2
\$\begingroup\$

For a Powerline Communication device that uses the frequency band 42 to 89 kHz. I need to choose a power supply that will not add noise in the powerline and kill my communication band. There are two devices with certain current requirements; the first needs 1 A and the second needs 4 A.

I have looked at the available power supplies and tried adding a mains filter to suppress the EMC noise out of it (see figure below).

enter image description here

This will attenuate the communication signal as well, due to the Cx cap on the line (see figure below). So my best option at this point is to choose a better PSU to solve the problem.

enter image description here

I am looking for two designs or models of a switching mode power supply that will output 1 A or 3 A that has a high enough frequency that it harmonics will not interfere with the communication band between 42 to 89 kHz.

\$\endgroup\$
  • \$\begingroup\$ Are you asking us to design it for you? \$\endgroup\$ – winny Jun 26 '18 at 10:28
  • \$\begingroup\$ No, I am hoping to find a power supply that has been designed for this purpose already. Or if there is a design that someone has seen with a high enough switching frequency that it will not effect the band I am interested in and also give 1A or 4A. \$\endgroup\$ – Gareth T. Jun 26 '18 at 11:30
1
\$\begingroup\$

If you are using a SOC PLC device from STM you could maybe have a look at the ALTAIR04- 900 device as it was a recommended power supply in a datasheet for one of their devices. Page 56 of this document. They also have an example of the power supply input EMI filter- designed for the CENELEC A Band which is the frequency band your are interested in (page 57).

\$\endgroup\$
  • \$\begingroup\$ This looks like a viable option. Wow, thank you. I need to do a little bit more research into it. Do you know of any other options that are similar to the Altair04 that will have the same effect? On and the EMI filter looks interesting. I don't know how the output works but I will ask that in a separate question. Thank you again. \$\endgroup\$ – Gareth T. Jul 3 '18 at 8:27
1
\$\begingroup\$

First, just assume the power line is noisy and design your protocol accordingly. A competent switching power supply by the device shouldn't make this substantially worse because you already assume there are switching power supplies hanging off the power line in various places.

Second, a line filter doesn't have to apply both to the power supply and the signal connection to the power line. This really should have been obvious. Put the line filter between the power line and whatever power supply you use, but tap the line for the purposes of the signal connection on the line side of the filter.

\$\endgroup\$
  • \$\begingroup\$ Oh, I see what you are saying about the line filter, and yes that is the way we have it. I will make it clear in the question. The element that I am concerned about in the line filter is the front cap (Cx in more filter designs) this will directly effect the comms signal in the line. I did change the characteristics of the font cap to not filter in my band, but if there is noise there because of the power supply I have chosen I see that as a bad design. \$\endgroup\$ – Gareth T. Jun 26 '18 at 11:36

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.