Powering Multiple 5V Chips

[Ripthorn]

As I continue working on my project that has an MCU, CMOS switch IC, and multiple PT2399's, I find myself needing some guidance here. The majority of the circuit is 5V, with input and output buffers at 9V. The current draw on the 5V rail is expected to be on the order of 200 mA. That is higher than the 78L05's 100 mA max output. That means I either need to use an LM7805 or multiple 78L05's.

Seeing how I am hoping to keep this SMD, I wonder if it might be easier to use the multiple 78L05's, one at each chip. This has a few advantages:

- 9V power is will be the only primary power trace to run, I don't have to worry about snaking a 5V trace and a 9V throughout the board
- Each 78L05 can be located very close to the power supply pin of each chip, which means the filtering will be more efficient, etc.
- 78L05 in SMD doesn't require any kind of heat sinking.

The drawbacks I see as:

- 78M05 (LM7805 in SMD) is a single chip and therefore less overall cost (not such a concern for me)
- 78L05 would take more board area
- 78L05 would require one extra filtering capacitor per chip, increasing component count

I'm not concerned about the nominal extra cost, it's really a question of simplicity and what will work for optimal power supply operation. Any thoughts would be much appreciated. I'm running an LM7805 on the breadboard and it is working fine, I'm mostly concerned about having multiple voltage traces on the board and not criss-crossing anything. Thanks!

[antonis]

I should use L788M05 https://www.st.com/resource/en/datasheet/l78m.pdf
IMHO, 200mA for SMD (DPAK) shouldn't cause any issue..
(but that's a rough estimation and you need to precisely calculate power dissipation..)

[bean]

A 78L05 will certainly run more than one PT2399 but if there's space I tend to use a separate regulator for each device. It has a small footprint in either package. Plus, you actually don't save an extra decoupling cap: you should be using one right next to the power pin of each 5v device whether or not you are using a single or multiple regulators.

Or, maybe the answer in your case is one 5v regulator for the PT2399s and another one for your MCU/CMOS.

[Ripthorn]

Interesting, perhaps I'll consider a separate regulator per device or something. I've read that a pt2399 can consume up to about 30 mA, is that correct?

[Rob Strand]

It's more normal to have a single 5V rail and to use local bypass caps if you need to keep the supply impedance low.  There's a few exceptions where you might want to isolate a sensitive part of the circuit with a separate regulator.   Another case is where you might want to keep the fault current low, since the 78L05 has a lower short-circuit current.

Something to realize is the 7805 and 78L05 do not have the same specs.  In general the 7805 will have much better performance in nearly all respects:
- more than 10 times lower thermal resistance.   The 78L05 can get hot internally with quite a small load.
- lower output impedance and better load regulation.
  Need to be careful how to compare apples to apples here as the 7805 figures are at 0.5A or 1A (depending on specific model) and the 78L05 is at 100mA so the actual load regulation for the same load can 5 to 10 times
better than the figure in the 7805 datasheet.   
  If you divide-up load then you are helping the 78L05 case because the load is reduced.
- better line regulation and lower ripple
  (also better line regulation due to self heating/drift which is *not* in the datasheet figures)
- Because of the lower output impedance and lower internal impedances the 7805 will show
  a lower dropout voltage.
  Again, for the many 78L05's the difference is reduced but you will probably find one of the 78L05 is the weakest link.


Despite the 150mA rating the maximum current you can pull from a 78L05 often be less than 150mA because the power dissipation can cause the regulate to over-temperature.   For 9V in 5V out and thermal resistance of 160 deg C/W, temperature rise for 35degC ambient  dT = 125 - 35 = 90degC, the thermally limit output current is 140mA.  For 12V in that will drop to 80mA.   Keep in mind the regulator is cooking inside and about to shut down at this point.  Some older datasheet quote higher thermal resistances than 160 deg C/W making the limit lower.

The 7805 can pull about 10 times the current for the same temperature rise.

[vigilante397]

+1 to the 78M05 in the DPAK, they're rated up to 500mA. Keep a decoupling cap on each device for sure, but one 78M05 should cover it.

[PRR]

LM7805 is available in several SMD packages. I can't see the benefit of multiple LM78L05s.

[Ripthorn]

Looks like I'll go for a single 5V regulator. Thanks for all the great info, guys!