I need the PWM signal, coming from the TL494, and the inverse of the signal. TL494 cannot make the inverse of the signal, that I need.
I can invert the PWM with a single transistor, like a BCxxx, or with a TTL/CMOS inverter, or I can connect the TL494's transistor's in the way, you can see on the schematic.
When I use TL494 in push-pull mode ( OUTC tied to VREF ) then both the outputs will have the same duty cycle. For example 30% 30%, but I need to drive the high-side MOSFET with 30% Duty-cycle, and 70% duty-cycle, the lower MOSFET.
I can invert the PWM with a single transistor, like a BCxxx, or with a TTL/CMOS inverter, or I can connect the TL494's transistor's in the way, you can see on the schematic.
When I use TL494 in push-pull mode ( OUTC tied to VREF ) then both the outputs will have the same duty cycle. For example 30% 30%, but I need to drive the high-side MOSFET with 30% Duty-cycle, and 70% duty-cycle, the lower MOSFET.
I was experimenting a little bit in my simulator.
I created a schematic, wich makes about 750nanosec dead-time.
Is that too much? Or too short?
T1 transistor is the TL494's output transistor. I applied 50% duty-cycle square-wave to the transistor.
On the schematic there is CD4069, becouse the simulator haven't got CD4049 in it's library. I'm going to use CD4049.
What do you think of this? Will it work in the real world?
I created a schematic, wich makes about 750nanosec dead-time.
Is that too much? Or too short?
An externally hosted image should be here but it was not working when we last tested it.
T1 transistor is the TL494's output transistor. I applied 50% duty-cycle square-wave to the transistor.
On the schematic there is CD4069, becouse the simulator haven't got CD4049 in it's library. I'm going to use CD4049.
What do you think of this? Will it work in the real world?
I have /now/ a few experiences. The efficiency is very low (about 70%) at 20A output current, becouse the U_in is far more higher than U_out, and /becouse of this/ the switching losses are dominating, not the ohmic (RDS_on) losses.
I couldn't decrease the switching frequency, becouse then I would need a huuuuge output inductor. The core is quite big, about 4cm outer diameter, and 2cm inner diameter. Only 4 turns are on the inductor, and I think it still saturates, when I go over 20A. I can't decrease the turns on the inductor, becouse then the dI would be too high. Now (with 4 turns) the core is very hot even if I use active cooling, with a small 486 CPU fan.
I will try to get/make an inductor core, wich can "hold" more energy.
I couldn't decrease the switching frequency, becouse then I would need a huuuuge output inductor. The core is quite big, about 4cm outer diameter, and 2cm inner diameter. Only 4 turns are on the inductor, and I think it still saturates, when I go over 20A. I can't decrease the turns on the inductor, becouse then the dI would be too high. Now (with 4 turns) the core is very hot even if I use active cooling, with a small 486 CPU fan.
I will try to get/make an inductor core, wich can "hold" more energy.
Hi danko
How is the project coming along. If not to good there is a thing you can do/think about it.You and I try to drive N-channel mosfet with IRxxxx, trying to drive its gate over the input(max circuit voltage). But there is much simpler way. Use P-channel mosfet and you dont even need IR, the PWM chip can drive it directly.
What do you think?
How is the project coming along. If not to good there is a thing you can do/think about it.You and I try to drive N-channel mosfet with IRxxxx, trying to drive its gate over the input(max circuit voltage). But there is much simpler way. Use P-channel mosfet and you dont even need IR, the PWM chip can drive it directly.
What do you think?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.