[tamasic1]

Hey all.

I’ve built 4 headphone amps previously and they’ve all turned out great. This weekend I started a new project and tried AMB’s Mini3. Not from the kit but from parts I had. Built it verbatim of the parts and schematic. It worked first time but …
I’ve read that the AD8397 is a bit bright. Sorry, but a bit bright is an understatement. A bit bright is when I open the bedroom door in the morning and a few of the sun’s rays are coming through the kitchen window as I make my way to the coffee pot. To me the sound I get from this amp is if I walked out into the yard in the morning and stared directly into the sun!
Sources are line level sony cd player and a new Sandisk MP3. Headphones tried are Sennheiser PX100 (32ohm), Koss (300ohm) and a cheap set of Behringers (unk). Same experience no matter what combination is used.
Will this mellow after a burn in period, is there a way to knock it down or is this just the way this amp is?

TIA

Amps built - Cmoy, Cmoy w/CF, Apheared, Starving Student

 

[FallenAngel]

Check current draw to make sure it's really "working properly", AD8397 isn't overly bright, but it's not slow/drawn-out.

 

[CodeToad]

Quote:

Originally Posted by tamasic1 /img/forum/go_quote.gif Hey all. Built it verbatim of the parts and schematic.

When you say verbatim did you use the same parts recommended on the mini3 parts list? Did you follow the intial checks and all of them passed correctly?

I've never heard anyone say a mini3 is bright before. Especially with PX100's.

 

[diditmyself]

Either there's something wrong with this amp or the amp you use as a reference.

 

[tamasic1]

Thanks all. Bad ground buffer.

FA - Current draw right on the money at 24.8 mA

CT - Only difference in parts is that I used caps and resistors I had in the parts box but they are all the same type and value. But you're right I should have went back to the site and read more. I've read several times about people who didn't like what they heard at first but after X amount of hours of 'burn in' said it sounded better. I didn't know as I've only built a couple amps previously.

 

[diditmyself]

Burn in will never change the sound dramatically. Such reports are either vastly exaggerated or the change is in the head of the listener.

It's often stated that RFI/EMI will influence the sound of amplifiers, and an amp could sound different depending on what activity is going on in the area.

 

[tomb]

The AD8397 is prone to oscillation unless the PCB design is just right. Oscillation can very often be mistaken for highs that are too strong and a lack of bass. We know that AMB's layout is good, so look for deviations in parts selection or construction. Did you mount the ferrites up above the level of the board, for instance? Just a thought ...

 

[cobaltmute]

Any high speed op-amp can go bad with small changes in layout. I had a LMH6643 oscillate because one of the feedback resistors was moved from the top of the board to the bottom. IT seems like the AD8397 can be cranky in the same way.

 

[tamasic1]

Hi Tom. I didn’t buy the kit, yet, and didn’t have a pcb. I created it on a rat shack pc board. Ferrites are above board. I also created my own adapters for the amp chips and believe I damaged the ground.

It was Sunday morning, I didn’t have any other chips and didn’t feel like waiting for one to be shipped so I hacked together a Sijosae buffer out of what I had. Worked right off the bat, everything tested correctly and the sound was amazing. Don’t need to use much of the volume with the low-z phones. I’ll replace the buffer next week when I do my parts order.

Thanks CodeToad for reminding me to go back to the check list. Spent 5 hours listening to the amp after fixing and it wasn't fatiguing at all.

 

[amb]

tamasic1, AD8397 is a very cranky opamp and is almost certain to misbehave when it's built on a perfboard and mounted on an adapter. It is extremely sensitive to trace inductances and a well-designed PCB is a must to tame it. Read the Mini³ development threads over at headwize.com for details.