- Joined
- Sep 12, 2007
- Messages
- 9,399
- Points
- 113
LPF Donation via Stripe | LPF Donation - Other Methods
Links below open in new window
ArcticMyst Security by Avery
Inverting buck/boost - Case Positive Diode Driver
- Thread starter rhd
- Start date
rhd
0
- Joined
- Dec 7, 2010
- Messages
- 8,475
- Points
- 0
I don't see why the inversion is necessary. Why not just flip the cell and put the switching and sense elements "after" the load?
Something like this crude paint sketch:
(BJTs are easier to draw than FETs)
That's smart. I think it would work in theory.
One slight glitch is that if your IC's FB is say 1.25V, that could very well be higher than the voltage at the current sense monitor's SENSE+ terminal. It certainly won't work with a ZXCT1009 that takes its power from the line being sensed, but there are plenty of current sense monitors that have independent power pins.
rhd
0
- Joined
- Dec 7, 2010
- Messages
- 8,475
- Points
- 0
I don't see why the inversion is necessary. Why not just flip the cell and put the switching and sense elements "after" the load?
Something like this crude paint sketch:
(BJTs are easier to draw than FETs
Many buck ICs are meant for driving P-Mosfets. What would you suggest for using them in your circuit? Drive a P-Mosfet that pulls the gate of the N-Mosfet high? I'm wondering whether the extra switching may add too much delay when dealing with high frequency (ie 1.5 to 2 MHz) ICs.
djQUAN
0
- Joined
- May 27, 2013
- Messages
- 1,154
- Points
- 63
An open drain P channel driving the gate of an N mos is not ideal due to not enough pull down current capability.
You could rig a boost controller IC in buck mode with the negative side switched rather than the positive supply.
You could rig a boost controller IC in buck mode with the negative side switched rather than the positive supply.
rhd
0
- Joined
- Dec 7, 2010
- Messages
- 8,475
- Points
- 0
Attachments
- Joined
- Aug 14, 2013
- Messages
- 2,655
- Points
- 63
It looks like you may be in need of either a
gate clamp or the MC34063.
It's funny reading over this thread and
seeing nearly the exact design issues I
have been dealing with for the past couple
years.
gate clamp or the MC34063.
It's funny reading over this thread and
seeing nearly the exact design issues I
have been dealing with for the past couple
years.
Last edited:
Hiemal
0
- Joined
- Dec 27, 2011
- Messages
- 1,443
- Points
- 63
This wouldn't work?
Wouldn't work because the N-channel's gate would drain far too slowly with only a 10k resistor as the pulldown.
That's why most gate driver IC's use a complementary push pull pair, so that the gate can be pushed high by one switch, and then pulled low by the other with nearly zero resistance both ways.
Last edited:
- Joined
- Sep 12, 2007
- Messages
- 9,399
- Points
- 113
Something like this gate drive idea should work. Just switch the PNP and NPN for inversion.
- Joined
- Aug 14, 2013
- Messages
- 2,655
- Points
- 63
I have used that before. Very fast and very
effective
effective
Something like this gate drive idea should work. Just switch the PNP and NPN for inversion.
rhd
0
- Joined
- Dec 7, 2010
- Messages
- 8,475
- Points
- 0
Something like this gate drive idea should work. Just switch the PNP and NPN for inversion.
Cyparagon / TLS, What do you think of this part, to accomplish the inversion above?
- Rohm QSZ2TR http://rohmfs.rohm.com/en/products/databook/datasheet/discrete/transistor/bipolar/qsz2.pdf
and is there need for the 560 and 22 ohm resistors shown?
Last edited:
- Joined
- Aug 14, 2013
- Messages
- 2,655
- Points
- 63
That is non-inverting. For inverting, you
need the collectors tied together. Nice part,
though
- Joined
- Sep 12, 2007
- Messages
- 9,399
- Points
- 113
The flyback driver example just amplifies the gate drive; it doesn't invert it. That device won't invert the gate drive either. The BJTs have common collectors, and need to be either separate, or have common emitters.
22 probably isn't necessary, but 560 is certainly needed to limit the base current. The IC's gate drive is basically powering a short without that. It may also need to be higher or lower, depending on what sort of gate current the switching IC can typically handle.
22 probably isn't necessary, but 560 is certainly needed to limit the base current. The IC's gate drive is basically powering a short without that. It may also need to be higher or lower, depending on what sort of gate current the switching IC can typically handle.
rhd
0
- Joined
- Dec 7, 2010
- Messages
- 8,475
- Points
- 0
Seems like the 560 should be lower. IC can typically handle 500mA gate current.
