Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Limitation of the A47 design
New Posts  All Forums:Forum Nav:

Limitation of the A47 design

post #1 of 7
Thread Starter 

Firstly thanks for all your help so far, I think I have nearly figured out how those little op amps actually work (calculating gain and whatnot) 

 

 

As I said in an earlier post I intend to make a desk amp, as an oppose to a pocket amp, and ideally this will  be the last one I need to build for a while so I am going all out.


I was looking at the Apheared 47 amp and it look like a good solid design, ideal for desk situations, He documented that it is limited to low impedance H/phones like Grados that make good use of a good current supply, and that these are not suited to higher impedance h/phones.

 

here is my question, is this because the gain is to low to generate a high enough voltage for high-Z /phones. If this is so, could you not add a switch to allow you to lower the 4.7k resistor for a 2k or 1k resistor to provide higher gain for highZ headphone which would benefit from higher voltages at a lower current?

 

Thanks.

 

Apheared 47 design

http://gilmore2.chem.northwestern.edu/projects/showfile.php?file=apheared_prj.htm

post #2 of 7
Thread Starter 

BTW: as I mentioned it will be wall powered with no need for portability so Wattage is not an issue for me as long as the amps can handle it.


Thanks again

post #3 of 7

 

Quote:
Originally Posted by dobby156 View Post

 

As I said in an earlier post I intend to make a desk amp, as an oppose to a pocket amp, and ideally this will  be the last one I need to build for a while so I am going all out.

 


Which earlier post? Why not tag this onto that thread? It is better to bump an old thread that you have a specific question about than to start another. At the very least link to it.

 

As JCX pointed out in an earlier post, not all opamps deal well with large swings on the inputs particularly when we swing near the rails (see this thread HERE, post 27,) which may limit their use as buffers if you want the CHA-47 to swing a lot of voltage. You are also limited to op amps which are unity gain stable for the "buffer" position (or the whole thing depending how you build the beast) but thats not too hard to do.

 

Now that thats out there, the real question of suitability boils down to how many volts do you want the amp to swing? Most people take no effort to learn and know this answer beyond "heh, as many as possible" which is a fun although somewhat expensive and challenging thing to design for. In many ways its wasteful... Designing an amp to swing 15vrms when you will never even use 1 full vrms...

 

How many volts do YOU want the amp to swing? After you know this, you can knock that off the list of questions before you build something... "will it go loud enough" yes/no. Its simple! 

 

To figure it out, there are a couple methods. The best one is to ignore the pencil pushers, build the amp, and listen. Does it go loud enough? Yes? congratulations. whoever told you it dosnt have enough gain must have been doing something very different from you. If it dosnt, put it on the FS boards. Theres a lack of people building anything you cant find on a PCB these days.. give it some time and you are almost certain to recoup your parts cost on a slightly unusual build.

 

A slightly less expensive method is to take the headphones you have now, and measure how much voltage swing it takes to drive them to your desired level. Whats that you say, your meter dosnt measure below 0.1v? Thats a bummer, but it should probably tell you something....

 

The second method is to look up the rated efficiency of your headphones, which is typically given in db/mw or db@1v(note! rms or p-p) bust out the calculator and work it backwards. 

 

Add a few db for headroom, and if you change source or find a particularly quietly recorded bit of music. Realistically, 12db will probably cover most people here. Many people will STILL find themselves within what they can cover with gain of 3, and possibly less. 

 

Once you know how much voltage swing you need you compare that to the voltage output of your source. For most people the voltage going into the headphones will be lower than what comes from the source. Its funny how much money we as a group spend on attenuators which we set to -50db so that we can have a +20db gain stage for a net output 30db lower than the input voltage.

 

A lot of people like to build with more gain than necessary because when you have gain you can keep the volume knob wayyyy down which gives a feeling of massive power from an amp. Its an illusion. There are a few very REAL benefits to building an amp with lower (but still adequate!) gain VS higher. Real benefits VS an illusion. 

 

With lower gain you can spin the volume knob up higher. This keeps the volume control in a better matched area. Even expensive pots have channel matching issues from time to time at low volumes... everything (even cheap stuff) matches well at high volumes.

 

With lower gain you have less "gain related noise" and noise in general. Simple.

 

Although its a little theoretical, GBP (gain bandwidth product) means you will have more bandwidth with lower gain. Newer opamps have sooooooooo much bandwidth that even with gain of "holly $#@% thats a lot" ! you still have way enough bandwidth but it never hurts to mention. 


Edited by nikongod - 7/3/10 at 11:43am
post #4 of 7
Thread Starter 

Sorry I didn't really need to mention the other thread here as it was not srictly relevant.

 

Thanks for the reply, it was helpful, and I take on board what you say about unnecessary gain. So is the limitation of a A47 design with the second Amp being used as buffer not responding well to larger AC voltages, not necessarily the gain?

 

So the buffered design is not suitable for high impedance h/phones?

 

Thanks

post #5 of 7

Maybe you should consider a design like the Szerkeres VE3...

There is not too much it can not drive.

post #6 of 7
Thread Starter 
Quote:
Originally Posted by Avro_Arrow View Post

Maybe you should consider a design like the Szerkeres VE3...

There is not too much it can not drive.





Quote:
Originally Posted by dobby156 View Post

 So is the limitation of a A47 design with the second Amp being used as buffer not responding well to larger AC voltages, not necessarily the gain?

 

So the buffered design is not suitable for high impedance h/phones?

 

Thanks


so it is not suitable?

 

I will take a look at that design though, it looks interesting.

post #7 of 7

It's not so much a question of suitability as it is "how much power can

you use?" A CMoy is a great portable design as it is generally easy

on batteries. An A47 is also a great portable design but a little harder

on batteries. A Buffered CMoy can be built as a desktop or portable

design. A full class A buffer like the Szekeres is power hungry but

also provides quite a bit of power to your headphones. The amount

of voltage it can put out depends on the voltage coming in as the gain

is less than 1. Many of the designs I explored in that thread are scaled

down Nelson Pass Zen series amps.

As Nikongod pointed out, many people have no idea how much voltage

or current their headphones require. Grados can be driven to ear shattering

levels with just a couple of volts...

My DAC can produce 5 volts and my VE3 can produce 150 mA.

My IEM's will melt if I ever used that much power on them...

New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Limitation of the A47 design