Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Suitable portable amp for B&W P5's
New Posts  All Forums:Forum Nav:

Suitable portable amp for B&W P5's

post #1 of 5
Thread Starter 

With the P5's being only 26 ohms impedance, I understand they will require large current to drive them (similar to Grado's?).


For this reason, I decided to upgrade my TLE2426 CMOY power section based on Tangents site. I was planning to use the TLE, BUF and an AD817.


That's when I started looking at the A47. The layouts I've seen use a similar buffered supply; minus the opamp. Does this reduce potential current output?


I note that the A47 uses two opamps in parrallel in the audio section however; is this achieving a similar effect? I wanted to use 2xAD797BRZ in my audio section, so didn't fancy needing to shell out for 4 if I can make do with 2 for the audio section (plus the AD817) in the power section.


i hope that makes sense?

post #2 of 5

The P5's are not particularly difficult to drive. I couldn't find a graph of their impedance vs. frequency, but most headphones benefit from being driven from a low impedance. A max. power in of 50mW equates to a current of ~44mA. You get 115dB @ 1V, which is plenty loud.


I'd still use paralleled AD797s if that is what you choose, but why not use a dual? NE5532s are cheap (dirt cheap) and widely available. I bought 10 for $3 US shipped from Thailand:- http://www.ebay.co.uk/itm/10-x-NE5532-Dual-Low-Noise-Operational-Amplifiers-/220759929025?pt=LH_DefaultDomain_0&hash=item33665130c1#ht_1738wt_931



post #3 of 5
Thread Starter 

Hi Waki


Thanks for your reply.  I was trying to get my head around the whole impedance thing because although the P5's have a super low impedance, they were designed specifically for the iPod / iPhone so I couldn't imagine that the tiny amp inside one of these units could provide more current than a CMOY.


I have noticed though that my current unbuffered CMOY distorts the bass frequencies even at low volumes.  I have it set up for a gain of 2 since the P5's are very sensitive, as you say.  44mA is just outside the capabilities of the TLE I believe, so I'll still go for the buffered supply.  Do you think the additional opamp is overkill in this design?


Headphone amp ground channel


What are the benefits to using two dual opamps for the audio section rather than one?


I'm trying to create the highest quality sound that my current level of expertise will allow; hence upgrading the CMOY in stages as it were. My thinking behind the expensive opamps was obviously to get the best components into my creation (I'm not bothered about cost to be honest).  With the amps being the most expensive parts, if I decide to build something better than a CMOY/A47 in the future, the amps can easily be transferred.  I'd read on head-fi that the AD797BRZ was basically the best opamp around (and it should be for the price!).


Thanks again for your feedback.

post #4 of 5

OK, a couple of points.


Although there are a number of '3 channel' or 'ground channel' designs on show on the web, the general consensus in the wider engineering community is that these don't actually perform as well as designs using a more conventional ground arrangement such as using the junction between two 9V batteries for the ground. Some test data on this subject is available, but I can't link to it because the member has been banned. Take a look at the O2 amplifier thread though...


This is not the whole story, because some sources insist that a 'virtual ground' or rail splitter design runs less risk of damaging headphones due to a DC offset at the output should one battery go flat while the other still has output, but given that there is measurable degradation supported by a theoretical rationale it seems a contradiction to me to spend a lot of money on opamps only to employ them in a dubious configuration.


There are other ways of building an amplifier to ensure that a DC offset does not damage headphones although they all add complexity to the design and reduce battery life. If, however, you feel confident that you can monitor the battery situation yourself (some distortion would almost certainly become audible before damage to the headphones occurred) and are not likely to fall asleep while using the amplifier, then this protection can be omitted entirely.


A number of searches on my part have failed to disclose anybody complaining that their headphones were destroyed by the failure of one battery out of two.


That said, I do think the additional opamp in your diagram is overkill.


The Grado RA-1 portable headphone amp uses a single NJM4556 opamp. This is a cheap device. Less than a dollar US, although they can be awkward to obtain at the lowest prices particularly as demand for hobby quantities has been high recently due to the O2, which uses 2. I understand your desire to have the very best, cost no object, and the AD797 is undoubtedly a very fine opamp, but all components need to be matched to their intended use and its output @ 50mA is less than the 70mA (times 2) of the NJM4556. Whether the other performance differences between the 2 devices are audible, as opposed to measurable, is open to question. I doubt very much that it is, but others will insist that they can hear one. Grado certainly choose to use the NJM4556 in a $400 product, which would seem to allow sufficient profit margin to permit the use of any opamp under the sun.


The principal advantages of a dual as opposed to a single are cost and convenience. Where 2 devices are being paralleled there is the additional factor that devices on the same die are well matched, which lends itself to paralleling. Crosstalk between duals as opposed to singles may be worse, but still exceeds the known threshold of audibility by a considerable degree.


Another device you could look at is is the TLE2062, which has a very low quiescent current. Although the distortion is higher, it is still probably below the threshold of audibility. Again, using a dual per channel would be recommended, just because it is nice to know you are not running a device close to its limits. Most amplifiers' distortion goes up with output.


The main disadvantage to the NE5532s I mentioned earlier is the 10mA quiescent current which eats into the battery life a bit.


A simple protection circuit:-




As shown with 8.4V NiMh PP3 batteries, when the total battery voltage drops by > ~5V the relay will drop out. The cost is 10mA. The relay will tolerate considerable excess voltage so regular alkalines can be used. With the 11V zener the relay should pull in as long as the batteries are not close to discharged. It will drop out and will not pull in if one battery is flat or disconnected. The relay is good for 2A.




Edited by wakibaki - 10/29/11 at 3:39pm
post #5 of 5
Thread Starter 

In case anyone is following this discussion, here's what I went for in the end:


I created a VERY simple CMOY variant, omitting the voltage splitter completely and simply using the junction between the two 9V batteries.  I haven't built in any fail-safes at this point...just bought two new Duracell batteries which should discharge at more or less the same rate (and should produce audible distortion when low on juice, before any damage can be done to my headphones).  The power supply has 2 x 470uF Panasonic FC electrolytics.  All resistors are 1% vishay.


I went for 2x AD797BRZ opamps on an adapter, set to a gain of 2.  I consulted the datasheet and used 0.1uF ceramic and 47pF tantalum capacitors from the  supply pins to the virtual ground.


I also added a 22pF bypass capacitor in the feedback loop (as per the data sheet for a gain of 2) to aid in stability.  I did not use an 'R5' resistor.  The signal path is quite literally straight in to the opamp and back out again to the headphones.


I measured the DC offset at the ouput jack with my iphone connected and playing music.  It was less than 5mV for each channel :-)


I first tested it with an OP2227 installed (in case it blew up!) once I was confident I swapped in the AD797s.  Big difference in clarity and sound stage.


It sounds wonderful needless to say; everything I've tried so far has felt like there was some kind of 'veil' or curtain between me and the music.  The curtain has been lifted.  I couldn't be happier.


Before I started designing this amp I ordered an O2.  Once it arrives I'll be interested in comparing the two amps - I'll let you know how I get on!





Edited by pjoliver182 - 11/8/11 at 7:06am
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Suitable portable amp for B&W P5's