Hello there,
I recently bought a HIFI sound card which came with 3 opamps and a dedicated headphone out for high impedance cans.
Ok, I tried it and it improves the sound vs. motherboard. But there is some very thick sound and the high frequencies are muffled.
So I skipped the headphone dual opamps and connected the headphone to RCA out. Better high frequencies and a slight loss of bass.
So I connected the headphones in balanced mode straight to the dac chip opamps. High frequencies came back to full power and muddiness gone.
Long post to say... is this a conspiracy that designers sabotage good sound with so many opamps when only 1 or 2 are required???
I recently bought a HIFI sound card which came with 3 opamps and a dedicated headphone out for high impedance cans.
Ok, I tried it and it improves the sound vs. motherboard. But there is some very thick sound and the high frequencies are muffled.
So I skipped the headphone dual opamps and connected the headphone to RCA out. Better high frequencies and a slight loss of bass.
So I connected the headphones in balanced mode straight to the dac chip opamps. High frequencies came back to full power and muddiness gone.
Long post to say... is this a conspiracy that designers sabotage good sound with so many opamps when only 1 or 2 are required???
Designers have to use lots of opamps because the factories which produce them cannot reduce production without creating problems with yield. If people used fewer opamps then there would be opamp mountains building up around the world. So they add extra ones as fixed tone controls (otherwise known as filters) and hope the user is wise enough to bypass the ones he doesn't need.
Some more enlightened designers do the same thing with rare NOS valves. In order to stop them piling up in random places they add them to their circuits (e.g. as unnecessary buffers before chip amps). This restores the equilibrium and prevents the world from wobbling too much.
Some more enlightened designers do the same thing with rare NOS valves. In order to stop them piling up in random places they add them to their circuits (e.g. as unnecessary buffers before chip amps). This restores the equilibrium and prevents the world from wobbling too much.
So true, they 'soften' the too perfect sound of chip amps by adding 'tube warmth'. this is so true everywhere DF96.
Billshury do you realize the difference is minor from all the opamps rolling??? Each opamp adds layers of stuff... There is no need for a headphone socket unless you listen to 120 DB and have loss your hearing.
The worst I have seen is mods with 'discrete' opamps after other opamps or in the middle of a chain of opamps with no regards for dac circuit.
There is forums with dozens of pages of posts on how the opamp rolling made a difference. Really? They don't make any!!
To give substance to my argument, the ad8620 can drive 1K load with capacitance or inductance +-1.5 V of the power supply with 0.01% distortion!!!!!!
The worst I have seen is mods with 'discrete' opamps after other opamps or in the middle of a chain of opamps with no regards for dac circuit.
There is forums with dozens of pages of posts on how the opamp rolling made a difference. Really? They don't make any!!
To give substance to my argument, the ad8620 can drive 1K load with capacitance or inductance +-1.5 V of the power supply with 0.01% distortion!!!!!!
You are joking right ?
The manufacturers (TI, Maxim, etc) have dozens of specialized headphones drivers. It's not their fault if soundcard makers are using sub-optimal ICs at the wrong place.
I am sure the designers have different considerations than an audiophile has. They have to use a separate opamp for driving the most common 30 ohms headphones. They have to use another buffer opamps to drive the the capacitance of longer cables attached to the RCA. Clearly attaching 600 ohms headphones directly to the DAC could give better sound, but the designer has to consider all common cases.
I disagree, the single opamp can drive multiple connections in parallel. Normal opamps can drive the capacitance of long cables of RCA.
I know many people would not believe me, In my setup I bypassed the output headphones 2x dual opamps and connected 300 ohms to regular RCA outputs (many claim 300 ohms headphones can't be driven with regular RCA outputs... are you joking?)
Tests: 1st db value is RCA modified to AD8620, 2nd value is directly from the DAC modified opamps LM456.
Tests volume 14 mv (0.7 uW) x2):
2000Hz + 2050hz : -38.8db VS -43.1 db, 2nd: -39.7db VS none , 3rd: -41.6 VS -50.8
20Hz relative to 1khz : -0.25db
1khz : 2nd 0 both , 3rd -47.3 VS -49.1 4th : 0 both; 5th : -47.5 VS -48.7
Subjectively through the headphone outputs the sound is muffled and speech hard to understand with very poor high frequencies.
Through RCA: speech clears up as well as high frequencies
Through DAC differential opamps : release fully the high frequencies and sounds loud with substance and full body, clear, bass comes alive, sounds overwhelming. Lyrics are very easy to follow.
I know many people would not believe me, In my setup I bypassed the output headphones 2x dual opamps and connected 300 ohms to regular RCA outputs (many claim 300 ohms headphones can't be driven with regular RCA outputs... are you joking?)
Tests: 1st db value is RCA modified to AD8620, 2nd value is directly from the DAC modified opamps LM456.
Tests volume 14 mv (0.7 uW) x2):
2000Hz + 2050hz : -38.8db VS -43.1 db, 2nd: -39.7db VS none , 3rd: -41.6 VS -50.8
20Hz relative to 1khz : -0.25db
1khz : 2nd 0 both , 3rd -47.3 VS -49.1 4th : 0 both; 5th : -47.5 VS -48.7
Subjectively through the headphone outputs the sound is muffled and speech hard to understand with very poor high frequencies.
Through RCA: speech clears up as well as high frequencies
Through DAC differential opamps : release fully the high frequencies and sounds loud with substance and full body, clear, bass comes alive, sounds overwhelming. Lyrics are very easy to follow.
"Rolling IC's"
In either a feedback power supply or a PPM line buffer/headphone amp ,
IC choice has a very minor effect.
On the PS - ne5532= -117db , OPAxxxx = <125db (Jung super-reg).
On the amp - ne5532= 10ppm , OPAxxxx = .5ppm - who cares ?
Would the difference of 9V/us versus 125V/us make a difference ?
Or if you used a 2000V/us buffer for the headphone amp ?
The buffer would make difference in THD @ 30R ! 300R is where most
8 pin IC amps really shoot up in THD ...
If you "roll" the wrong op-amp into a circuit optimized for a specific one
in a lazy,cheap chinese design , it could even oscillate.
"The sound" ... yes , If you drive your load with the chip. Buffer the
IC with discrete's and you would not hear much of a difference.
OS
In either a feedback power supply or a PPM line buffer/headphone amp ,
IC choice has a very minor effect.
On the PS - ne5532= -117db , OPAxxxx = <125db (Jung super-reg).
On the amp - ne5532= 10ppm , OPAxxxx = .5ppm - who cares ?
Would the difference of 9V/us versus 125V/us make a difference ?
Or if you used a 2000V/us buffer for the headphone amp ?
The buffer would make difference in THD @ 30R ! 300R is where most
8 pin IC amps really shoot up in THD ...
If you "roll" the wrong op-amp into a circuit optimized for a specific one
in a lazy,cheap chinese design , it could even oscillate.
"The sound" ... yes , If you drive your load with the chip. Buffer the
IC with discrete's and you would not hear much of a difference.
OS
The specs will get better with 30 ohms load. The higher the impedance the more difficult for transistors
Wow I was so taken by this I had to count up the number of op amps in my Ecler Sclat 100 mixer.
236 op amp stages.
I built a Cmoy style headphone amp but with LM317/337 PSU, and far too much headroom to be safe......
Sounds great into high impedance headphones like my Beyer DT100. Terrible with low impedance headphones designed to run off low voltage equipment.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.