[nysulli]

Quote:

Originally Posted by Pars Actually, with a mini3-fied PINT, because of the lack of input caps and the bipolar opamp(s), the amp will see a variable resistance from the pot on its input. The offset will therefore climb as the volume is increased (because the input bias currents change with pot level). <20mV at full volume isn't too bad.

i realize that without those caps you have a variable resistance on one side of the opamp, and thus its pretty difficult to balance the input and output sides of the opamp, but even so, i would think one could do much better balancing the load on either side to not have dc offset vary from 1mv to 20mv, even being a noob builder when i made my minified pint, i was able to sort out my configuration such that my offset varies from .5mv to at most 1mv at full volume, doubling of the offset i can see, but varying 2000% over the full swing of the volume pot just doesn't sound right, if it was 10mv to 20mv, i wouldn't think that much of it, its just the order of magnitude change that makes me wonder

 

[Filburt]

Quote:

Originally Posted by Andrea wow, well, seems just like what I'm after, too... we have a common comparison term in the LM6172, which, when push comes to shove, we don't seem to 'hear' that differently! I'll let you know my impressions.

Okay, great! I hope you enjoy it. I'm also liking being able to use 4 AAs/AAAs instead of 9V batteries since I have a bunch of rechargeables. I couldn't hear any sonic difference between 6V and 9V, but it does use slightly more power at 6V which I think is common.

I haven't had trouble with oscillation on it; just make sure to have proper bypassing and hopefully you'll be OK. It seems to be less cranky than the 8397, at least, in my experience. I should probably switch to having another 8058 on the ground since it is so much faster than the 8397, but it does seem to run fine as-is. Not sure what I'll use on the ground if I change it

 

[Andrea]

I'm still totally uncertain, to say the truth - the LMH6654 has the high current, good speed, and low distortion to work fine as a buffer; besides, it would seem to tolerate a 12V power supply (as needed for the preamplifier op-amp - which I'd choose between AD8620, AD746, and leaving the standard OPA2134 - to give appropriate performance) better than the AD8057, having a 13.2V "absolute maximum", vs. 12.6V. I wonder if the latter observation is truly significant?

 

[Filburt]

Quote:

Originally Posted by Andrea I'm still totally uncertain, to say the truth - the LMH6654 has the high current, good speed, and low distortion to work fine as a buffer; besides, it would seem to tolerate a 12V power supply (as needed for the preamplifier op-amp - which I'd choose between AD8620, AD746, and leaving the standard OPA2134 - to give appropriate performance) better than the AD8057, having a 13.2V "absolute maximum", vs. 12.6V. I wonder if the latter observation is truly significant?

The "absolute maximum" by AD I've found to generally be conservative. I'd suspect it won't matter and I wouldn't worry about.

 

[Andrea]

Hmmm, ok, so it's a tie again...

I'd love to try the AD8057 firstly in the Guinea Pig Xenos (to see how powerful, and full sounding with vocals it is), but... I don't have the heart to abuse it anymore

and I also have two of them AD's only!

 

[Filburt]

Quote:

Originally Posted by Andrea Hmmm, ok, so it's a tie again... I'd love to try the AD8057 firstly in the Guinea Pig Xenos (to see how powerful, and full sounding with vocals it is), but... I don't have the heart to abuse it anymore and I also have two of them AD's only!

Hehe! Oh well, I guess you'll have to wait for the other Xenos then?

You may want to get an AD8058 in case the 8057s don't work as well as a buffer but maybe the 8058+6654 would work well...*shrug*. I'm still having no troubles over here with in the PINT, though.

 

[Andrea]

The LMH6654 is a fantastic chip for headphone driver purposes.

The AD746 (as preamplifier op-amp) has let me fully appreciate it... (see here, http://www4.head-fi.org/forums/showt...=196170&page=8 )

 

[Filburt]

Hmm, I'll have to try the AD746 some time (presumably w/ buffer). Hey, can you also try the AD9631 if you have a chance? Spec'd at 50mA output current and 130mA short-circuit. Seems to drive lower impedance loads pretty well, and is low distortion, while still being very low open-loop gain for an op-amp. May be worth trying and comparing to AD8057.

 

[Andrea]

Wow... and I thought that the LMH6654 was too wide-bandwidth

jk

Definitely interesting numbers... but that 17 mA quiescent current for a single chip SOIC package is borderline... could be too much for my Xenos, for lack of proper heat dissipation. Am I overly cautious? Also, comparing datasheets, the LMH6654 could have a little less distortion when gain is +2, like in the Xenos.

 

[Andrea]

The chip itself is very promising. In which kind of existing design would you use it?

 

[Filburt]

AD8057/58 has higher bandwidth than the 9631, so I think we're OK considering this chip looks as though it may be geared towards at least as good, if not better, stability.

The Iq is definitely high, but I don't think it's *too* high necessarily. AD8397 is spec'd to have a max Iq of nearly that per channel. I think you might be OK on heat; only way to find out, though, is to try it! ;D

The harmonic distortion on the 6654 might be slightly lower in some spots, but I strongly suspect it has significantly higher quantities of other sorts of distortion caused by its much slower slew rate and much greater dependence on feedback. I'm really not concerned with the differences here, but maybe it's something for you to worry about. I'm much more concerned about other difference where I think the 9631 will be significantly better (such as TIMD).

I'm not sure what I'll use these for. I purchased boards from dsavitsk for his AD1865 NOS DAC to try it out, and maybe I'll use them for the gain stage or I/V stage there.

I may also try doing using them as a sort of buffer/output stage as your Xenos uses.

I'm also considering trying the AD8055/56.

Yeah I know, these are fast chips. Honestly, I haven't had any trouble making them work. Maybe I build unusually robust and stable circuits, or maybe there is some alarmism going around about fast chips. I suspect I'm not competent enough to pull off the former, so I'm thinking maybe the latter is more probable

 

[Andrea]

I know well the AD8056... I had no problems with using it even in a basic design such as the Zero Audiocraft, where I used it for the output buffer position.

Sound wise, it's alright. Not the last word in naturalness & balance... I preferred the LT1361 in the same amp, it felt both clearer and more natural. To be honest I think the LMH6654 is much better! Of course the AD8055/6 are not quite like the AD8057/8.


I don't know if the AD9631's lower dependence on feedback due to the lower open loop gain matters - graphs just show that in gain of +2 it distorts IIRC around 5-10 dB more than the LMH6654 at 100 KHz - particularly the 3rd order distortion. And that with a more than triplicated quiescent current. Not very efficient

Besides, the LMH6654 is higher current -- it can output up to 150 mA (if not continuously). And its slew rate, in my perspective, is already overkill for audio.

Note this -- National's new LM4562 is only 20V/us, and yet its IMD is specified as lower than 0.00005% IIRC.