The Objective2 (O2) Headphone Amp DIY Project

Administrator
Joined 2007
Paid Member
Thank you Mooly, just removing the resistors r17 and r21 is ok? no need to bridge anything?

If its working as you want then its OK although personally I would then link out R16 and R22 as a I mentioned to make it a true unity gain buffer stage.

I have another question! What does the Mosfet Q2 and Q1 do?
P Ch MOSFET Vgs > 24V Vds > 25V Rds < 0.5 ohms IPAK GDS
N Ch MOSFET Vgs > 24V Vds > 25V Rds < 0.5 ohms IPAK GDS

I just realized i have them reversed and as soon as i plugged in a 12v power supply, they both burned...

Forget my last message, i was asking for U5 and U6
12 Volt Positive Fixed Regulator TO-220
12 Volt Negative Fixed Regulator TO-220

If you've fitted the regulators incorrectly and they burned then you need new replacements.
 
If its working as you want then its OK although personally I would then link out R16 and R22 as a I mentioned to make it a true unity gain buffer stage.

If you've fitted the regulators incorrectly and they burned then you need new replacements.

The Regulators are burned for sure, but the amp still works on battery power, that's a good sign i didn't burn anything else. I'm building 2 more amps as pure desktop versions - this was my first attempt at the O2 and i wanted to have a baseline.

What do you mean a "true unity gain buffer stage"? Just to make sure i understand, i remove R16 and R22, as well as R17 and R21, could i remove the gain switch since it's effectively useless? Would i need to bridge anything then?
 
Administrator
Joined 2007
Paid Member
You remove R17 and 21. That allows that switch position to give unity gain.

Test the amp and confirm its OK gain wise.

If you never ever want to use the other switch positions to give different values of gain then I (that is to say me personally :D) would probably then short out R16 and R22. A true unity gain buffer has the - opamp input connected to the output pin directly. If you leave the resistor and cap in place then detailed testing (squarewave testing with a scope) could reveal a problem if you leave the cap and resistor in place. No point removing them... just link them out.
 
Mad Dogs
https://mrspeakers.com/product/mad-dog-leather-alpha-pads-comfort-strap/

Xonar DG optical -> Fiio D30k (output voltage is 1.5v)

DIY Objective2 (stock components)

I don't know if I'm getting the most from my amp because it doesn't get as loud as I thought it would. To make sure I'm not clipping the input on the amp, I create a 1khz sine wave at 0 db and adjust the volume till I hear no clipping. Then I drop it down a couple notches just to be safe.

I usually listen to music at around 85-90db depending on the song. I use a Radioshack SPL meter and a cardboard cutout so it covers the entire ear pad. Not perfect but should give me an idea. Peak SPL at max volume was 110db. According to this review http://www.innerfidelity.com/images/MrSpeakersMadDogA.pdf the headphones only needed 0.52mw for 90db. For 117db I should need around 250mw which I think the O2 should be able to do at 50 ohms.

Does this seem right?
 
You remove R17 and 21. That allows that switch position to give unity gain.

Test the amp and confirm its OK gain wise.

If you never ever want to use the other switch positions to give different values of gain then I (that is to say me personally :D) would probably then short out R16 and R22. A true unity gain buffer has the - opamp input connected to the output pin directly. If you leave the resistor and cap in place then detailed testing (squarewave testing with a scope) could reveal a problem if you leave the cap and resistor in place. No point removing them... just link them out.

Thank you Mooly, much appreciated!


Now i need to find instructions on a relay to get rid of that "pop" when my headphones are plugged in or when the amp is turned on. I've only read 20 pages of this thread out of 415...:eek:
 
Administrator
Joined 2007
Paid Member
Thank you Mooly, much appreciated!


Now i need to find instructions on a relay to get rid of that "pop" when my headphones are plugged in or when the amp is turned on. I've only read 20 pages of this thread out of 415...:eek:

You're welcome :)

(I've never seen a practical circuit for the 02 for muting although it should be easy to rig something up. What you could try... and it won't damage anything... is to see whether the opamps can stabilise if a dead short is placed on the headphone output, that is to say on the output side of those 1 ohms. If it did, then you could arrange for a normally closed relay to be across the output. After a short delay it would open and all would work. On power off it would be arranged to drop out instantly. If it didn't work with the 1 ohms then I would be tempted just to increase those until it did, perhaps 2.2 or 4.7 ohm. Its the easiest way to do it and avoids passing the signal through any contacts.
 
250mw into 50 ohms is 3.5 volts rms. That's a tall order for an opamp based output stage... the 02 won't do it even with its paralled opamps.

3.5 vrms is 5 volts peak. That gives a peak current of 100ma, way in excess of what the two 2068's can deliver.

I glanced over the AGDR booster project for the o2, is that something that would help in my situation or would I be better off looking at a different amp?
 
Administrator
Joined 2007
Paid Member
Right... :)

But is it distorting (clipping) when you turn it up. If so then you need a higher current capability in the output stage. 4556 or even 4560 opamps (from memory) could give you that extra capability for the output stage.

If it doesn't distort (you have the volume up full and its still not loud enough) then you might just need a bit more gain... dead easy, just change the gain setting resistors.
 
Now i need to find instructions on a relay to get rid of that "pop" when my headphones are plugged in or when the amp is turned on.

There is a modification for that here. :)

Its the same circuit I used with the headphone relay on the O2 Booster Board to eliminate the O2 turn-on and turn-off pops. I would recommend using that Booster Board circuit rather than the older mod post in the link since it has a couple of modern tweaks. Works exactly as Mooly says. There is a turn-on delay to allow the power rails to stabilize and the chips to thermally stabilize, which nukes the turn-on thump. Then in the case of the Booster Board circuit it gets some advance notification of turn-off from the gate of the mosfets, to allow the relay to open before the mosfets have started cutting power to the output chips, to get rid of the turn-off thump. Works great. :) There are several posts to that effect in the Booster thread from folks who built it.

The schematic and BOM (parts list) for the entire O2 Booster Board circuit are in the Google Drive link in the first post of the Booster thread. You could easily whip one up yourself on a proto board. It probably wouldn't fit though in the standard O2 B2-080 case (in that link above I later discovered that and had to use the B3-080 taller case). To get it to fit in the B2-080 case with the Booster Board I had to use a single DPDT sub-miniature relay and mount it upside down underneath the Booster Board. :) Still might be possible to do that with a proto board though.
 
Last edited:
Administrator
Joined 2007
Paid Member
Isn't the NJM4556 the "stock" opamps? Its what I'm using.

Ah, my mistake, for some reason I was thinking it used 2068's throughout.

In that case I would have thought the 4556 should have more than enough output. Another thought. Are you boosting the extreme (and probably inaudible) bass so much that its overdriving the 'phones and/or that at extreme inaudible LF, the impedance has fallen to a low value ?

Perhaps a simple LF roll off would solve the issue without any audible effect. You could try reducing value of C12 and C13.
 
Administrator
Joined 2007
Paid Member
It really does seem then that you need more voltage swing (into your particular phones) than the O2 can deliver.

This is where you could do with hanging a scope on the output and seeing where the problem really is, see what voltage swing it can deliver before clipping into your phones.