Re: [EE] looking for negative 5V regulator or circuit

I get it - you need a rail 5V below your incoming +12V supply. And presumably the 
+12V supply may not be (well) regulated, hence the need for the "7V" rail to be 
precisely 5V below the whatever the main supply is.

It could be that all of your logic chips (and micro?) will run off this 7V regulated rail 
(Vcc =+12V, GND=+ 7V), you don't say.

On the other hand you may have a conventional +5V regulator for the low side 5V 
logic circuitry (micro), requiring a logic level shifter to your high side 5V logic.

If the latter is the case and you have +5V available, then an interesting idea would 
be a switched capacitor voltage converter (run by the micro if it suits)... charge cap 
to +5V wrt ground, then isolate both sides of cap, switch high side of cap to +12V 
and neg side provides 7V, repeat. More effort than a negative regulator (unless you 
can't find one)... fairly low losses, fairly rough regulation, way more fun :-)

On 11 Jun 2025 at 10:36, Dwayne Reid wrote:

> I see that I was not sufficiently clear about what I am doing.  My
> apologies.
> 
> I am doing hi-side switching from the incoming +12 Vdc supply.  It looks
> like the easiest and least expensive way to do this is to use 74HC595 shift
> registers driving MMBT4403 transistors.  THe emitters of the transistors
> are tied to the incoming +12V rail.  So are the VDD pins of the shift
> register and the PIC that is driving the shift register.
> 
> So - I need to create a -5V rail with respect to that +12V rail.  In other
> words, about 7 Vdc.  My initial thought was to use a 7905B regulator with
> its input tied to incoming ground and the bias pin tied to the +12V rail.
> That gives me -5 Vdc with respect to the incoming +12V rail.  The problem I
> have is that the 7905 data sheet says that worst case bias current is 8 mA
> which is really high for a vehicle that might be sitting idle (turned OFF)
> for weeks on end.
> 
> So - I'm looking for a -5 Vdc regulator that has significantly-lower
> quiescent current than the 7905.  My fall-back position is something like a
> MPSA63 voltage-follower being fed with a 6.2V Zener or, as was suggested
> here, one of the low-current TL431 variants.  I would use something a bit
> beefier than the MPSa63 but you get the idea.  I'm wanting to get the
> quiescent current down to a couple of hundred uA.
> 
> I estimate the total current on that -5V supply is less than 50 mA,
> depending on how many shift register / transistor stages I need.  Might be
> as many as 16 outputs, each transistor getting a bit more than 2 mA drive.
> 
> As always, suggestions and comments gratefully accepted!
> 
> dwayne
> 
> On Tue, Jun 10, 2025 at 12:43PM Dwayne Reid <[email protected]> wrote:
> 
> > Good day to all.
> >
> > I'm working on a project where I need a negative 5 Vdc rail relative to
> > the incoming +12 Vdc rail.  My first thought was a bog-standard 7905B
> > regulator (the "B" suffix denotes -40C operation) but this is a
> > battery-powered system and the worst case bias current of 8mA is way too
> > high.
> >
> > I spent some time looking at Digikey offerings but they all seem stupid
> > expensive.  Can anyone suggest a part that is reasonably inexpensive and
> > works down to -40C?  Current is fairly low - should be less than 20 mA
> > worst case.  I sort of prefer a package larger than SOT-23 just because of
> > the fairly minimal heat dissipated.
> >
> > My alternative is a 6.2V Zener diode feeding a PNP Darlington pass
> > transistor.
> >
> > Suggestions and opinions greatly appreciated!
> >
> > dwayne
> >
> -- 
> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
> View/change your membership options at
> https://mailman.mit.edu/mailman/listinfo/piclist