Positive regulators for negative supplies?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm interested in using a positive regulator to regulate a negative supply in addition to a positive supply. More specifically, I'm looking at using the LT1763 low noise regulator to power some op-amps as a front end to a power amp. I don't see any reason why I could not use the same regulator for a negative supply, as long as my positive and negative supplies start out completely separated. I plan to use two secondaries to generate the unregulated voltages, regulate them with two "positive" regulators, and tie them together at the output of the regulators to obtain a bipolar supply.

Can anyone tell me if this is a bad idea, and why?

Thanks,
John
 
you can get +/- from a single ended output -- I put the schematic up on www.diyaudio.com, but the original is at the office. Basically, I take one output, rectify, filter and regulate with a 7815. I pull the negative rail from the "Ground of the 7815 and the positive rail from the obvious...I create a floating ground by connecting the positive and negative together with a 560R resistor. This might be adequate. The resistor has to be small enough in value (I use 560 ohms) so that the floating ground stays in place. If you further regulate each of these rails with LM317LZ and LM337LZ regulators you don't want the current on the adjust pin pulling or biasing the regulators --

You can ditch all this and use the the LT1026 +/- doubler inverter --+ 5 volts in +/- 10 volts out. See linear tech's website. If you are only driving a few opamps it might suffice (you can pull 30ma from the chip.) I posted a view of the output here last week.<p> You will need a small choke on the output (I use 62uH) -- Linear Tech has a circuit on their site which uses the LT1026 with a single LDO regulator to derive bipolar supplies. If you need a couple LT1026's to play around with I can let them go for a buck (half of the Digikey price)., contact jack@tech-diy.com
 
Warning!

Hi,

I once used a +ve regulator to produce -5.2 volts for a very expensive chip (a days wages at the time).
During testing, I found I needed to attach my 'scope ground clip to something.
I chose "common ground". The moment it touched, there was a small flash, and the chip stopped working.
Of course, the regulator had been bypassed by the signal cables and the probe earth.
From then on, I always use +ve and -ve regulators in their proper positions.

Cheers,
 
JohnG said:


Since there are no major objections, I think I will go with my plan.

The LT1763's look like very nice regulators.




John,


What makes you think this is a good regulator? This doesn't mean I don't think it is, just that is not being used for audio applications as far as I know.

Being low-dropout doesn't mean too much, unless you're going to use it with batteries in portable situations.

Things like output impedance might be more important, and that data is not there.

In my opinion you should compare any new regulator with the ones being used until now and reported as good sounding, like LT317 or LT1085. And better still try them and see how they compare.


Carlos
 
Carlmart

it's probably the lowest, or one of the lowest noise linear regulator available -- 20uV, and the dropout is very low. The "noise" per se, of an LM317 is anywhere from 1 to 5mV with the same amount of capacitance (well, capacitance doesn't really reduce noise, it reduces ripple.)

This noise level gets the unit down to the noise level of precision reference diodes.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.