Forum Rule: Always post complete source code & details to reproduce any issue!
Results 1 to 7 of 7

Thread: Opto Isolated Mosfet Driver Board

  1. #1
    Senior Member
    Join Date
    Mar 2013
    Posts
    651

    Opto Isolated Mosfet Driver Board

    Im posting this hear because I believe it may have a use to some who use Teensy's.

    I have been researching 3d Printers for the past few weeks and I have seen several people use SSR's or standard relays to run the High Wattage Heat Beds and figured there should be a better way to drive a high power load with logic signals. Many 3d Printers are setup to drive 12Aish loads but anything more then that and bad things can/do happen. I have seen a few Mosfet Driver boards but none of them had Isolation on them. Which is something I like to use when driving large loads with micro's.

    I figured since the printer kit I am looking at comes with a Relay and I plan to replace it ASAP I figured I would just create the board I wanted and give it enough functions to be useful for more then just 1 thing.

    It uses cheap off the shelf Dip4 Opto Couples to drive low RDSon n-channel MOSFETs. This limits them to around 1Khz switching frequency.
    Its laid out to support either SMD or Thru-hole components(4pin SMD opto fits the Dip pads). The only exception is the 30A rated screw terminals for the load. The Pos supply and digital lines can be direct wired.

    My goal is to be able to support up to 30A and up to possible over 100VDC(reason for Z1).
    The basic Kit I might supply would be for a 12-24V Load and a 3.3-12V Digital input. The output limits would be specific to the Fet used and the amount of cooling.
    I plan to build 3 prototypes with different setups, D2pak vs TO220 vs TO220+Heatsink to see what works and what does not work or at least where the limits need to be.
    If those pan out as well as the simulations/math then I would make a larger kt batch based on interest/feedback.





    Also.....
    I have looked at making an even more powerful version for those insane 120VAC 600-1000W beds but the capacitors needed would almost quadruple the size of the board so I wont do it unless someone asks for it. The upside to it would be reduced current, the downside is 80-170VDC hurts
    Last edited by Donziboy2; 02-01-2016 at 06:09 PM.

  2. #2
    Senior Member
    Join Date
    Mar 2013
    Posts
    651
    Just noticed I use the Dpak instead of D2pak.... Doh.

  3. #3
    Senior Member
    Join Date
    Mar 2013
    Posts
    651
    Fixed most of my early mistakes, only thing new I learned is that thermal isolation is a godsend and those Stock Eagle thru hole parts have very small pad sizes. Thats something I will fix for my other projects.


  4. #4
    Senior Member Constantin's Avatar
    Join Date
    Nov 2012
    Location
    In the yard with a 17' Dia. Ferris Wheel
    Posts
    1,408
    Neat!

    However, do consider making the board longer so that the isolator can actually isolate. The isolation rating only holds true if the area under the isolator is free of traces. You need to keep all low signals to one side, all high side signals on the other.

    Also, I tend to disable the polygon thermals for high load applications like this. Makes soldering harder (flux well, need a good hot iron) but also helps carry all that current without toasting the copper traces.

    Looks like you are using two sides of the board for the the high-amp circuits, which is good. However, I'd add more 23 mil vias that you can then drench with solder to tie the two sides together.

  5. #5
    Senior Member PaulStoffregen's Avatar
    Join Date
    Nov 2012
    Posts
    23,784
    I like those screw terminal parts!

  6. #6
    Senior Member
    Join Date
    Mar 2013
    Posts
    651
    Quote Originally Posted by Constantin View Post
    Neat!

    However, do consider making the board longer so that the isolator can actually isolate. The isolation rating only holds true if the area under the isolator is free of traces. You need to keep all low signals to one side, all high side signals on the other.

    Also, I tend to disable the polygon thermals for high load applications like this. Makes soldering harder (flux well, need a good hot iron) but also helps carry all that current without toasting the copper traces.

    Looks like you are using two sides of the board for the the high-amp circuits, which is good. However, I'd add more 23 mil vias that you can then drench with solder to tie the two sides together.
    Well, the isolation is more for using 2 different grounds then anything else, I dont see any need to move it out to much further, im not shooting for 5kV. I will probably make the board bigger to increase cooling for the D2Pack(and more via's, they are a SOB to solder lol). I did some testing today and miswired the massive brake resistors to 0.5 instead of 1ohm. Got some good results doing it lol, at 12V, 500Hz and 50% Duty Cycle the D2Pack just could not keep up (24A peak and 17A RMS). The TO220 did ok(60C constant on package at 500Hz) with a large heatsink attached, I think there is room for improvement, the switching looses seem to be pretty bad at 1KHz so I need to have a closer look(from the time I applied power it only took about 1 minute to go over 100C on the Packages).

    Quote Originally Posted by PaulStoffregen View Post
    I like those screw terminal parts!
    30A rated
    https://www.digikey.com/product-deta...8197-ND/316833

  7. #7
    Senior Member
    Join Date
    Mar 2013
    Posts
    651
    Did some testing at work, the D2Pak case gets to around 85C running at 0.7ohm load at 95% Duty Cycle (500Hz, 12V, 17Apk, 16.5A RMS).
    I took the TO220 board and added a large heatsink , it had no problems at the same levels as the D2Pak and seemed fairly stable at around 70C case temp.

    I need to increase the size of the PCB to add space for a larger HS/larger copper pad for the D2PAK.
    I found reducing the switching frequency also lowered temperatures fairly nicely, about a 10-12C drop going from 500 to 250Hz.

    edit..
    Got another 10C drop going from 250 to 100Hz
    At lower frequency's the two packages have close to the same case temperature, both are sitting happy around 50-55C.
    Last edited by Donziboy2; 02-24-2016 at 06:04 PM.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •