[uelover]

Quote:

Haha, not that I've heard of. My understanding is that 1x gain is "unity gain", and I haven't heard of anything lower than that.
A lot of people have also commented that a 1x gain is only really suitable if you're listening at very low levels, so realistically speaking, there probably isn't need for it.

My DAC lineout is pretty loud and my ES5 is pretty sensitive. So, perhaps, the lower the better will apply in my case =)
 

 

[jseaber]

No need to e-mail us. There's a "notes" field in the checkout process for...making notes. 

 
Quote:

firev1: Thanks again. Wow, I guess the O2 interest really is a global phenomenon.
 
Everyone: if you're buying from JDS or Epiphany, do ask about customising the gain settings on your particular unit. I spoke to JDS, and John informed me that he could change the low gain setting if need be.
Hope that helps you out if you're no DIY-er (like me).

 

 

[firev1]

     Quote:

My DAC lineout is pretty loud and my ES5 is pretty sensitive. So, perhaps, the lower the better will apply in my case =)
 

 
I know what you mean, my DAC output measures 2.4 - 2.6V and I barely use beyond 8-9 o'clock even on 1X on Desktop with my TF10s. 

 
Quote:

My O2 vs V200 thread is nearly finished now. Got all my photo's done and most of it written up now. Should be posting it soon'ish. 

Looking forward to it
 

 

[mikeaj]

For lower gain you could use another O2 as a pre-amp to an O2.  Use the volume control on the first one to attenuate.  Yep.  

 
Well seriously, this is where some amps with multiple gain settings (including negative) would be useful, and/or multiple stages.  That increases complexity, cost, etc. though.
 
I've got like 0 + epsilon background in op amp circuits, but the non-inverting amp configuration the O2 uses has gain 1 + R_a / R_b, so you're not going to get lower than 1 with any resistance values.  I'm not sure what kind of configuration would be most suitable for negative gain.  Of course, you could use a resistive divider network somewhere along the line, or equivalently, something like another potentiometer on a second gain/attenuation stage.  Actually, if you put an additional pot before the input stage (that's where the pot is on many audiophile amps, instead of between the input and gain stages like it is here), then you could reduce the input voltage some, which admittedly reduces SNR.
 
Another thing to consider is a volume attenuator / hiss buster circuit or adapter for the IEMs.  For each channel, put a resistor R_1 in series and another resistor R_2 in parallel with the load.  If R2 << impedance of the IEMs, you get roughly R2 / (R1 + R2) of the voltage you would have gotten, to the IEMs.  Adjust R1 accordingly to be larger than R2, so you get a lot of attenuation.

 

[maverickronin]

Quote:

Can O2 drive K501 well? I can't find any review with K501.

I couldn't say for sure, but I'd imagine it would be ok.  Its great with my K601s and many others like it with K701s.
 
Quote:

Another thing to consider is a volume attenuator / hiss buster circuit or adapter for the IEMs.  For each channel, put a resistor R_1 in series and another resistor R_2 in parallel with the load.  If R2 << impedance of the IEMs, you get roughly R2 / (R1 + R2) of the voltage you would have gotten, to the IEMs.  Adjust R1 accordingly to be larger than R2, so you get a lot of attenuation.

 
What does that do to the output impedance the IEMs see?  It really should be as low as possible with BA drivers.
 
Another option, if you have a good 24 bit DAC, is to lower the volume in software.  The extra bit depth will keep you from losing dynamic range with 16 bit recordings and nothing actually uses enough that you'd all the dynamic range in a 24 bit one.

 

[mikeaj]

Quote:

What does that do to the output impedance the IEMs see?  It really should be as low as possible with BA drivers.

 
I've never seen or done the analysis for this or any similar situation, and I don't really have the background in feedback theory, but I don't imagine the effect would be great.  The amplifier drives this new, probably more resistive load, rather than the original load, and it has no idea what's going on with the headphone drivers.
 
Certainly the FR would be mostly unchanged since the voltage is getting scaled mostly constant over frequency.  As for phase, correcting back EMF, etc., I'm not sure.  Do those things really matter that much?

 

[maverickronin]

Quote:

I've never seen or done the analysis for this or any similar situation, and I don't really have the background in feedback theory, but I don't imagine the effect would be great.  The amplifier drives this new, probably more resistive load, rather than the original load, and it has no idea what's going on with the headphone drivers.
 
Certainly the FR would be mostly unchanged since the voltage is getting scaled mostly constant over frequency.  As for phase, correcting back EMF, etc., I'm not sure.  Do those things really matter that much?

That would still raise the output impedance  as measured from the output of the attenuater adapter unless R2 was really small though, right?  With impedance curves like this you want the output impedance to be as low as possible.
 
Isn't the math the same as this example on the Meier site?  (At the bottom of the page.)
 

 
Ignore the O2's .5 ohm Z out, let his Ro = R1 and Ra =R2 which makes Ro part of the adapter.  Then the output impedance of the amp + adapter should be (R1*R2)/(R1+R2) or did I mess something up?

