LM317 and LM350 FOUR terminal regulators!!!
May 10, 2006 2:10:50 GMT
Post by gns on May 10, 2006 2:10:50 GMT
The LM317 and LM350 (1.5 and 3 Amps respectively) positive voltage regulators (and their negative counterparts) give better performance than a fixed regulator.... Really?
The great thing about using adjustable voltage regulators is you have more control over the results and therefore can obtain a better performance - the part itself cannot do it for you.
You can reduce the ripple (hum-noise) through bypassing the adjust pin to ground using a 10 micro-farad tantalum capacitor and it works a treat, but using the plastic package TO-220 types you can replace ripple by low frequency instability which will really upset the sound of a high gain circuit such as a phono preamplifier. It can really make it boom, so like most things are double edged swords, the solution given here could actually be used by the booom-booom brigade in reverse....
The solution is in the title of this thread. Four terminals? These are three terminal devices surely? Not if you count the tab!
Usually you take the output-to-adjust pin resistor from the output pin - it seems to make sense, except that the small length of pin emerging from the device does actually have resistance. It may be inconceivably low, but resistance it is. And it is also in series with the load.
The result is that the adjust pin voltage "micro-gyrates" in sympathy with the signal at the amp stage, and in RIAA stages where bass frequency gains are up to 10 times higher than at mid frequencies, and 100 times higher than high frequencies, the bass seems to get unnaturally boomy.
That's because of a thing called positive feedback through the power supply rails - in other-words you just created a low frequency oscillator, which although it won't normally free-run, can sound really bad.
The solution is to star out from the output of the chip itself so there can be no pin Resistance whatsoever, but that's impossible!!!
Or is it?
The tab is connected to the chip's output! There's your solution!
By taking the adjust pin resistor from the tab you get rid totally of any series resistance, and you get in return, stability.
For headphone amps and other audio circuits, you're not going to hear it as an immediately obvious improvement, but it is there, and all these little things add-up to a better sound.
G
PS. I will be doing some "kit" projects at our website www.audiocontrol.co.uk sometime in the not too distant future... ;D
PPS. Users of the T03 or TO39 versions of the above can do the same by taking the output from one mounting screw, and the adjust pin resistor from the other
The great thing about using adjustable voltage regulators is you have more control over the results and therefore can obtain a better performance - the part itself cannot do it for you.
You can reduce the ripple (hum-noise) through bypassing the adjust pin to ground using a 10 micro-farad tantalum capacitor and it works a treat, but using the plastic package TO-220 types you can replace ripple by low frequency instability which will really upset the sound of a high gain circuit such as a phono preamplifier. It can really make it boom, so like most things are double edged swords, the solution given here could actually be used by the booom-booom brigade in reverse....
The solution is in the title of this thread. Four terminals? These are three terminal devices surely? Not if you count the tab!
Usually you take the output-to-adjust pin resistor from the output pin - it seems to make sense, except that the small length of pin emerging from the device does actually have resistance. It may be inconceivably low, but resistance it is. And it is also in series with the load.
The result is that the adjust pin voltage "micro-gyrates" in sympathy with the signal at the amp stage, and in RIAA stages where bass frequency gains are up to 10 times higher than at mid frequencies, and 100 times higher than high frequencies, the bass seems to get unnaturally boomy.
That's because of a thing called positive feedback through the power supply rails - in other-words you just created a low frequency oscillator, which although it won't normally free-run, can sound really bad.
The solution is to star out from the output of the chip itself so there can be no pin Resistance whatsoever, but that's impossible!!!
Or is it?
The tab is connected to the chip's output! There's your solution!
By taking the adjust pin resistor from the tab you get rid totally of any series resistance, and you get in return, stability.
For headphone amps and other audio circuits, you're not going to hear it as an immediately obvious improvement, but it is there, and all these little things add-up to a better sound.
G
PS. I will be doing some "kit" projects at our website www.audiocontrol.co.uk sometime in the not too distant future... ;D
PPS. Users of the T03 or TO39 versions of the above can do the same by taking the output from one mounting screw, and the adjust pin resistor from the other