[morfic]

Obviously sounding good is most important, the DC offset fries headphones, so it's beyond technicalities of RMAA measurements. (much loved amps often look pretty wonky in RMAA)
the higher the offset, the more likely it'll cook the phones, the loss of dynamic range aside (the DC offset shifts the phone "to the side" by the amount of the DC offset)

And yes, that page is where the whole idea of running 2 bypass caps per rail to ground is coming from. It's also him who i got the 20mV safe offset threshold from.

 

[PatOMalley]

We have output caps and the LM is not being used for gain am I right? Just for impedance matching. And LM is stated to have .1mV offset, lower than the stock opamp. I think you could actually remove the output caps.

 

[lazpete]

Back from a week in the sun, I opened it up, and replaced the stock op for a LM4562NA opamp, and at the same time I added a pair of CV4015 (EF92) mullards to the equation. And my new ref tune, old woodstock with van morrison on a brand new 180gr vinyl, just sounds fantastic. The acoustic guitar on the tune starting a new life is so real. Im gonna give this setup plenty of time, to see whrere it takes me.
This is listening trough a pair of K601, gain on low.

 

[morfic]

You ought to try some CV4014 Mullards.

Pat: opamp does voltage amplification, yes.

 

[lazpete]

Quote:

Originally Posted by morfic /img/forum/go_quote.gif You ought to try some CV4014 Mullards. Pat: opamp does voltage amplification, yes.

What would be the primary difference to the CV4015 Mullard?

Br
Lasse

 

[morfic]

Vocals are clearer, bass goes deeper, no loss of high end detail.

 

[morfic]

Well, i just measured the DC Offset of the LT1364 in the I+ And it is at 28mV, almost 1.5x what Tangent recommends as a safe value to ensure you do not cook your phones.
So for anyone interested in using this opamp, it's not a suitable choice in the I+ w/o changing the circuitry to compensate for the increased DC offset.

For anyone interested in opamp rolling, stick to FET based opamps or check that the bipolar transistor's "Input Bias Current" is in the nA or less range (LME49720/LM4562 are this low and thus they work, FET based opamps are in pA range)
The uA input bias current which is more typical for bipolar opamps can cause a large enough DC Offset to be worrysome.
Data sheets are easily available via google, easy to double check this before using any bipolar opamps.

LM4562/LME49720 are a notable exception, 1mV of DC offset is nothing to worry about.

 

[lazpete]

morfic, thx for the interesting input. Gotta check the CV4014 in a near future, thats for sure.

 

[bronth]

Quote:

Originally Posted by morfic /img/forum/go_quote.gif Well, i just measured the DC Offset of the LT1364 in the I+ And it is at 28mV, almost 1.5x what Tangent recommends as a safe value to ensure you do not cook your phones.

Interesting find. But IIRC LD I+ has separate transistors as its output chain (i.e. buffers) thus the op-amp itself actually doesn't work as a headphone driver. Any thoughts?

 

[morfic]

The buffers are not there to kill DC offset.

I measured what came out of the headphone jack, not the chip's outputs to ground, so what i measured goes to your headphones, biasing your headphones at best and cooking the coil at worst.

I just noted it here since i know it's a popular choice and that hardly anyone checks into this, and I too only learned about any of it because i wanted to build a bare bones amp (cmoy) to see if opamp difference are more noticable there, not sure on more noticable, just know i there too like the LME49720 best.

Oh and the OPA2227 that i got with my I+ i tried in the cmoy and although it is a (di)FET, performed horribly, unstable, distorting, just a side note

 

[PatOMalley]

I read that the 2228 is more stable in cmoy than the 2227. In any event I like the LM4562 which I have been using for a couple of years in several things, cambridge 640p pho pre, project headbox, Bugle phono pre, now the LDi+.

 

[morfic]

I'm with you on the LM4562/LME49720, no matter how often i try opamps, i always end up with it, sounds best, and now i even know it is a "safe choice"

 

[taso89]

I used an LT1364 for a while in this amp and after reading the posts regarding DC offset, I put a spare LM4562 I had lying around. Didn't notice a diff. in SQ but if I can avoid cooking my cans I will do so

 

[lazpete]

Quote:

Originally Posted by morfic /img/forum/go_quote.gif You ought to try some CV4014 Mullards.

tried to chase some down, hade some probs tough. Any idea where to find a pair or two?

Br
Lasse

 

[lazpete]

I'e accually reveresed my setup, and went back to the russian EF95's, and at first glance, listening to a live recording with Ryan Adams, they are a little "brighter". It might be the fact that they are not warmed up properly. The low end is definetly there.

One thing ive been thinking about, is how tubes that are used and burned in, react to be used again, a few weeks later. Do they still need to be burmed in again, or is warming up sufficient?

Br
Lasse