0
\$\begingroup\$

I'm trying to figure out if my power distribution setup is hurting my circuitry.

Below is a picture of the setup for one device:

series regulator

The power supply is a wall adapter rated for 6VDC/900mA output.

The middle item is a special power hub I created from two pins of each USB plug and it is powered from the wall adapter through an LM2940 regulator with the 10uF cap on the input and the 22uF cap on the output.

I then connect various USB to C+ plugs to power other circuits off the same hub (hence the right-most box in the diagram). Power to those circuits is passed through its own LM2940 regulator as well.

Some circuits contain a max232 chip for interfacing the circuit with a PC but when I test the circuits in my setup, my baud rate must be very limited, and I'm thinking with two LM2940 regulators in series, I'm not getting enough voltage or current to the max232 and I feel my only way(s) out could be to ditch the hub and power each circuit from its own wall wart or use a wall wart that can deliver higher current.

any ideas?

\$\endgroup\$
  • \$\begingroup\$ Max232 does not need much current... It is unclear what your doing though. USB devices typically require more power then should be used with a linear regulator? \$\endgroup\$ – MadHatter Jul 3 '19 at 23:41
  • \$\begingroup\$ Can you measure all outputs and MAX232 bipolar voltages, and share 6V inputs , don’t cascade then use a <1V CMOS LDO for 5V out . (Lots of part’s do this) Then you are good to go. beware of LDO drop series losses and Rja temp rise. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Jul 4 '19 at 2:28
  • \$\begingroup\$ I'm going with Elliot's answer. I initially suspected that I wasn't providing my circuits with enough power, and his answer confirmed it. \$\endgroup\$ – Mike Jul 4 '19 at 19:08
1
\$\begingroup\$

No, this won't work. The LM2940 requires a minimum input voltage of 6.25V for a 5V output. So, your wall wart is not providing a high enough voltage. Furthermore, you can't connect the output of one LM2940-5 to the input of another LM2940-5.

You need an ac/dc power supply that provides at least 6.25V. The inputs of both of the LM2940s should be connected to this power supply.

\$\endgroup\$
  • \$\begingroup\$ one would think this is is obvious but Op has no mention of bipolar charge pump errors for RS232 and low ESR ratings for cap surge currents. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Jul 4 '19 at 0:56
  • \$\begingroup\$ I'm not too fluent with charge-pumps but I'm replacing the 4.7uF electrolytics with 1.0uF MLCC caps for lower ESR. I read somewhere titaniums are easier to destroy so I'll try MLCC first. Nevertheless, I think the real answer (regardless of capacitor values) is to go with Elliot and not chain the regulators and if I'm still unlucky, then I'll just have to buy the next higher voltage adapter available from the local shop and pray I don't fry the regulator too much because a higher input voltage heats the regulator (especially 9 or 12VDC) \$\endgroup\$ – Mike Jul 4 '19 at 19:07
  • \$\begingroup\$ Then again, I was wondering if my circuits weren't drawing enough current to begin with because someone here said the MAX232 doesn't use much current and if I'm not mistaken, if not much power is taken from a wall adapter, then the voltage reported is higher. I measured the same adapter (one labeled 6V 900mA) once directly with a voltmeter and the voltage reported was almost 19VDC which would allow the LM2940 to deliver 5V output. In any event, I'll stop the chaining for sure and if my circuit draws too much current then its time for a power adapter replacement. My adapter is unregulated. \$\endgroup\$ – Mike Jul 4 '19 at 19:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.