Carrier resudials = speaker cable will become an antenna and spew emi/rfi all over the place.
Which of the board designs sucks the least ?
The order to boardhouse have to be sent before the end of the day.
Btw why do you think all my board layouts sucks ?
Which of the board designs sucks the least ?
The order to boardhouse have to be sent before the end of the day.
Btw why do you think all my board layouts sucks ?
Eva, eat this: i replaced the 22ohm gate resistors with 100ohm resistors and now the fets are COLD!!!! No heating!
There ya go, my circuit is working properly, on a breadboard
Btw i´ve ordered the real boards now.
There ya go, my circuit is working properly, on a breadboard
Btw i´ve ordered the real boards now.
- Power supply loops must be shorter.
- The ground plane has too many cuts in the power loop zone, where it shouldn't have any.
- Gate drive loops may be made longer without much problem.
- Wider PCB tracks are less inductive.
Look at the PCB and try to visualize the paths that cross-conduction (reverse recovery) spikes and inductor current follow when switching is taking place.
- The ground plane has too many cuts in the power loop zone, where it shouldn't have any.
- Gate drive loops may be made longer without much problem.
- Wider PCB tracks are less inductive.
Look at the PCB and try to visualize the paths that cross-conduction (reverse recovery) spikes and inductor current follow when switching is taking place.
Well with the recommended clearance of 20mil, thats what happen to ground planes, and its too late now, the boards are ordered.
I´ve tried everything and its impossible to get the ps loops shorter. Note that i also have small caps right at the fets, not only the lytics.
Btw i dont think the ground plane is too cut up as you say, it doesent even have any connections with the power components, except the speaker output and the caps.
And where would you put all the components that are now around the fets if neither long traces or gp cuts are allowed ?
Frankly i feel like i should ask fumac instead since he dident like the way your board was laid out, and that you are the only person on this forum to discredit others board designs. I start to think its all jealousy because other´s boards are actually better.
Anyways heres a shot of the groundplane only: http://diymania.hv4all.com/selfosc classd/new/ground plane.png
Btw i have a pair of class d caramps and they dont even have a groundplane.
I´ve tried everything and its impossible to get the ps loops shorter. Note that i also have small caps right at the fets, not only the lytics.
Btw i dont think the ground plane is too cut up as you say, it doesent even have any connections with the power components, except the speaker output and the caps.
And where would you put all the components that are now around the fets if neither long traces or gp cuts are allowed ?
Frankly i feel like i should ask fumac instead since he dident like the way your board was laid out, and that you are the only person on this forum to discredit others board designs. I start to think its all jealousy because other´s boards are actually better.
Anyways heres a shot of the groundplane only: http://diymania.hv4all.com/selfosc classd/new/ground plane.png
Btw i have a pair of class d caramps and they dont even have a groundplane.
http://diymania.hv4all.com/selfosc classd/new/selfoscdampbrdlasttry.png <-- lemme guess, worst this far.
Tekko, if you don't want to heed advice that you asked for then don't. Simple as that. You're obviously free to build your amp as you see fit.
You are still not figuring out all the high current high frequency paths.
Most of the switching RF current will try to flow from one supply rail to another through the small ceramic decoupling capacitors. Note that the ground plane is in the way from one capacitor to another... Gate drive current will also flow through the MOSFETs to the supply rails and then to the ground plane through the ceramic caps, then it will flow to the ground pin of the gate driver and to its +Vcc pin through the corresponding supply capacitor. Switching current in the lower RF range will flow through the main electrolytics.
You need to set up a little PSpice circuit simulation engine in your brain in order to produce good layouts 😀😀😀
Most of the switching RF current will try to flow from one supply rail to another through the small ceramic decoupling capacitors. Note that the ground plane is in the way from one capacitor to another... Gate drive current will also flow through the MOSFETs to the supply rails and then to the ground plane through the ceramic caps, then it will flow to the ground pin of the gate driver and to its +Vcc pin through the corresponding supply capacitor. Switching current in the lower RF range will flow through the main electrolytics.
You need to set up a little PSpice circuit simulation engine in your brain in order to produce good layouts 😀😀😀
http://diymania.hv4all.com/selfosc classd/new/selfoscdampbrdlasttry1.png <-- worst/crappiest one this far eh ?
- Status
- Not open for further replies.