Guru
Joined: 13/10/2014
Location: AustraliaPosts: 1688
| Posted: 11:12pm 08 Mar 2024 |
An earlier version of the PWR board showed them on Q1 to Q8 as well, there was an earlier circuit during the first design stages that had the Beads in the Gates.
There was a lot of work carried by Wiseguy and Poida investigating exactly what was happening with the FETS, Ferrite beads and ringing in inverters, and they were moved to the FET Drain. If you can fit them by all means try them, in may case everything looks good now, much better then the typical inverter power stage ringing.
Now the ferrite beads that Wiseguy and I use will not fit over the wide part of the FET leads.
There is also limited space on Q1-Q8 to fit them to the Drain, that lead is bent straight down after the wide shoulder and onto the +48v power plane just below the FET body.
Also the Q1-Q8 drains connect to the power supply +48v rail, this has a lot less HF switching noise compared to the Choke / Toriod input terminal T4 for Q9 to Q16.
Fitting just the low side Ferrite-beads made both the HI and LO side PWM output very clean and almost identical in my testing when measured with a 380mHz bandwidth DSO.
You will also notice that I did NOT show Diodes D7 to D18. The original design was using HY4008 FETS, however I had a failure of one of those Diodes when using HY5608W.
NOTE: That happened under an extreme almost 30kW pulse condition, I really brutalized that the Big inverter and Wiseguys power board during the first 6 month of torture testing.
The Diodes turned out to be a weak link and underrated compared to the body diodes in the HY5608W, which are the recommended Power FETs to use. Wiseguy and I decided to leave them out as they are not really needed with the HY5608W FET.