I've been studying the design of the super regulator(s) with the intention of building a SMD version for myself. I think I understand most of the circuit functionality (with the possible exception of the startup constant current source - analog electronics is not my forte) but have hit a stumbling block when it comes to an alternative for the LM329.
I've come across the ADR5045, which lists output voltage noise as 6.6 uVrms for 0.1 Hz to 10 Hz and 280 uVrms for 10 Hz to 10 kHz. Now the LM329 datasheet lists it's RMS noise as typically 7 uV for 10 Hz to 10 kHz. Would this mean the ADR5045 is a poor substitue as its 10 Hz to 10 kHz figure is higher, or should we instead look at the 0.1 Hz to 10 Hz figure as the output of the regulator is filtered by a ~2.7 Hz low pass filter?
I've also looked at a regulators such as the ADR435, which produces a 5V reference output from an input voltage of 7V to 18 V. (Datasheet ) Now the catch for this part is the minimum Vin is 2V above Vout, meaning it can't be a drop in replacement for the LM329.
But if my output rails are +/- 10V, I wonder if I could do this (omitting the capacitors and most of the supply side of the circuit):
By tying the output of the second ADR435 to ground, it sees 15V from Vin to GND, while providing me with a -5V reference. Only thing is I doubt the GND pin can be hooked up in that way.
Alternatively, based on Fig. 36 in the ADR435 datasheet, I could try this:
Ignore the AD8610 tag, it could be any op-amp, I just used the first part on the list. In this design, I can imagine that the voltage noise being fed into the 825 is then a function of the op-amp, and based on that I don't know if the dual ADR435 design would be a better solution. Or have I got it wrong and neither one is feasible?
I've come across the ADR5045, which lists output voltage noise as 6.6 uVrms for 0.1 Hz to 10 Hz and 280 uVrms for 10 Hz to 10 kHz. Now the LM329 datasheet lists it's RMS noise as typically 7 uV for 10 Hz to 10 kHz. Would this mean the ADR5045 is a poor substitue as its 10 Hz to 10 kHz figure is higher, or should we instead look at the 0.1 Hz to 10 Hz figure as the output of the regulator is filtered by a ~2.7 Hz low pass filter?
I've also looked at a regulators such as the ADR435, which produces a 5V reference output from an input voltage of 7V to 18 V. (Datasheet ) Now the catch for this part is the minimum Vin is 2V above Vout, meaning it can't be a drop in replacement for the LM329.
But if my output rails are +/- 10V, I wonder if I could do this (omitting the capacitors and most of the supply side of the circuit):
An externally hosted image should be here but it was not working when we last tested it.
By tying the output of the second ADR435 to ground, it sees 15V from Vin to GND, while providing me with a -5V reference. Only thing is I doubt the GND pin can be hooked up in that way.
Alternatively, based on Fig. 36 in the ADR435 datasheet, I could try this:
An externally hosted image should be here but it was not working when we last tested it.
Ignore the AD8610 tag, it could be any op-amp, I just used the first part on the list. In this design, I can imagine that the voltage noise being fed into the 825 is then a function of the op-amp, and based on that I don't know if the dual ADR435 design would be a better solution. Or have I got it wrong and neither one is feasible?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.