CHANGSTAR: Audiophile Headphone Reviews and Early 90s Style BBS

  • December 31, 2015, 11:20:03 AM
  • Welcome, Guest
Please login or register.

Login with username, password and session length
Advanced search  
Pages: 1 [2] 3 4

Author Topic: The Opamp Thread  (Read 9053 times)

0 Members and 1 Guest are viewing this topic.

Solderdude

  • Grab the dScope Kowalski!
  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +206/-4
  • Offline Offline
  • Posts: 907
  • No can do skipper, the dScope was terminated
    • DIY-Audio-Heaven
Re: The Opamp Thread
« Reply #10 on: April 13, 2013, 06:43:44 PM »

See, to drive the worst case Hifiman HE-6 to 115 dB (a reasonable level for a hard impact), you need about 2 W power, therefore ~130 mA at 15V (+/-7.5V) (2W) per channel. Almost all opamps cannot do this unbuffered. It takes just about every ounce of power out of BUF634. Of course you'll also need near zero output impedance too.

To drive the HE-6 to 115dB you need 6.3W. This means 18Veff. For this you need an amp with at least +/-28V rails. A 70W/4Ohm amp is capable of this. To reach 6W the output stage must be able to provide 500mA peak.

But to drive something easier like HE-500, you only need ~30 mA at 15V (600 mW), which quite a few opamps can manage without a dedicated buffer, but not that AD8597, which can only source ~18 mA. Some opamps can source more current at lower voltages, but it is not one of them.

To drive the HE-500 to 115dB you need 630mW. This means 5Veff. For this you need an amp with at least +/-9V rails. to reach 630mW the output stage must be able to provide 140mA peak.

One of the few opamps that can deliver this (without needing an additional buffer) is the OPA551/552. Even the NJM4556 is limited to around 100mA where most opamps manage around 20 to 40mA and often not even symmetrical.

Numbers change when the 115dB level is abandoned and a good 105dB (average level 95dB) is needed.

To drive the HE-6 to 105dB you need 0.6W. This means 5.6Veff. For this you need an amp with at least +/-9V rails. to reach 0.6W the output stage must be able to provide 160mA peak.
To drive the HE-500 to 105dB you need 63mW. This means 1.5Veff. For this you need an amp with at least +/-4.5V rails. to reach 63mW the output stage must be able to provide 56mA peak.

In both cases your average opamp won't be able to cut it.

I won't burn my fingers on the subject of 'sound'.
« Last Edit: April 13, 2013, 07:01:03 PM by Solderdude »
Logged
Use your ears to enjoy music, not as an analyser.

AstralStorm

  • Speculation and Speculums
  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +250/-164
  • Offline Offline
  • Posts: 559
  • Warning: causes nearby electronics to go haywire
Re: The Opamp Thread
« Reply #11 on: April 13, 2013, 10:03:27 PM »

You're right about HE-6. I though it is more efficient. It's actually 83.5 dB/mW not 86 dB/mW I used. It's also 50 Ohm, where I used 40 Ohm.
What you said is why I actually want to drive opamps in my device-to-be off +/- 15V giving about 12.5V swing from the ADA4898's. It is doable in a portable device even with required reasonably high current capability, just more expensive and a bit heavier. (I'm about to do that as well.)

I wasn't too much off with HE-500 though, except in the current calculation I used short circuit current - where the actual clipping happens. (of course I didn't tell that, did I?  :)p12 ). Yes, the Leckerton obviously won't do for HE-6, since it does not have a differential output to double the swing and even ADA4898 or single BUF634 would current clip.  Would have to stack a pair at the very least. AD797 would not clip if not for the power supply, it has very high short circuit current rating. (250 mA AFAIR)

I actually forgot though about one most important thing. Neither of the opamps is R2R. In UHA-6s mkII, OPA209 is limited to +/-4.3V and ADA4898-1 is limited to +/-4.1V. Either is just short of driving HE-500 with full 105 dB dynamic range, but still reasonable. Way off for any drivability of HE-6.

Still probably loud enough. I always wondered where that recommended peak 110 dB figure came from. Do we like to listen to thunderstrikes via headphones or something? Even a full percussion or orchestra hit shouldn't be that loud - that is a painful level. Even the most dynamic music has at most 60 dB dynamic range, so a comfortable quite loud level of about 60 dB SPL average would require up to 90 dB SPL peak. Averaging at 80 dB SPL is dangerous to hearing - but let's not derail the thread to there.

Remember that the current output of the opamp as rated is for constant drain at the rated supply voltage, which is often +/-15V for historical reasons. When you drive it lower, the average capability can be somewhat higher, for example, OPA209 which is rated at 25 mA can source 40 mA at +/-12V supply, bit more at even lower supplies.

I have a few reservations though: care to show the Veff equation you used? Plain old RMS, yes? But we were talking about peak drive.
I've of course assumed near zero output impedance. Leckerton is actually rated (and measured) at 0.4 Ohm, not a really important difference.

