[smrtby123]

I have most of the parts ready to build the Apheared 47 amp for my set of Grado SR80's, but I am debating what to do about power. The tutorial by Feist has it powered by a wall-wart, but I wanted to make mine portable. A friend of mine told me to use AA NiMH or NiCd instead of 9V because of the lower output impedence, but is there another way to power this with the 18-24V needed so I don't have 12-16 AA's crammed into a case?

Edit: Also any suggestions for an enclosure that would fit that many batteries

 

[jcx]

the choice today would be to use a op amp with enough output current and rail-to-rail swing like the AD8397 - then a single 9V gives listenable levels with Grados

of course willingness/tolerance for hand assembly of surface mount components is required

 

[Fred_fred2004]

What about a discrete rail splitter they handle the current and you could use 9volts, or use two rechargable 9volts and trickle charge them overnight

 

[smrtby123]

I was planning on using the standard burr-brown op-amp and - wanted to avoid 9 volts because the capacity will be lower. Using that many AAs will have to suffice I suppose.

 

[jcx]

the A47 scheme isn't attractive with limited supply V - it is limited by least of the op amp's common mode input range and output V swing capability and there is extra V drop in the summing resistors

for battery power you should really look at op amps designed for rail-to-rail output

the OPA132/4 waste 2 cells worth of V from Each rail to get to even 30mA output current

if you want to stick with thru-hole parts there are fewer options, none as good as AD8397 at supplying large current near the supply V

AD823 would be a OK thru-hole battery power candidate - with >5V supply it will give more current closer to the rail than the OPA

 

[smrtby123]

Ok few more follow ups:

1. So would I be better off doing a simple Cmoy with a different op-amp like the AD8397 with a TLE2426 instead of the OP2134?

or

2. Should I construct a discrete rail splitter like Sijosae's (HeadWize: DIY Workshop > Apheared 47 with a Discrete Rail Splitter) and still use the OP2134 amps in the design?

3. Could I power both versions with a simple 9V?

 

[jcx]

to split a battery supply for Grado low impedance cans you need much more current than the tle2426 can sink

the mini^3 project is one solution The Mini³ Portable Stereo Headphone Amplifier
extensive build support thread on Headwize

another option with a stack of single cells is to simply use the center of the battery stack as ground

as an experienced instrumentation design engineer I have several criticisms of the 3-channel active ground hype - the passive approach can be good too at lower parts count and and less supply current drain


if we could decide on a small 4-pin connector for protables then a balanced drive portable would be cool

 

[cobaltmute]

Use AAAs instead of AAs. 12 will fit into a Hammond 1455J1201

That's what I use in my Pimeta:


The yellow at the back taped together is 4x 3AAA 800mAh battery packs. They were $5 each.

 

[smrtby123]

That was the suggestion my one friend made was to use battery packs. So basically what I'm going to have to do is what Apheared did in his original article which is have a dual power supply having a center tapped battery.

If I go with a AAA stack is it still usable with Ben Feists implementation of the A47? ( I have been trying to stick as close to that as possible for my first amp.)
Or should I just build a standard Cmoy with an AAA stack and an AD8397 for higher current output.

btw cobaltmute thats some great work on the pimeta, although your awesome construction is slightly lessened by the bumblebee battery pack lol.

 

[cobaltmute]

The original A47 is a Cmoy using a second op-amp for current gain. Ben Fiest's using a a buffered ground channel (not spec'ed in the original article) for greater ability to sink ground current.

I don't have Grado's, but to my understanding they are low impedance cans that needs lots of current (not a lot of voltage swing). Some here say that you can just use a buffer circuit (no gain) to drive them. I would be tempted to just build a Cmoy with the AD8397 if you want to do it yourself or a Mini^3 if you want a kit.

 

[cobaltmute]

Oh the layout for my Pimeta case comes from tomb. He did his with 2x9v in the back.

 

[smrtby123]

I have been reading a lot about the Mini^3 but I have it on the to-do list, but I wanted to avoid a PCB for my first crack at this. I'd rather ruin a $2 radioshack board than a $12 PCB. I was reading a few other threads and people have said that Feists A47 design is pretty much impossible to run off battery power, and that the A8397 is not simply droppable into a Cmoy with some modifications.
Since the TLE2426 will starve the amp of current could I simply use a buffer circuit for the power with a AAA pack like tangent suggests:

(layout by Tangentsoft)
and than use the A8397 in tangents Cmoy layout? If anything I think I may just go with a tangent Cmoy since it will work with my Grado's I was just trying to get something a little more specific since they can sometimes cause problems with the virtual ground in a Cmoy.