12 April 2010: See NP-100v12: DIY 12AU7 (ECC82) Tube / IRF510 MOSFET Headphone Amplifier for the latest update to this project.

I have an update to the 12AU7 / IRF612 headphone amp hybrid I made, this one is smaller and has less parts. It is a great design for a new builder who wants tube sound at a low price. It has about 10db of gain and about 250mW of power, great for Grado or other low Z headphones. The issue is that there isnt much noise rejection from the power supply, so I use a 12v lead acid for the PS. The whole amp will fit on a standard Radio Shack proto (72x95) board and only 8 components per side. I will post some scope pics as soon as I get my scope back, but the whole thing draws about 250mA, so with the 9Ah battery I have it last a few days between charges.

Hi Rogers,

Thanks for the update and great work! Yeah, the hardest part of these is the power supply. A 12V battery eliminates all those worries. When I built my Class A MOSFET Headphone Amplifier I tired a bunch of different supplies and was shocked at how bad some of them were.

Did you build in a jack/switch so you can charge the battery without removing it?

Have you thought about trying a LM317 as a CCS instead of the load resistor? That would improve the efficiency which is not an issue running of the wall, but would be nice for a battery.

Hi Gio,
Yes, since the pic I have added a charge switching circuit and external posts. I have a wallwart type battery charger and over the last two days it lasted about 18-19 hours. I also built a LM741 based low voltage circuit http://www.reuk.co.uk/LM741-OpAmp-Voltage-Indicator.htm to tell me when to charge the batt.
I have considered the LM317 and also other CCS's like the dual fet and another mosfet. That is what i like about this circuit is it is easy to experiment with and the chassis is an old supermarket PA system that I got for $1. It has a lot of room under the hood for more mods.

That enclosure is perfect.

Did you play with C2 to see Improve the response?

I'll update the project page and post a link back to here.

Cheers.

For this amp the 1u input cap works great at nominal voltages, The original design had a 36 volt B+, but this little guy amplified the noise more than anything, so decided to move down to a 24 then 12volt B+. The battery is a Vision CP1290 which is used in APC ups's, I have a bunch of old ones and it is rated at 20 hours at 450mA of draw. This unit only draws 250mA, so I am going to play with adding another parallel mosfet leg, with a more efficient CCS as you stated. With a DC supply it works great.

Ok, keep us posted on your progress. I personally like the CCS better. A complimentary MOSFET is likely a better choice over the LM317.

What frequency is that square wave?

The wave was at 1khz, I ran it at 100hz-to 10khz and it was fine, maybe a small amount of ramping at 100hz, but not as much as we experienced before, This can be resolved by increasing the input cap to 2u or so. Here is the total noise, it is pretty consistent with small homebrew amps of its like.

Are you getting a power on pop? Going to 2uF might make that even worse.

I think that noise is pretty good. You should drop a zobel into the simulation and see if it helps the noise on the high end.

Definitely have some pop, I have the 1u in right now and do not have any issues on the lower end. It sounds pretty good. I built a nelson pass time delay for another amp, but it required a 24v supply, I just have to tweak the RC and get a 12v relay, the only thing is it only handles power on pop, not power off. Amb's e12 does both but it is for dual rail supply. I will have to do some research. For now I just unplug my phones, $175 SR125's I am not taking any risks.

I guess you could also wire up a mute switch on the output ground as well. Yeah, you don't want to be killing Grado headphones.