--
I need to ramble less and get more to the point.
« Last Edit: April 13, 2013, 10:57:34 PM by AstralStorm »
Logged
For sale: Hifiman HE-500; Paradox; Brainwavz B2. PM me if you would like to buy them.

fishski13

  • Mate
  • Pirate
  • ****
  • Brownie Points: +79/-1
  • Offline Offline
  • Posts: 424
Re: The Opamp Thread
« Reply #12 on: April 14, 2013, 12:20:06 AM »

it appears Solderdude's numbers for the HE-6 are based on the measurements from IF, http://www.innerfidelity.com/images/HiFiMANHE6.pdf , which puts the sensitivity at 77dB/mW.  there's probably some manufacturing variances though.

his Veff is RMS, but the peak current calcs are based on Vp-p.

the 120dB benchmark is usually found in powered pro monitors for deaf rock musicians who want the mix played back at concert levels.  also, you may want some extra headroom for speakers in larger rooms if you're sitting at a greater distance from the speakers.  105dB is reasonable for HPs as it gives you some headroom for quieter mastered recordings - 5-8dB should be enough even for early 80's mastered digital recordings.  also, unlike speakers, the driver distance remains the same. still, anything greater than 95dB is too freaking loud for my tastes.
Logged

Solderdude

  • Grab the dScope Kowalski!
  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +206/-4
  • Offline Offline
  • Posts: 907
  • No can do skipper, the dScope was terminated
    • DIY-Audio-Heaven
Re: The Opamp Thread
« Reply #13 on: April 14, 2013, 07:54:20 AM »

care to show the Veff equation you used? Plain old RMS, yes? But we were talking about peak drive.

fishski13 answered the question perfectly !
Logged
Use your ears to enjoy music, not as an analyser.

Anaxilus.

  • Dikus Beligerantis Analmorticus
  • Pirate
  • **
  • Brownie Points: +65535/-65535
  • Offline Offline
  • Posts: 577
Re: The Opamp Thread
« Reply #14 on: July 09, 2013, 08:09:53 AM »

Posting my recent opamp impressions:

4898-weighty, thick, and rich, maybe not as much as 627.  Organic sounding.  Better blackground, than 209 but less detailed/resolving, less plankton.  209 digs deeper.  Vocals less dry than 1611 but still a hint of dryness. A bit muddled at times, but still decent separation. Slightly bigger sounding bottom end.

209-Analytical signature, digs deep.  Easier to hear recording noise and artifacts.   String plucks and applause is more thin and brittle.  Slight haze/veil despite picking up tons of detail.  More blended, less separation and articulation of dynamics.  Overall less refined signature than the others.  Vocals less dry but could also because it stands out less in overall signature which is more coherent and balanced than the others.  Therefore, any vocal issues could stand out/apart less.  I do think vocals are more liquid than the others in actuality on more listening. 

4627-Better dynamics and note articulation than 209. Perhaps best macro and micro dynamic balance of the bunch w/ lots of clear inner detail.  Excellent timbre!  Clearer, blacker than 209 w/ better separation and imaging.  Widest SS of all.  Less recording noise emphasis than 209 but actually more resolving.  Superior resolution, excellent spatial precision, best blackground, fantastic macro and micro dynamics. Most natural and accurate timbre.  The clear cut overall winner.

1611-Very fast, best of the bunch. Analytical type sig. Decent timbre, not as good as 4627 or 4898 though, better than 209.  Not sure why, could be a minute extra bit of treble brittleness, ringing type treble sound.  Driest vocals.  Wider SS than 4898.  4627 has cleaner, clearer note articulation and more accurate timbre.  1611 might have sharper, more precise sound but less blackground and timbre is less accurate and fleshed out. 

Note: All opamps have more than sufficient run time w/ the exception of the OPA1611 which has the fewest by far.
 
« Last Edit: July 09, 2013, 06:58:45 PM by Analixus »
Logged
If you do not change direction, you may end up where you are heading - Lao Tzu

CEE TEE

  • Master controller of all scores
  • Mate
  • Pirate
  • ****
  • Brownie Points: +98/-338
  • Offline Offline
  • Posts: 839
  • Need More Time To Loaf Around
Re: The Opamp Thread
« Reply #15 on: July 09, 2013, 05:41:24 PM »

^Thank you, I've been waiting for something like this so that when I finally try the Leckerton I know which opamps you have been listening to. 
Logged
sound soft harmonics rich bile rich rhyme

AstralStorm

  • Speculation and Speculums
  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +250/-164
  • Offline Offline
  • Posts: 559
  • Warning: causes nearby electronics to go haywire
Re: The Opamp Thread
« Reply #16 on: July 09, 2013, 06:36:46 PM »

Any GSM interference issues of note? OPA209 was atrocious in this regard... I could grab OPA1611 for a spin, as well as actual AD797s.
Logged
For sale: Hifiman HE-500; Paradox; Brainwavz B2. PM me if you would like to buy them.

