Anna Conda
Banned
- Joined
- Oct 13, 2011
- Messages
- 159
- Helped
- 48
- Reputation
- 292
- Reaction score
- 140
- Trophy points
- 1,323
- Location
- Melbourne
- Activity points
- 0
There is one thing about the IR2110 series to be aware of: if you have a pretty loose layout and poor decoupling of the power stage, then, with an inductive load, when the upper switch turns off the midpoint can swing to below gnd for a short while, this has the effect of bringing into play the intrinsic internal diodes in the 2110 chip and can be destructive and or detrimental to operation, buffering will not necessarily solve this problem as it is the tie point in the middle of the power xtor pair that still connects to the IC, this is possibly why GDT's help the operation so much, as it cuts this connection and the IC (2110) is now unaffected by the source of the upper devices going several volts below gnd.
You see on some datasheets for hi/lo 1/2 bridge drivers, a voltage limit spec for this and sometimes a duration or dv/dt spec as well.
You see on some datasheets for hi/lo 1/2 bridge drivers, a voltage limit spec for this and sometimes a duration or dv/dt spec as well.
), by moving the main polarized storage caps from the main's bridge and mounting them directly to the half-bridge's pos/neg rails. Since I'm only testing right now the storage cap consists of a single 470uF cap but by mounting it directly to the half of the bridge I was testing along with the 1uF filter cap that was suggested on here, the switching issues seem to have went away while driving with the IR2110. So far I have only tested it to 90VDC but as I said, nothing got hot, and the frequency counter sat rock solid, as did the average DC voltage of the bridges output. Also after doing this I was able to reduce my gate drive resistors down to 10R paralleled with a reverse diode to speed turn-off time, this also allowed me to reduce the gate-emitter resistor down to 1K as well.
