Hi there all,
Do op amps performance improve with higher operating voltages.
In a cd player I have they run at +/-5V. Would there be a sonic improvement if I run them at +/-15V.
I've been told mixed answers and I am trying to clear this detail up before I make the modification
Many thanks
Paul
Do op amps performance improve with higher operating voltages.
In a cd player I have they run at +/-5V. Would there be a sonic improvement if I run them at +/-15V.
I've been told mixed answers and I am trying to clear this detail up before I make the modification
Many thanks
Paul
Most normal opamps have slightly better performance at higher voltage but +- 5 volts is nowadays not a very low voltage. The AD8610 (my favorite at the moment) has excellent performance at low voltage.
If you interested in a particular opamp, check the datasheets first.
If you interested in a particular opamp, check the datasheets first.
The op amps in question are NE5534's. From what I hear they're quite good for HIfi but I might change them if anyone can advise a good alternative.
I've read the data sheets and other stuff but could find nothing documenting performance to op voltage.
I've read the data sheets and other stuff but could find nothing documenting performance to op voltage.
I don't think the 5532/34 cares much about supply voltage. The one opamp that I know for sure that improves noticeably at higher voltages is the OPA2604. Run it at +/-22V and it rivals the best I have heard. Run it at +/- 15V and it is merely 'good'.
You could try replacing the NE5534 with OPA627.
I also like the LM6171 very much (properly bypassed!).
But at +/-5V, don't even think of that.
You are very limited with that voltage.
You shoud first find a way to rise that voltage.
I think your cd-player has only one PSU for digital and analogue sections, and that's why you only have +/-5V.
You could make a small +/- regulated PSU just for the op-amps.
The improvement would not be small, even with the NEs.
I also like the LM6171 very much (properly bypassed!).
But at +/-5V, don't even think of that.
You are very limited with that voltage.
You shoud first find a way to rise that voltage.
I think your cd-player has only one PSU for digital and analogue sections, and that's why you only have +/-5V.
You could make a small +/- regulated PSU just for the op-amps.
The improvement would not be small, even with the NEs.
I didnt read the link to the last post but I do know this. I have an active xover that uses tl07x. They are operated from batteries on approx +- 7 volts. When the batteries are running out the voltage can drop down to +- 2-3 volts. I hear no difference in sound quality until the batteries are completely dead when the whole shebang fisrt looses bass and then very quickly distorts. This loss in quality happens extreamely quickly. I was listening to a CD and thinking how brilliant it sounded (I hate distortion by the way so wouldnt have thought it sounded good if distortion was present). The CD ended I got up changed CD and in that time alone the sound went from Good to Bad. Ofcourse this is only from 2-7 volts etc and I dont know what they sound like using say +-25volts. But I have substituted in OPA627's for a comparison in the Mid range and trebble section and noted an improvement in transparancy and all those micro things that only a trained ear can hear. But the 627 did the same thing sound remained constant until the very end. Oh and the 2604's is the loss in sound quality over the whole frequency range or just the higher spectrum. Im going to rebuild the Active xover permenantly using 627s for bufferes and mid/trebble and then 2604's on the bass, still off the batteries. Does the lower operating voltage affect bass? I mean that might be hard to judge because if you used the 2604 in a preamp and it changed the higher frequencies this has been known to affect bass.
Cheers Matt
Cheers Matt
Batteries and voltage
Matt,
You won't get to any conclusion because of the low voltage you're using in that active crossover.
If you raised the voltage to something between +/-15 to +/-18volts, then you would hear improvements.
But with those TL07x I really don't know, I simply don't use them.
I have made a very simple headphone amp, for my tests, with an LM6172.
It works with 4 9v alkaline batteries in series.
In fact, it plays very well, but as you can see, I don't cut on voltage.
It's important.
It could play with only 2 9v batteries.
But the results would not be the same.
And yes, low voltage normally affects bass performance.
Matt,
You won't get to any conclusion because of the low voltage you're using in that active crossover.
If you raised the voltage to something between +/-15 to +/-18volts, then you would hear improvements.
But with those TL07x I really don't know, I simply don't use them.
I have made a very simple headphone amp, for my tests, with an LM6172.
It works with 4 9v alkaline batteries in series.
In fact, it plays very well, but as you can see, I don't cut on voltage.
It's important.
It could play with only 2 9v batteries.
But the results would not be the same.
And yes, low voltage normally affects bass performance.
Non linear capacity
It's a fact that non linear parasitic capacity gets more constat and gets lower value as voltage rises, its very usual in every transistor, doesn't matter it's inside a IC or in a discrete way.
It gets worse as its sometimes multiplied by miller effect, if cascode is not used (not in TL, OPAs do), and it causes distortion at high frecuency, and IMD. Also more voltage means less capacity , and it means more speed and less TIM and slew-rate limit.
Power consumption will increase as voltaje rises, so thermal noise will also increase. If noise figure is low, it won't be noticed.
For Vbe and Cbe a similar thing happens, the more current, the more linear, but it cannot be controlled with voltage.
It's a fact that non linear parasitic capacity gets more constat and gets lower value as voltage rises, its very usual in every transistor, doesn't matter it's inside a IC or in a discrete way.
It gets worse as its sometimes multiplied by miller effect, if cascode is not used (not in TL, OPAs do), and it causes distortion at high frecuency, and IMD. Also more voltage means less capacity , and it means more speed and less TIM and slew-rate limit.
Power consumption will increase as voltaje rises, so thermal noise will also increase. If noise figure is low, it won't be noticed.
For Vbe and Cbe a similar thing happens, the more current, the more linear, but it cannot be controlled with voltage.
Attachments
- Status
- Not open for further replies.