Anaxilus.

  • Dikus Beligerantis Analmorticus
  • Pirate
  • **
  • Brownie Points: +65535/-65535
  • Offline Offline
  • Posts: 577
Re: The Opamp Thread
« Reply #17 on: July 09, 2013, 06:58:11 PM »

Any GSM interference issues of note? OPA209 was atrocious in this regard... I could grab OPA1611 for a spin, as well as actual AD797s.

I did run into this at The SHOW w/ the 4627 when I was in a particular room and couldn't get a signal.  For the most part, it I care enough about SQ at any given moment I activate power saving and disable my mobile data which helps quite a bit for me.
Logged
If you do not change direction, you may end up where you are heading - Lao Tzu

fishski13

  • Mate
  • Pirate
  • ****
  • Brownie Points: +79/-1
  • Offline Offline
  • Posts: 424
Re: The Opamp Thread
« Reply #18 on: July 16, 2013, 02:50:50 AM »

Posting my recent opamp impressions:

4898-weighty, thick, and rich, maybe not as much as 627.  Organic sounding.  Better blackground, than 209 but less detailed/resolving, less plankton.  209 digs deeper.  Vocals less dry than 1611 but still a hint of dryness. A bit muddled at times, but still decent separation. Slightly bigger sounding bottom end.

209-Analytical signature, digs deep.  Easier to hear recording noise and artifacts.   String plucks and applause is more thin and brittle.  Slight haze/veil despite picking up tons of detail.  More blended, less separation and articulation of dynamics.  Overall less refined signature than the others.  Vocals less dry but could also because it stands out less in overall signature which is more coherent and balanced than the others.  Therefore, any vocal issues could stand out/apart less.  I do think vocals are more liquid than the others in actuality on more listening. 

4627-Better dynamics and note articulation than 209. Perhaps best macro and micro dynamic balance of the bunch w/ lots of clear inner detail.  Excellent timbre!  Clearer, blacker than 209 w/ better separation and imaging.  Widest SS of all.  Less recording noise emphasis than 209 but actually more resolving.  Superior resolution, excellent spatial precision, best blackground, fantastic macro and micro dynamics. Most natural and accurate timbre.  The clear cut overall winner.

1611-Very fast, best of the bunch. Analytical type sig. Decent timbre, not as good as 4627 or 4898 though, better than 209.  Not sure why, could be a minute extra bit of treble brittleness, ringing type treble sound.  Driest vocals.  Wider SS than 4898.  4627 has cleaner, clearer note articulation and more accurate timbre.  1611 might have sharper, more precise sound but less blackground and timbre is less accurate and fleshed out. 

Note: All opamps have more than sufficient run time w/ the exception of the OPA1611 which has the fewest by far.

my opamp rolling is limited to the AMB M3.  interesting to compare impressions, even with a MOSFET output buffer.

despite the fact that the OPA1611 sounds "fast", with excellent attack and decay, it can't carry a tune in the M3.  it sounds very ethereal but a bit dry/chalky and boring, lacking dynamics and the ability to carry a tune.  it sounds "slow" to me in the musical sense.  keeping the OPA1611 in L/R and changing the G channel to the ADA4898 improves things though.  a bit more timbre and dynamics.   

the ADA4627-1 has excellent timbre, speed, and can carry a tune.  transients and micro-detail/plankton are a little more obvious.  very detailed.  L/R/G in the M3 and can't complain.   
Logged

kiteki

  • Guest
Re: The Opamp Thread
« Reply #19 on: August 11, 2013, 07:09:09 AM »

Has anyone actually directly compared OPA627 and ADA4627-1?

The price difference is immense.  Hypothetically, let's say it's $25, if a manufacturer is going to ship 1000 DAC / Amp / PMP's, that'll cost them $25,000 extra, well $50,000 if you need stereo.

In China, you can now see thousands of so called OPA627 IC's selling for something like $1.50.  I received two for free recently, and they sounded like ####.

An interesting note from diyaudio...

"Test on sonic difference,
Anyway, I made listening test on ADA4627 Brz and 4637 Arz.
Sonically they are not much difference. Check the spec, 4627 has a 18 V/us slew rate, normally this should be enough for an audio purpose OP. 4637 has much faster slew of 100s. My feeling is the resolution of 4637 is slightly better,but the major difference is the sense of the instrument locations. It is more alive and sharper image for 4637. In other words, 4627 is more flat. I vote for the 4637, even it is a cheaper A version used."

http://www.diyaudio.com/forums/solid-state/150222-ada-4627-ada4627-1-outperform-burr-browns-opa-627-opa627-2.html#post2883823
« Last Edit: August 11, 2013, 07:42:17 AM by kiteki »
Logged
Pages: 1 [2] 3 4