Here is the op amp circuitry in the O2 headphone amp:
When the amp is set to 1X gain what is the purpose of the first stage with U1?
This amp is pretty much just a good power supply powering some op amps, the idea is that op amps distortion levels are much lower and consistent than discreet designs, the main problem is low power output so 2 high current op amp are used in parallel. it does sound very clean but a bit thin and 2D.
To minimise signal degradation as much as possible and maybe even improve sound quality I was thinking about bypassing the U1 section entirely and inputting straight into the pot since the output buffers are actually driving the headphones.
The OPA1612 sounds very clean and accurate in the U1 position and but even it has some degree of distortion, having none at all would be ideal but is it possible? I can just put some jumpers on the inputs and outputs of the U1 socket to test it but just want to hear some expert opinions on it first.
When the amp is set to 1X gain what is the purpose of the first stage with U1?
This amp is pretty much just a good power supply powering some op amps, the idea is that op amps distortion levels are much lower and consistent than discreet designs, the main problem is low power output so 2 high current op amp are used in parallel. it does sound very clean but a bit thin and 2D.
To minimise signal degradation as much as possible and maybe even improve sound quality I was thinking about bypassing the U1 section entirely and inputting straight into the pot since the output buffers are actually driving the headphones.
The OPA1612 sounds very clean and accurate in the U1 position and but even it has some degree of distortion, having none at all would be ideal but is it possible? I can just put some jumpers on the inputs and outputs of the U1 socket to test it but just want to hear some expert opinions on it first.
The front end has some gain so you would lose that if you bypassed it.
It also has an RF filter which I would leave in to stop radio interference.
It also has an RF filter which I would leave in to stop radio interference.
ok I see, my version has as an open contact on 1 switch position for 1x gain, should have clarified.
tested and it sounds very good, much more natural, the best its ever sounded.
also tried bypassing the RF filter to avoid having a resistor in the signal path, sounds very slightly different (maybe worse) or no different. going to leave it and enjoy the new sound for now.
tested and it sounds very good, much more natural, the best its ever sounded.
also tried bypassing the RF filter to avoid having a resistor in the signal path, sounds very slightly different (maybe worse) or no different. going to leave it and enjoy the new sound for now.
Last edited:
It seems like a lot of silicon in the way, if you just need a buffer.
I once did it using a single BUF634, open loop, with a simple servo for DC offset correction.
Signal goes through 2 BJT junctions, from input to HP.
I once did it using a single BUF634, open loop, with a simple servo for DC offset correction.
Signal goes through 2 BJT junctions, from input to HP.
Headphones are all over the place in terms of impedance, sensitivity... Creating an "universal" design such as the O2 forces you to plan for some gain. If you build for just one pair of headphones, it's another thing.
I see, its a case of versitility over sound quality. There is very little reason to buy a new headphone amp when this can be easily modified to be much better. Its crazy how simple this stuff really is.
using 4 OPA1612s in parallel per channel dramatically increased sound quality and gave a significant increase to current output compared to the 2 high current NJM4556s.
powering the op amps directly from the regulators made huge improvements to sound, for desktop use is there really any reason not to power them this way? regulator were heat sinked first but their barely even warm
Now im interested about adding a push/pull output stage. AFAIK this is only thing that distances this amp from an extremely powerful and extremely expensive op amp based amp like the Violectric V200/280
In the most extreme scenario could all 4 op amps have their own pair of transistors, totaling 4 NPN and 4 PNP per channel?
using 4 OPA1612s in parallel per channel dramatically increased sound quality and gave a significant increase to current output compared to the 2 high current NJM4556s.
powering the op amps directly from the regulators made huge improvements to sound, for desktop use is there really any reason not to power them this way? regulator were heat sinked first but their barely even warm
Now im interested about adding a push/pull output stage. AFAIK this is only thing that distances this amp from an extremely powerful and extremely expensive op amp based amp like the Violectric V200/280
In the most extreme scenario could all 4 op amps have their own pair of transistors, totaling 4 NPN and 4 PNP per channel?
Last edited:
It's crazy how simple this stuff really is, yet numerous basic aspects of the O2 design seem to have passed you by.
I suggest to do a little homework before further second-guessing of the O2 project.
You might start here: The Objective2 (O2) Headphone Amp DIY Project
Dave.
Last edited:
thanks for the suggestion, you are right. Im learning on a need to know basis on here but it isnt very practical. I bought my O2, didnt build it as it seemed impossibly complicated at the time and then gradually I went from swapping op amps a few months ago to having a (very) basic understanding of every part of the circuit, which seems crazy now.
Actually, the circuit isn't crazy. It's very logical. If you assume that the distortion induced by the first opamp is negligible (which was the designer starting point), it makes a lot of sense to have the volume control in between two stages.
First of, it allows for a well defined input filter. If the volume pot was before everything, it would interact with the filter.
Then, if you need gain, it's better to have it on a lightly loaded opamp. Distortion goes down as load goes up while increasing gain reduces the amount of feedback you can apply. It also means that you can run the output opamp at unity/low gain, to maximize feedback around the opamp working harder.
First of, it allows for a well defined input filter. If the volume pot was before everything, it would interact with the filter.
Then, if you need gain, it's better to have it on a lightly loaded opamp. Distortion goes down as load goes up while increasing gain reduces the amount of feedback you can apply. It also means that you can run the output opamp at unity/low gain, to maximize feedback around the opamp working harder.
- Status
- Not open for further replies.