Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Looking for help on an idea - IPod powered amp!
New Posts  All Forums:Forum Nav:

Looking for help on an idea - IPod powered amp!

post #1 of 9
8/16/07 at 1:35pm
Thread Starter 
Hey guys,

I've been thinking for a while that this would be a wonderful project to build, but unfortunately, I never allocated enough time to get it passed planning.

The idea is a portable amplifier powered directly from the IPod dock. The amp is to be TINY and perhaps would even have the dock connector built into the board.

The IPod gives 3.3V on Pin 18, Ground is common Ground (Pin 1 and 2), they are also connected to Audio LineOut ground.

All of my background on very low voltage headphone amps and single power supply amps comes from Sijosae's very low voltage amps described here and here.

Both of these amps use the TDA1308A Class-AB headphone driver, and the battery life would be wonderful, but perhaps an opamp would sound better, let's say the AD8397 which should run beautifully at 3.3V and +/- 3.3V. For the wonderful performance of this opamp at low voltages, I wouldn't mind sacrificing a little battery life on it

Now onto the power scheme, I don't think virtual ground is possible since IPod V- is tied to Audio Ground. We can either use output caps and tie output ground to V- or we can use the LM2664 to get ourselves an equal V- at around 90% efficiency. I don't know about the performance of this chip or how badly it could mangle the psu noise, but it's worth a shot since I really wouldn't want to use output caps (size requirements mostly).

The LM2664 can offer 40mA of current, this should be enough for the AD8397, but not much more. Because of the high output current of the AD8397 (300mA), would the "ground" be able to suck it up?

I'll try building a schematic of this thing as soon as I have some time, which unfortunately is not too often, but I hope to have one on the weekend. If anybody has a little time to dedicate to drawing this up, I would be very grateful, you'll probably do a better job too since I've never made one before

Have I missed anything trivial that could prevent this from being possible?
Calling all interested DIY gurus to chime in and help out with moving this thing forward.
Reply
post #2 of 9
8/16/07 at 4:41pm
This is a pretty interesting idea, I'm surprised there hasn't been much movement on this in the past. Personally I'd probably try to avoid locking it to a specific DAP device if you want this to be generally useful...but if it's just for personal use...you know how it goes. Because such a small, SMD reliant board will need to be created it's going to be the sort of thing that a proper fab house will be very nice for, and if you can use any device with it, there're more DIYers to share the cost with.

I just poked around looking for other appropriate ICs and found an unusual suspect around here - Maxim. It looks like they've got some interesting headphone amplifier parts that would be perfect for this project - provided they sound okay. For example the MAX4409 is a class AB amp with THD+N under 0.02% up to about 60mW into 16R, 95dB SNR and ~80dB of supply rejection. As an added bonus, it has built-in charge pumps to generate an internal V- and provide a ground-referenced output, so no output caps required. Specs seem better than TDA1308 in every area. A chip worth experimenting with anyway, if you can handle the 0.65mm pitch TSSOP package.

This should be pretty easy to realize. Don't forget that you'll still need input caps even if you reference the output to ground (either with your own inverter or a builtin one). Also keep in mind that +/-3V3 is not going to be enough to drive your high impedance phones to decent levels; you'd need to generate both + and - rails with a switcher, and you'd probably need 'large' inductors to do so (1cm on all sides, at a guess).

Great idea though!
Reply
post #3 of 9
8/16/07 at 5:24pm
Isn't there a Class D headphone amplifier chip out there?
Reply
post #4 of 9
8/16/07 at 5:40pm
Thread Starter 
Quote:
Originally Posted by threEchelon View Post
Isn't there a Class D headphone amplifier chip out there?
Haven't seen many, and I thought Class-D was mostly for speakers. Maxim has the MAX9769.

error401 - I've seen the MAX4410 (actually, I have one of these) and I was playing with the LM2661, but unfortunately, without a PCB, it's really hard to work with it. After a few unsuccessful attempts with dead-bugging this thing, I gave up when I burned up my 5th LM2661.

