the PC shuts down.