Re: [N8VEM-S100:1137] S-100 SMB and S-100 bus extender PCBs status

[David Riley <frave...@gmail.com>]

That's not how linear voltage regulators work.  The voltage drop across the regulator is equal to the difference between the input and output; thus for an 8V input and a 5V output, it will drop 3V (and at 1.5A, it will burn 4.5W, so you better have a decent heat sink).  You'll have the same heat dissipation with a 7805 or an LDO regulator (with a switching regulator, it's a completely different story, but we're not talking about those here).

What a low-dropout regulator gives you is a better *minimum* voltage drop.  The minimum drop is there because there needs to be some headroom to bias the transistors into conduction (think of it as a few Vbe drops).  So, for example, you can run an LDO with a 1V minimum dropout from 6V if you're outputting 5V, whereas a 7805 needs about 8V to operate properly (otherwise, you'll wind up with somewhere around (Vin - 3V) depending on temperature).

Some LDO regulators have a separate, very low current higher-voltage bias supply to allow the supply voltage to be nearly equal to the output voltage; LDO equivalents to the 7805 don't, since they're trying to mimic the 7805 pinout.  I've used ones that use a 3.3v bias supply to achieve an ~0.1v dropout, which of course you can really only do with a FET output stage.

Another ancillary benefit for a low-dropout regulator is that it is generally a much newer part with lower noise and power supply rejection than a regular 7805.  That won't help with the power consumption or heat dissipation, but if you have a noisy power supply, it can help clean it up a lot.

There's certainly no reason not to use one vs. a 7805 unless it's significantly more expensive or less robust.

All this is, of course, why modern electronics use switching regulators (except where linear regulators excel, e.g. noise-sensitive or cost-sensitive circuits where you can afford to throw away efficiency).  You can achieve a greater than 90% efficiency on a circuit like this with a switching regulator, whereas the linear solution operates at a fixed 62.5% efficiency.


-Dave


On Oct 10, 2012, at 2:27 PM, Tom Lafleur wrote:

> the Vdrop (~500mv) on the part is 1/2 of the 7805 (1v) , so less power
> dissipate in the part...
> 
> 
> 
> 
> On Wed, Oct 10, 2012 at 11:10 AM, monahanz <mon...@vitasoft.org> wrote:
>> Interesting Tom, have not used them. Are you sure the L4940-5 will generate
>> less heat. The S100 8V line is kind of fixed at that. I know the L4940-5 can
>> work with voltages just above 5V, but I would think it would have to burn
>> off the excess 8 Volts as heat anyway.
>> 
>> BTW for people building the SMB, be sure to get the 1.5A voltage regulators
>> ( Jameco # 1130069 etc) not the 1A ones.