Body
Hi Ron,
I see no decoupling capacitors across the voltage supply terminals of the MOSFET drivers. This is asking for trouble. The current when charging the input capacitance of the rightmost MOSFET has to go from the voltage supply to terminal J1 and through the green path to reach the gate of T6 and return through the yellow path to J1 and back to the voltage supply. When discharging the gate of T6 the blue path through leg 4 of U3 and back to terminal J1 and the yellow path back to T6 is taken.
The big loops formed stores magnetic (inductive) energy that kicks back with voltage spikes when the current is rapid changed. Current can't be discontinuous in inductive circuits which makes them induce voltage to compensate. This may cause instability or even destruction of weak components. Consider it like instantly changing the rotational speed of a flywheel. It will break the axle or something else by the kickback. The yellow return path is shared with the wide ground path. This path may have voltage drops when switching big or fast current. The gate of T6 is in series with this path and directly sees any voltage induced along this path. The normal practice is dividing this track into two separate tracks. A wide one for power and a narrow one for control signals. The yellow path should follow the green path the short distance between T6 and go to a decoupling capacitor at leg 4 of U3. This decoupling capacitor is like a small local power supply or battery close to where the energy is needed. Read Ivor Catt about transmission line theory to know why. Catt had his troubles back in the 1950's when working with the very fast emitter coupled logic.
Place decoupling capacitors as close as possible to the voltage supply terminals of the MOSFET drivers (leg 1 and 4). Decoupling capacitors may be needed across the positive and negative voltage supplies of the MOSFETs. I.e. across the drain of the top MOSFET to the source of the bottom MOSFET.
Minimize the big loops formed. Don't share signal paths with power paths to avoid instability.
Regards
Ole
---In [email protected], <iron1of1@...> wrote :
Hi Warren, all,
Here is the start of my motor controller, making the PCB with Target 3001 (free version) and a GRBL CNC mill. Do the design in Target 3001 and it spits out the G code to run the CNC 2418 mill. The sense resistors will be on another board.
https://www.youtube.com/watch?v=QiJPdxUojRo&feature=youtu.be https://www.youtube.com/watch?v=QiJPdxUojRo&feature=youtu.be
How is your's coming along?
Ron