I got an lm317 337 regs feeding a +-24v buffer. And from the same reg id like to power an opamp at +-15v. So i need a circuit to drop 7 volts. Initially i had passive shunt reg in mind but could there be more elegant way to achieve my goal? The input voltage is already fixed so i dont need another voltage reg per say. Maybe a cap multiplier thats configured for 7v difference will be better for the opamp rail's pssr? But the shunt is a constant current draw which i understand would keep the 317 337 happy.
What are your thoughts?
What are your thoughts?
+1 just add suitable additional nested 3-pin regs.
The 'noise' such 3-pin regs add is fundamentally 'white' - but since the raw supply is itself already regulated, the compound regulation against the insidious stuff of the raw supply, is immense.
- And since the 'noise' output of such a nested regulator with a usefully-low output impedance is far from being correlated with the signal the opamp is handling it does not, in the least, affect the opamp performance.
😀
The 'noise' such 3-pin regs add is fundamentally 'white' - but since the raw supply is itself already regulated, the compound regulation against the insidious stuff of the raw supply, is immense.
- And since the 'noise' output of such a nested regulator with a usefully-low output impedance is far from being correlated with the signal the opamp is handling it does not, in the least, affect the opamp performance.
😀
So the line rejection of 7812 is preferable to resistor> zener simple shunt reg? My thinking was the R of shunt reg would be upwards to 500ohm and thats a really good way to combat the high freq hash that lm317 doesnt handle
'good way to combat the high freq hash that lm317 doesn't handle'
That is so,so easy: rather than just rely on a 'raw' DC supply - add a little R-C decoupling in front of 3-pin regs.
Regardless of the raw DC supply type - just add 1-10ohms in series (depending on how much current you need to supply and therefore how much IxR voltage drop you can afford from the raw supply), then, c. 100-1000uF as local bypass at the regulator Vin to 0v rail.
This utterly kills any influence from the raw DC supply. 10ohm+100uF is a low-pass filter at c. 160Hz ; from whatever might be on the incoming supply - it adds 20dB rejection at 1600hz, 40db at 16Khz ... so easy; so cheap; so reliable.
There are many more advanced ways of trying to do 'better' but a dab of R+C, in terms of utter simplicity/ ease of implementation/ reliability & cost ... unbeatable.
That is so,so easy: rather than just rely on a 'raw' DC supply - add a little R-C decoupling in front of 3-pin regs.
Regardless of the raw DC supply type - just add 1-10ohms in series (depending on how much current you need to supply and therefore how much IxR voltage drop you can afford from the raw supply), then, c. 100-1000uF as local bypass at the regulator Vin to 0v rail.
This utterly kills any influence from the raw DC supply. 10ohm+100uF is a low-pass filter at c. 160Hz ; from whatever might be on the incoming supply - it adds 20dB rejection at 1600hz, 40db at 16Khz ... so easy; so cheap; so reliable.
There are many more advanced ways of trying to do 'better' but a dab of R+C, in terms of utter simplicity/ ease of implementation/ reliability & cost ... unbeatable.
Far preferable.
You have this back to front. A Zener shunt generates RF hash; an LM317 regulates it away.