Search

Idea Log

a record of ideas I've come up with

Category

Ideas

Idea: Incorporate speaker profiles in computer audio output.

I just read that this article about Apples HomePod speaker performance, and read how the DSP corrects for the actual speakers natural imperfect frequency response.

Ideally you want you audio to have a flat frequency response.  Audio in = audio out at every frequence. Speakers aren’t perfect so this doesn’t happen.  Some are louder at lower frequencies others are louder at mid frequencies, and so on.

What the HomePod does is increases the volume on frequencies the speaker is poor at, so the speakers output has a near perfect frequency response.  This isn’t difficult for a DSP to do if you know the speakers frequency response curve, and this is common thing measured for every speaker.

So my idea is, What if you could enter you speakers “Brand” and “Models number” into your computer and have its DSP auto-correct for your speakers imperfect frequency response.  This should be fairly easy to do.  A database of speaker profiles would have to be created, but after that computer DSPs already accept custom frequency response profiles (bass-boost, vocal, etc)

I have a similar thing with my digital camera and graphics editor.  I can enter my cameras lens in the graphics program and it will correct for my lenses slight distortions in color, geometery, etc.  And this is much more complex than speaker profiles.  Its hard to believe no ones done this.

Some possible challenges.  Speakers distort if volumes are too high volumes.  Having DSPs correcting for speakers poor outputs at certain frequencies  may require the volume to increase at that frequency to make the distortion noticeable.  So these factors would have to be added to the speakers profile.  Such as max volume profile curve to each frequency.  Another factor is the amplifier, or any other stages between computer and speaker.  Audio amplifiers ideally have flat frequency responses, but this is not always true.  So these profiles may want to be added as well.  But these are minor difficulties and easy to correct.

So if anybody want to take up this idea and make it reality, go for it.

Advertisements

Laptop batteries with longer lifespans

I have taken apart a bunch of failed laptop batteries to harvest the lithium-ion/polymer batteries for random projects.  By doing this i have noticed that often only one cell out of many has failed, and often even the failing one can be saved.  Lithium-ion battery cells are considered ‘dead’ if the voltage goes below 2V. “The Li-Ion cell
should never be allowed to drop below about 2.4V, or an internal chemical reaction will
occur where one of the battery electrodes can oxidize (corrode) through a process
which can not be reversed by recharging. If this occurs, battery capacity will be lost
(and the cell may be completely destroyed)”.[1] So if one cell becomes dead the whole battery pack is useless.  If there are 4 cells in the battery pack, and one cell is bad, 3/4 are still useful.   I think most users would be happy if they could continue to use a battery with 3/4 the charge, rather than having to buy a new battery pack each time one cell dies.

My idea is two parts.  First is attempt to recover the ‘dead’ cell.  Second is bypass the dead cell.

Idea one:  Recovery a dead cell. I have found that i can recover most dead cells, by applying a small current until the 2V is reached, then charging like normal.  I am no expert on the chemistry of the cells, by lithium-ion batteries are volatile so this may be dangerous.  I havnt had any problems but there could be a good reason cells below 2V are considered dead.  To impliment this, it wouldnt take to much circuity to add this to a battery pack.  Adding a small increase in parts cost and probably no increase in battery size.

Idea two:  Bypass a dead cell.  There is no reason a dead cell couldn’t be bypassed.  This would lower the max/min voltage of the pack.  for example a 4-cell pack has a max voltage of 16.8V(4.2×4), average voltage of 14.8V(3.7×4) and a min voltage of 12V(3×4).  With one dead cell this would cause the voltages(max/ave/min to be 12.6V/11.1V/9V.  If you applied this to the same laptop it probably wouldn’t work, however this can be solved using two methods.  First and most effective is to have the laptop accept the lower voltages when a cell is bypassed.  When the power comes out of the battery if goes through a DC-DC converter to lower the voltages to usable levels(usually 12V/5V/3.3V) for the laptops devices.  If that converter is designed to accept voltages of a wider range, it can easily use batteries with bypassed cells.  Another method for allowing the use of bypassed cell is putting a DC-DC converter in the battery pack.  This would make the pack larger and less efficient but would allow the pack to be used on an unmodified laptop.  Ether way the bypass circuit would increase the size of the battery pack’s circuits and the cost of added parts.  The circuity accounts for about 1%(estimated) of the battery packs volume.  The bypass circuitry would probably double the size of the circuity.  Adding a DC-DC converter to the pack would increase the size by 10%(guess), and lower the efficiency by 5%(guess).

Using ether of these ideas will increase the battery packs lifespan significantly saving users a lot of money, and hassle with dead batteries.

20150515-1234-5158
Example laptop battery.  Although there are 6-cells they are actually 3- sets of 2 cells in parallel, so with respect to voltage, it is the equivalent of 3-cell

 

References

[1] http://www.ti.com/lit/an/snva533/snva533.pdf

Blog at WordPress.com.

Up ↑