You are here

VESC6 PCB questions

7 posts / 0 new
Last post
ESCam
Offline
Last seen: 2 years 1 month ago
Joined: 2018-10-22 23:31
Posts: 24
VESC6 PCB questions

Hello,

I'm not sure if these belong in the original or third party section, so I posted in general. 

I'm hoping to build a basic lightweight FOC ESC based on the VESC6 design for my RC aircraft and also hoping to avoid common issues/mistakes others have seen in the past when developing ESCs. Please let me know if you have any info or advice related to the following questions or anything else related. I should need about 70Adc in max with a 7S battery (~28V). 

  1. What are the copper pour thicknesses? 1oz or more? Can heavier pours cause issues with fine pitch components?
  2. How are the bulk capacitor values and specs selected? What unit limits affect the cap selection? Has anyone tested anything in a smaller package?
  3. Has anyone tested smaller shunt resistors like these ones? They look like similar specs. CSS2H-2512R-L500FE​.
  4. Are there any other questions or concerns I should be asking about, but don't know to ask yet? 

Thanks for any and all help,

Cameron

TechAUmNu
Offline
Last seen: 2 weeks 5 days ago
VESC Free
Joined: 2017-09-22 01:27
Posts: 575

1. So long as you use very short and wide connections and keep everything very close together you can probably get away with a 4 layer board at 1oz.

2. Biggest thing for bulk caps is inductance on DC link. You need to keep the electrolytic caps as close as possible. Also try to get ones with low ESR to stop them heating up too much.

For ceramic capacitors, they need to be directly across the DC input to the fets, if you look at the VESC6 the caps are on the other side of the board directly ontop of the fets with vias going through to the pads on them. 

3. I use that size of shunts on my boards. Larger shunts will handle more heat better and spreads the current across the pcb a bit more.

galp
Offline
Last seen: 1 year 6 days ago
VESC Free
Joined: 2018-10-25 19:08
Posts: 49

For 2 oz copper you can go down to 0.2 mm min clearance / pitch. 0.15mm is possible, but expensive.

Lower value shunts provide lower voltage signal and lower signal to noise ratio when amplified. You'll have to find a balance between efficiency and noise.

Start your placement with power stage. Watch out for crosstalk from phases and mosfet gate traces. 28V swing in few hundred nanoseconds will couple to anything in close proximity. Use via shielding for signals if necessary. Also pay attention to driver - mosfet gate loop and keep it as short as possible to prevent ringing

You can use Saturn PCB toolkit to get via and conductor properties such as max ampacity, inductance, resistance, etc. http://www.saturnpcb.com/pcb_toolkit/

F4metz
Offline
Last seen: 2 weeks 6 days ago
VESC Free
Joined: 2018-07-22 15:47
Posts: 5

Hi,

i myself had a very similar project lately. I agree with the tips of the others above.

If you have not allready, i would recomend you to look at the VESC MINI4 project (https://vesc-project.com/node/246). that might be, what you need as a reference. He actually uses exactly these shunt resistors you were asking about.

regards

ESCam
Offline
Last seen: 2 years 1 month ago
Joined: 2018-10-22 23:31
Posts: 24

Thanks everyone for the great suggestions! They have been very helpful. 

ESCam
Offline
Last seen: 2 years 1 month ago
Joined: 2018-10-22 23:31
Posts: 24

How are capacitors sized for current/power dissipation? The only related value in the​ EKZN630ELL681ML20S datasheet I see is 2.38A ripple current and 63VDC, but the values in the simulation from the video are 55ARMS, much higher than the rated ripple. Is it just a power/heat calculation (I^2*R)?

What power values/limits have people used in the past successfully or unsuccessfully? Does heat sinking the capacitors help much (does the internal heat propagate to the case well)?

I am only running motors around 28V and 60A max DC in, and want to reduce the capacitor size and weight, so I'm thinking I could reduce the bulk capacitor values, but do not want to reduce it too much.

This video was very helpful in explaining the circuit and simulation variables. https://www.youtube.com/watch?v=x6lPdI9OVQg

TechAUmNu
Offline
Last seen: 2 weeks 5 days ago
VESC Free
Joined: 2017-09-22 01:27
Posts: 575

The main parameter for power dissipation is ESR. You can simulate roughly what current will go through the caps but the power measurements are a bit rubbish. They don't heat up nearly as much as you would expect. Heck they don't even get warm on my new board. You can certainly reduce the size of them a little if you want. Just don't reduce them so much that the voltage spikes get too big. Also if you want to make them smaller, don't have really long battery wires!

I haven't tried heatsinking the caps, but in VESC6 and 75/300, they are right next to aluminium case so that probably helps a bit. You can attach them to the heatsink with thermal pad if they are getting too hot.

If you are running lower voltage you can reduce the voltage rating of the caps, you will get the same capacitance in a smaller case size.