Negative voltage regulators...compromised?

[hollowman]

A DIYer recently noted to me that for symmetrical power supplies (e.g., +/- 15VDC), it's best to use two positive regulators, connected in series (like batteries), as opposed to a positive (LM317) and a negative (LM337).

Reason: specs of negative regs, like 337, are inferior to positive regs like 317.

Yet, I continue to see several PS designs for various projects like chip amps and IVs use negative regs. So I wonder how important the spec. diffs. between pos. and neg. regs are for real-world practices?

 

[AndrewT]

Hi,
irrespective of the quality issue between twin +ve reg and +ve &-ve regs, it is cheaper to build +ve & -ve due to saving the extra winding connection and the extra bridge rectifier that are required for twin +ve.

 

[hollowman]

Measured diffs between 337 and 317

Thx for your reply ...

irrespective of the quality issue between twin +ve reg and +ve &-ve regs, it is cheaper to build +ve & -ve due to saving the extra winding connection and the extra bridge rectifier that are required for twin +ve.

Yup -- I noticed the extra parts/cost issue.
Given the diff specs between neg and pos regs, how much measured diff. (e.g. using o'scope or analyzer) is there between the two rails? That is, diffs that are important for HQ audio apps?

 

[fernando_g]

I've always used complementary LM317 and LM337, and prior to that LM78xx and LM79xx regulators for my projects which did require dual polarities. Never experienced any problems that I could relate to the supplies.