 

[hekeli]

Quote:

 My one complaint about the build is that the 3.5mm socket feels a little stiff when inserting a jack, and possibly contributing to this is that the socket move slightly when using them, as if the face plate has a little room for the sockets to wiggle. It's a relatively minor issue though, and honestly what I would expect out of a amp built on a DIY design.

I was bothered by the jacks at first too. But I've done a lot of inserting and removing since, and it's better now, not an issue really. JDS even offered to replace the unit, which is nice, but I don't think it would have made a difference.
 
Would have no problem paying few bucks more for better jacks though.

 

[thehadi]

Quote:

I couldn't say for sure, but I'd imagine it would be ok.  Its great with my K601s and many others like it with K701s.

Thank you

 

[hekeli]

Quote:

For all the people in EU. Bear in mind that if your order from JDS Labs like I've done you need to pay the import duty plus the admin charge from the postal services (Royal Mail in my case) which added up to the total. Precisely I got charged £26.00 in excess to the $162 (Shipped) from JDS in the US. I didn't know about Epiphany Oliver at that point or else would have gone for that to avoid the unnecessary customs charge.

 
It's surprising that there really are people who don't know about customs etc.. even our customs agency has a "guide for dummies" for buying on the net.
 
Total price to Finland (EU):
 
Epiphany: 117 GBP = 183 USD
JDS labs: 160 USD + 37 USD (customs) + 15 USD (power) = 212 USD
 
So it would have been a measly 29 USD difference here. Bear in mind that not all packages end up in customs, so you might get lucky. But I happily paid that for the "JDS trust" and better specced power (link) of my own choosing.
 
Of course there is the chance of needing warranty etc, but from here it's pretty much the same shipping to US or UK. Might just take longer.

 

[KGee]

I just finished by O2 build a few days ago and am impressed with the sound so far, although I'm waiting for my HD650 to arrive.  Hopefully I'll be blown away.
 
I'm also thinking about clipping/removing R19 and R23 to get 1x gain instead of 6.5x with the gain switch in.  I'm not sure if 1x gain is a good match for HD650 vs. 2.5x gain, but I'm barely at the 10 o'clock volume position on 2.5x gain with my MDR-7506.

 

[hekeli]

Quote:

I just finished by O2 build a few days ago and am impressed with the sound so far, although I'm waiting for my HD650 to arrive.  Hopefully I'll be blown away.
 
I'm also thinking about clipping/removing R19 and R23 to get 1x gain instead of 6.5x with the gain switch in.  I'm not sure if 1x gain is a good match for HD650 vs. 2.5x gain, but I'm barely at the 10 o'clock volume position on 2.5x gain with my MDR-7506.

As it has been stated, unless you never intent to use low output portable sources, you can't even use the 6.5x gain. I clipped mine by just removing one head of the resistors (pulled with small pliers) and left them hanging in the air. So in theory it should be dead simple to solder them back if needed.

 

[vkvedam]

Here we go!!
 
!
 

 

 
Did I do something wrong?

 

[firev1]

Quote:

Here we go!!
 
!
 
 
 

It looks to me that R17 is kind of clipped which may be the source of your gain problems. 
 
 

 

[mikeaj]

Quote:

That would still raise the output impedance  as measured from the output of the attenuater adapter unless R2 was really small though, right?  With impedance curves like this you want the output impedance to be as low as possible.
 
Isn't the math the same as this example on the Meier site?  (At the bottom of the page.)
 

 
Ignore the O2's .5 ohm Z out, let his Ro = R1 and Ra =R2 which makes Ro part of the adapter.  Then the output impedance of the amp + adapter should be (R1*R2)/(R1+R2) or did I mess something up?

Oh, that looks right then (taking without proof, and without going through it myself).  Generally you do make R2 really small.  With say Ro (R1) = 16 ohms and Ra (R2) = 1 ohm, Ro * Ra / (Ro + Ra) gives you 0.94, not that bad.  Voltage to the headphones would be approximately 1/16th of normal, so 20*log10(1/16) = -24 dB gain.  Those are about reasonable numbers for both output impedance and attenuation, for those multi-balanced armature IEMs that have something like 130 dB SPL / 1V.
 
 
 

It looks to me that R17 is kind of clipped which may be the source of your gain problems. 

 
That should effect one channel but not both.  Also, R17 looks a little rough but not cut through.  To me, it looks like the PCB trace is damaged, where the resistors were taken out.  Thus the resistor is no longer connected to the circuit where it should be, so you have an open circuit and 1x gain.
 

 
I've marked it with red arrows.  There's no longer a connection between the spots indicated by the arrows.  The start of the arrow is where the gain resistor is.  It's supposed to be connected to the rest of the PCB trace, but the damage looks like the connection may be broken now.  You could manually solder a couple wires (insulated) between the spots the resistors are and the points where the PCB traces end, at the next solder point on the PCB for the next part.  i.e. bypass the PCB trace and just do that portion point-to-point.