Although the MAX4409 (almost same chip) should be an easier design to implement, I'm still leaning towards the somewhat better sounding AD8397, plus we can skip the input caps.

What was that amp that used this maxim chip? Little Dot Micro+? Great battery life, not as great performance as opamp based amps.
Reply
post #5 of 9
8/16/07 at 5:48pm
Looking at your sig, you seem to use low impedance headphones except in your home rig, where you probably won't use the ipod powered amp anyway.

So... maybe you don't even have to bother with a voltage inverter chip.
The AD823 goes down to 3v single supply, the AD8066 according to the datasheet down to 5v, but in reality I have succesfully used it down to about 2v (only with quite a bit higher input voltage noise)

The AD823 only provides 16mA output power, not really an impressive figure. But... you could add a discrete buffer. According to some people here, they also go down way low.

Alternatively, you could use a voltage inverter like Sijosae used it - only problem could be the noise produced by the voltage inverter. 6.6V can power quite a range of different amp and buffers - for example the LMH6321 buffer (minimum working voltage 5v)

Eventual noise problems from the iPod and/or voltage inverter could also be conquered by a LDO regulator as a postregulator.
Reply
post #6 of 9
8/16/07 at 6:24pm
Quote:
Originally Posted by FallenAngel View Post
Although the MAX4409 (almost same chip) should be an easier design to implement, I'm still leaning towards the somewhat better sounding AD8397, plus we can skip the input caps.
Probably not a bad idea, AD8397 is fairly appropriate aside from it's high Iq of 15mA per amplifier (30mA is pretty significant waste on a device with only a few hundred mA of battery). This is 6x more than the Maxim chip - but this is Head Fi after all, and maybe it's worth it, I haven't heard this Maxim chip, but it integrates the inverter and has decent specs so it'd probably make a good 1-chip ultra small solution. Maybe not the best small-as-I-can-but-still-audiophile solution though .

I'm wondering how you can skip the input caps though. If you're taking the output from the line out, is it not biased to 1/2Vcc like most DACs would be, or are the output caps actually in the device itself? I would've thought Apple would've skipped these caps internally and put them in the dock, but I don't know for sure.

balou - I thought one design goal was to avoid output coupling caps...but good point on other options that'll run on +/-3V3 supplies, not that I have anything against the AD8397 but maybe you can find something with lower Iq that still sounds good. e.g. the AD823 uses just over 4mA a channel.
Reply
post #7 of 9
8/17/07 at 1:55am
Thread Starter 
The IPod already has output caps on it's DAC, probably cheap **** (hense the IMod's success).

I have personally always been quite fond of AD8397 in the 3 amps I've tried it in, despite it's high current draw of 25mA (average when tested - but up to 30mA by specs).
Reply
post #8 of 9
8/17/07 at 2:38am
Kinda running around in a circle here aren't you?

What is the point of bypassing the internal underpowered crappy output stage and then adding another output stage that suffers from the same power limitations back to the device?


But far be it for me to come along and crap on your idea without offering you something.

http://www.analog.com/en/prod/0,,765...AD8592,00.html

check the chart on the bottom of the page and you may find some class d chips that are "suitable" headphone amps but.....
Reply
post #9 of 9
8/17/07 at 3:19am
Thread Starter 
Quote:
Originally Posted by MisterX View Post
Kinda running around in a circle here aren't you?

What is the point of bypassing the internal underpowered crappy output stage and then adding another output stage that suffers from the same power limitations back to the device?


But far be it for me to come along and crap on your idea without offering you something.

http://www.analog.com/en/prod/0,,765...AD8592,00.html

check the chart on the bottom of the page and you may find some class d chips that are "suitable" headphone amps but.....
Buddy! Glad you could chime in, but can I have the same thing in English now

Specs look nice (as far as I could tell, which unfortunately isn't saying much). Page 10 of the schematic shows "Figure 35. A PC-98 Compliant Headphone/Line Out Amplifier With Gain" which looks quite simple but we still run into output caps which I was hoping to avoid. There is mention of input caps as well, not sure about those though.
Reply
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Looking for help on an idea - IPod powered amp!