0
\$\begingroup\$

I am working on the hardware of a boost converter . I will just pose a question on my boost converter since I assume the same knowledge can be applied to my buck converter. Well this is my first converter design and also my first PCB layout design so any positive criticism will be appreciated. I have attached my layout below for everyone to see. PCB layout 3D view

Well firstly is the layout good enough to reduce EMI? For a maximum 7 A continuous current is 2 mm/78.7 mil enough? For the system testing, let us say I want to test the inductor current or any other components current What measures can I add to my PCB to perform such test. Is maybe adding a header where the current will flow and then connecting the header with an appropriate size wire where you can place your current probe good practice? Or maybe breaking the track and adding a wire there instead? Any suggestions on this will be appreciated. Is a single layer design sufficient? In the the case that I do not have a lot of components.

I have to also use voltage and current sensors for my system, how do I incorporate their design within this PCB if I want to have that on a different PCB?

I have added the schematic of the PCB below.

PCB schematic

\$\endgroup\$
7
  • 3
    \$\begingroup\$ Add the schematic. What controls the switching on Q1? But at a glance there's lots of long thin conductors enclosing huge loops, so it looks like a pretty good transmitter antenna. Did you pay any attention to the recommended layout guidelines wherever you got the schematic? \$\endgroup\$
    – user16324
    Commented Oct 2, 2020 at 12:43
  • 4
    \$\begingroup\$ Look for boost converter controller datasheets and App notes, there will be plenty of recommended layouts. TI has a very nice App note showing EMI differences for the same components with different layouts (of buck though, not boost) which I cant find now :(. Clearly the loops are not as short as possible and the tracks could easily be way wider. \$\endgroup\$
    – Wesley Lee
    Commented Oct 2, 2020 at 13:26
  • 2
    \$\begingroup\$ Are you familiar with current loop area and why you should try to minimize it? Usually that involves putting the switching components very close together. Given the high current, you should probably use copper planes as well. \$\endgroup\$
    – user1850479
    Commented Oct 2, 2020 at 14:06
  • 2
    \$\begingroup\$ What's the value of C1, C2, L1, what's your switching frequency, input voltage range, output voltage and current? For layout, C1, C2, L1, Q1 and D1 should sit as close to each other as possible. I doubt you will manage well without a ground plane, i.e. two layer board. \$\endgroup\$
    – winny
    Commented Oct 2, 2020 at 14:38
  • 2
    \$\begingroup\$ Minimalistic. A schematic should always have values on it. Essentially, everything winny said. Planning on current & voltage monitoring on a separate board does not help EMI (increases loops). Layout could be tighter. No ground plane increases EMI. Inductor sizing is important for 7A continuous, especially with CPU control doing boost and buck. I believe this is a good point to reconsider your approach. \$\endgroup\$
    – StainlessSteelRat
    Commented Oct 2, 2020 at 15:57

1 Answer 1

Reset to default
1
\$\begingroup\$

Let's Discuss Step by Step,

1. Circuit Diagram

  • For any Buck or Boost converter use few Bulk capacitors in parallel to reduce ESR.
  • Use Decoupling MLCC or PF capacitor parallel to Bulk Capacitors.
  • In P3 connector use the remaining pin as GND.

2. PCB Layout

  • Never use the steep bend of any data or power line.
  • Try to minimize High current loops.
  • Use Ground Bottom Plane, and Thick Positive line.
  • Keep Mosfet Drain Line away from Inductor to minimize the noise.
  • Place Power Mosfet and Diode on one of the edges of the PCB so that later you can add heat sink if needed.

Regarding your Current and Voltage sensors, post the issue or discuss here, we will solve the issue.

Best of luck 👍🏻

\$\endgroup\$
3
  • \$\begingroup\$ Thanks for your response. This also clears up a question I had on capacitor selection. Do you have suggestions on this "Is maybe adding a header where the current will flow and then connecting the header with an appropriate size wire where you can place your current probe good practice? Or maybe breaking the track and adding a wire there instead? " \$\endgroup\$
    – Kevin
    Commented Oct 2, 2020 at 15:49
  • \$\begingroup\$ For this, I would suggest, add a shunt resistor of around 10mΩ/1Watt in your current path where you want to measure current. Then use Normal DSO to measure current. Placing a header is not recommended as the Contact resistance of the header will cause an error in measurement. Check the method here "allaboutcircuits.com/technical-articles/…" \$\endgroup\$
    – Deepak
    Commented Oct 3, 2020 at 7:31
  • 1
    \$\begingroup\$ Thanks for your input ,its is very much appreciated \$\endgroup\$
    – Kevin
    Commented Oct 3, 2020 at 15:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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