I am working on a PCB design with STM32 MCU. On the PCB there will be 2 MAX490 transceivers and one bluetooth module. Communication speed will be 921.6 Kbit/s for all communication elements. Bluetooth module uses 3.3V as MCU and transceivers use 5V supply voltage. Board will be powered by a 12V battery.

So i need 2 buck converters for to step-down 12V to 3.3V and 5V. But i am concerned about their EMI performances. Would these converters make any interference on the comm lines?

  • \$\begingroup\$ Whats stopping you running everything on 3v3? At least you'll remove 1 further switcher. Alternatively you could switch 12v to 5v then linear regulate 5v to 3v3. As for interference that entirely depends on your filtering design and pcb implementation \$\endgroup\$ – theguitarfreq May 24 at 12:36
  • 1
    \$\begingroup\$ Sure they will and so it's the engineers job to design the PCB and schematic to minimize the interference to acceptable levels for decent comms integrity. \$\endgroup\$ – Andy aka May 24 at 12:36
  • \$\begingroup\$ @AndyWest Transceiver input supply voltage range is 4.5V to 5.5V. I don't they would work decently on 3.3V. \$\endgroup\$ – Teo Laferla May 24 at 12:41
  • \$\begingroup\$ @TeoLaferla Yes but could you use an alternative such as: mouser.co.uk/Maxim-Integrated/RS-422-RS-485-Interface-IC/… \$\endgroup\$ – theguitarfreq May 24 at 12:43
  • \$\begingroup\$ @AndyWest The sensors connected to these MAX490s are using full-duplex comm. The ICs from the link you've shared are half-duplex. \$\endgroup\$ – Teo Laferla May 24 at 12:48

It can cause communication issues if one side is earth grounded with an EMI filter to ground and the other has a loop antenna for cabling on a floating supply without earth grounding.

Even with differential drivers, receivers and a balanced cable, CMRR is compromised due to the leakage capacitance in the ungrounded side.

A workaround is to use two floating SMPS and a CM choke that operates >> 200 Ohms in the bandwidth of the Rx and CM noise or earth grounding both sides to re-direct leakage capacitance across the SMPS transformer.

I have proven this to PhD post-grad workers in their EE lab with 10yrs industry experience in SMPS using USB on a 1m cable at 1.5Mb/s USB1. This is similar to your application near 1Mbps.

  • \$\begingroup\$ Would you recommend using a different board for step down circuits? \$\endgroup\$ – Teo Laferla May 24 at 14:25
  • \$\begingroup\$ Different from what? This issue is Common mode noise between two terminals. One is battery powered, what is the other side using? \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 May 24 at 14:49
  • \$\begingroup\$ I mean another board for buck converters. Board and battery will be in a casing so i can choose the buy one commercial buck converter and just place it somewhere in casing and take power supply from this commercial buck converter. It wouldn't be a professional thing to do of course but i think a power switcher would make drastic noise on the MCU board. \$\endgroup\$ – Teo Laferla May 24 at 15:28
  • \$\begingroup\$ You have not defined current or power specs. OKI makes a 3 terminal SMPS 1A at digikey. but I asked what about the other end. Is it known>? or the Logic level interface ground? where does it come from? You need to show a block diagram of all the interfaces with DC , data and earth ground. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 May 24 at 15:44

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.