diyAudio (
-   Headphone Systems (
-   -   O2 LiPo Daughterboard (

ryuji 10th June 2013 12:31 PM

O2 LiPo Daughterboard
I do not have the parts to breadboard prototype with yet, only battery is ordered so far. ~2 weeks to get that. Free time lately is a luxury so progress into this project may be slow.

Im designing a daughterboard for the o2 that will run on a single 2000 mAh LiPo cell, and still hopefully leave room for the odac to fit.

Lets start with the battery i chose:
3 7V 2000mAh Lithium Battery Rechargeable Polymer Li Po for Nav MP3 GPS B 853460 | eBay

The dimensions are almost perfectly sized for fitting inside the 'battery compartment' and it has a built in protection IC so i dont need one.

MCP73841-420I/UN Charge management controller is fairly self-explanatory:

I will be running a inverting boost converter set at +-15v output, then using the O2's regulators to drop it to normal 12-ish volt operating voltage. this should filter out any switching noise etc etc. I will bypass the mosfet low voltage cutoff as it is not useful anymore when being fed with a boost converter. Might need to use two separate boost converters to get the current i need instead of an monolithic one

When/if i get to the stage of installing the odac ill rig it to charge off of either usb or the (converted to 12vdc) external supply.

The only thing thats going to be annoying is the battery managment ic's are all surface mount. I am having difficulty finding a high-ish current inverting boost converter in dip format, so i ultimately need to make my own pcb design. 'Dead Bug' until then :p

I have a friend that offered to help me hand etch a pcb but if there ends up being at least a few people interested (maybe with a protection IC included in the design) i would get the pcb formally run off and sell enough to make up my additional costs.

gastro54 10th June 2013 06:28 PM

You are going to have significant IV losses through the + - 12V regulators with that configuration. (15V - 12V) * (24mA) * 2 = 144mW that you're dissipating at idle. That's the listed DC idle current using just the batteries that doesn't take into account the quiescent current of the regulators. All of this gets worse once you actually start driving a load. It should also be noted that these losses are absent when using the dual 9V batteries since the regulators are not needed.

There will also be switching losses from the boost / inverting converter. Do you know what the conversion efficiency will be for the positive and negative rails (each one is different) over the expected range of current given your 3.2-4.2V LiPo input? I'd guess that 80% is optimistic.

Lets estimate that the 3.7V * 2000mAh battery gives you 7400mWh of energy. 7400mWh * 80% = 5920 mWh.

Lets estimate that the two default 9V * 250mAh * 2 batteries give you 4500mWh.

Just to be able to match the dual 9V batteries, the LiPo solution needs to be able to run for 9hrs. 9h * 144mW from before is 1300mWh, and that's at idle! 5920mWh - 1300mWh= 4620mWh. That 1300mWh that you turned into heat at idle is going to blossom into a much uglier number when you start driving something. As you can see, there are big issues here that need to be addressed if you want the LiPo powered solution to be advantageous over two 9V batteries.

ryuji 10th June 2013 11:34 PM

you have a good point on the dissipation losses on the regulators, I had switching noise in my head and not efficiency when i considered doing it that way. Probably much better off with outputting 12v directly. A bunch of the boost converters i was looking at earlier were claiming over 88% efficiency but that is probably for low current needs.

Never thought it would be a easy project. Thank you for the well thought out critique

gastro54 11th June 2013 12:13 AM

To minimize the power wasted I'd recommend putting LDOs (low-dropout-regulators) on your board after the boost / inverting controllers as you will still need a way to prevent switching noise from getting into the audio power rails. Set the output of the DC-DC converters to just above/below (V_out +- V_dropout) of the positive and negative LDOs. The linear regs used in the O2 have dropout voltages of 2V and 1.3V which is horrible for battery powered-usage. This way, you'll minimize the power wasted in the linear regulators since the voltage drop across the regulator can be made very small with LDOs. One thing to watch out for in this solution is the turn on transient. If you power everything on at once, you'll get uneven rise of the positive and negative rails, which can be dangerous for anything connected to the outputs. You'll need to figure out a way to delay activation of the LDOs until the voltages coming out of the DC-DC converter(s) have stabilized.

ryuji 11th June 2013 12:18 AM

I noticed that when i was simulating some inverting controllers in spice, quite a bit of time difference between settling of each rail. simplest but not the best solution is perhaps to leave the dead battery protection circuit in place, this would prevent premature powerup and thus dc offset?

gastro54 11th June 2013 12:47 AM

Yup - the battery protection circuit would help a lot with the turn-on transient.

ryuji 12th June 2013 01:18 AM

2 Attachment(s)
I think i found a controller that is suitable. it even was designed to permit configurable undervoltage lockout which would remove the need for a battery protector in all cases.

ryuji 12th June 2013 05:39 AM


with such a high switching frequency, i somewhat doubt it will be necessary tho. ill see how it goes without in prototype unless its strongly recommended against. Might also be useful to use potentiometers in place of feedback resistors so its possible to dial in the voltage you want

ryuji 12th June 2013 11:20 AM

put in order for rest of the parts today including the ldo regulators. ill update the thread when i get all the items and get time to work on prototyping

Amperage 14th June 2013 07:37 PM

Losses on these designs are definitely something to consider. I've built a number of headphone amps using boost converters to generate a true +/- rail and have managed to maintain 10hour+ run times.

My most recent amp is using a TPA6120A2 headphone output driver with a MC33078 used as a input buffer. The amp is using a 2200mah 3.7v li-po. Pretty close to the spec you chose and I'm achieving 15+ hours avg. run time by employing a simple feedback circuit from the headphone outputs to vary the boost converter voltage output. +/- 6v until the positive voltage swing at the headphone jack reaches 3.5v then the circuit changes the boost converter's feedback loop to push the output to +/- 12v.

Unless you are driving lower ohm headphones pretty loud it never "shifts" out of the lower voltage range and nearly halves the idle current draw until the extra headroom is required. Higher ohm loads would of course coax it out of the low voltage output mode more often.

As far as the LDO's following the boost converter. I generally run between 1.2 and 2mhz switching frequencies then filter some of that high frequency noise with ferrite beads. Some of it makes it through to the headphone output but it's a very low level. less than a couple millivolts of ripple.

The boost regulator I've been using for the higher powered amplifier designs is the LT3579. datasheet links below.

The idle current draw at the battery on the TPA6120 based amp is 98mA with the boost converter at +/- 6v and 172mA at +/- 12v. With no load on the boost converter output power consumption is 11mA so the converters are pretty easy going on quiescent power consumption.




This is the battery charge/management chip I use. It's a Qfn-16 package. not all that diy friendly if you aren't used to small surface mount work


All times are GMT. The time now is 04:23 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio