100mA at +-10Vdc would cover most Headphone options.
Some headphones are specified for absolute maximum 5Vpk
Some headphones are specified for absolute maximum 5Vpk
thanks for the pointer. the 2SC6144 and 2SA2222 look like they would be nice devices for experimenting in regulators.
fyi, they are in stock and similarly inexpensive at mouser also.
mlloyd1
fyi, they are in stock and similarly inexpensive at mouser also.
mlloyd1
Thanks Jack! (Looks like the 3.7C/W ones I have won't fit.)
If I may, one other general question regarding +/- regulators... I understand why in, for example, an audio amplifier the NPN transistors on the + side need to match the PNP transistors on the - side. But in a voltage regulator is there really any need for matched transistors? We just need a good -ve supply and a good +ve supply, no?
If I may, one other general question regarding +/- regulators... I understand why in, for example, an audio amplifier the NPN transistors on the + side need to match the PNP transistors on the - side. But in a voltage regulator is there really any need for matched transistors? We just need a good -ve supply and a good +ve supply, no?
If you have two separate circuit boards which do not have identical parts and identical layout, one of them will perform worse than the other. Perhaps 1% worse, perhaps 80% worse. If you can live with this, no further effort is required.
By coincidence, I also have a 'The Wire' balanced headphone amp and I logged to this thread today to look for the parts to make these regulators.
It it the case then they they won't do for driving the output stage of such and amplifier? I'm using 300 Ohm Sennheiser 800's, so not so demanding I guess, but still, will this SR do it?
I already have a discrete reg supplying it. But when I added 4 x 100u Black Gates across the headphone amps existing 4 x 330u polymer caps, the bass was quite a bit 'bigger'.
So I had to wonder if better a regulator would provide even better sound.
Can I use this SR as it is or should I fit Mark's suggested transistor and will that work straight away as I am in no position to properly test and develop? I need a finished design that I can simply assemble and use (and enjoy!).
Thanks.
It it the case then they they won't do for driving the output stage of such and amplifier? I'm using 300 Ohm Sennheiser 800's, so not so demanding I guess, but still, will this SR do it?
I already have a discrete reg supplying it. But when I added 4 x 100u Black Gates across the headphone amps existing 4 x 330u polymer caps, the bass was quite a bit 'bigger'.
So I had to wonder if better a regulator would provide even better sound.
Can I use this SR as it is or should I fit Mark's suggested transistor and will that work straight away as I am in no position to properly test and develop? I need a finished design that I can simply assemble and use (and enjoy!).
Thanks.
Last edited:
Sensitivity of the HD800's is 102dB/1W -- the idling current of the LME49600 can be as high as 18mA -- shouldn't be a problem for the SR.
Thanks Jack,
Should I use a load resistor, as suggested in recent pages, to get the lowest output Z and keep it there? Will the Z change as the transient load changes? Ie, each louder transient drops the Z, then it raises again? Does that matter at all to the sound quality? Would it exaggerate the transients? Or change the rest of the sound while a transient occurs?
I don't play these headphones at ear damaging level, relatively quietly I expect, but I've no idea how much current they draw during normal usage.
I've forgotten the Headphone amp schematic just now, but there are two 49600 per channel, so that would be about 36mA quiescent. I was thinking to use a separate ±reg for each channel, just in case that was beneficial.
Should I use a load resistor, as suggested in recent pages, to get the lowest output Z and keep it there? Will the Z change as the transient load changes? Ie, each louder transient drops the Z, then it raises again? Does that matter at all to the sound quality? Would it exaggerate the transients? Or change the rest of the sound while a transient occurs?
I don't play these headphones at ear damaging level, relatively quietly I expect, but I've no idea how much current they draw during normal usage.
I've forgotten the Headphone amp schematic just now, but there are two 49600 per channel, so that would be about 36mA quiescent. I was thinking to use a separate ±reg for each channel, just in case that was beneficial.
Last edited:
What would be a reason to use the Superreg over a paralleled LT3042 like this:
The Wire - All Boards and Kits Explained Here! - diyAudio
(the Wire LT3042)
My particular application would be a studio mixed with phono sections and a bunch of op-amps. Still trying to find something that beats the stock SMPS in terms of sound quality.
The Wire - All Boards and Kits Explained Here! - diyAudio
(the Wire LT3042)
My particular application would be a studio mixed with phono sections and a bunch of op-amps. Still trying to find something that beats the stock SMPS in terms of sound quality.
LT3042, right at the bottom of that page.
Yes, can someone here estimate how would that compare to those in the super reg comparison? For sound quality for the headphone output. I noted the comment that lowest possible noise was less preferable to dynamic response.
Jack?
Yes, can someone here estimate how would that compare to those in the super reg comparison? For sound quality for the headphone output. I noted the comment that lowest possible noise was less preferable to dynamic response.
Jack?
You'll have to listen to them yourself and report back. There are new low noise, high psrr regs from Analog Devices, Tex Instruments and LT in the past couple years.
I also would like to know what determines the dynamic response of power supply. Low noise is secondary in some applications.
I have tried replacing a SMPS with linear one and transient attack and dynamics were subjectively reduced.
The dynamic response of the power supply is determined by the gain of the pass transistor and the speed of the error amplifier.
Low noise is nice, but .... well it's like the dog-food problem in linear programming.
So what type of known ps has the best dynamic response? I am willing to give it shot. need +/-18v and 1amp. Recommendations?
Perhaps a more pertinent question to ask here at least is to ask whether anyone has scope shots of this regulator's transient response to a stepped load.
Transient response is intimately related to a regulator's output impedance...
The Linear Audio article on regulators shows the output impedance graphs. I'd recommend reading the whole article, but Jan has put the colour graphs up the downloads section of the website: Downloads | Linear Audio
http://linearaudionet.solide-ict.nl/sites/linearaudio.net/files/V4 JW F8.pdf
The Linear Audio article on regulators shows the output impedance graphs. I'd recommend reading the whole article, but Jan has put the colour graphs up the downloads section of the website: Downloads | Linear Audio
http://linearaudionet.solide-ict.nl/sites/linearaudio.net/files/V4 JW F8.pdf
And, as mentioned up a few bumps, Zout is related to the gain of the pass transistor, etc., etc.
TI once pointed out, in response to a National Semi app-note that reducing the noise of the LM317 by decoupling the adjust pin resulted in deterioration in transient performance.
So, there are many tradeoffs.