Yup. I recover a whole 1/4 watt out of the 22 put in, and it's GREEN!
I wouldn't have bothered except that I really wanted more than 3V of gate drive, there wasn't enough current out of the snubber to just use a linear regulator, and I need voltages for the other 'lectronics anyway. So -- yet another regulator.
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
Didn't find your answer? Ask the community — no account required.
W
Winfield Hill
The efficiency of all of these schemes depend greatly on model values for MOSFETs, transformers (or inductors), and various parasitic elements. Inductor flybacks have no leakage-inductance loss issues, and do well on many other accounts. With a Schottky diode working at the 15V intermediate level, its inefficiency could be below 2.7%. The HV diode loss should be much less. The other losses are a matter of picking the right parts.
Thanks,
- Win
P
Phil Hobbs
I keep mine updated, and, whadda ya know, there was a 2N7002 right there. I only put it in so that the efficiency calculation would take the gate drive into account--it works fine without it.
Cheers
Phil Hobbs
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
J
Joerg
No idea why they used another symbol for that, it makes no sense. But yeah, updating might be needed soon on my side because people use lib parts a lot. I am a bit hesitant though because in the past that also caused older parts to get dropped from the libraries and then my own sims would no longer run.
Regards, Joerg
http://www.analogconsultants.com/
J
Joerg
No kidding, a car company did that. A teeny amount of brake power fed a teeny battery and they boasted about recuperating brake energy and all that.
I always try to furnish >10V in gate drive.
Regards, Joerg
http://www.analogconsultants.com/
T
Tim Wescott
Yes, winning is simulation is just the semifinals. Getting close to what you simulated is what makes for the real win.
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
T
Tauno Voipio
There is no 2N7002 in my copy of LTSpice (OS X version, just updated).
There is only a hole with three wires ending in empty. Which way was your transistor connected?
-TV
P
Phil Hobbs
Weird. It's connected as a source follower. Alternatively just connect the pulse source to the gate resistor of the main FET.
Cheers
Phil Hobbs
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
M
M Philbrook
last few updates of LT have excluded that model.
Why? who knows.. I used to use it much for basic concept testing.
Jamie
B
Bill Sloman
On Sunday, 24 January 2016 23:37:16 UTC+11, Bill Sloman wrote:
n
.
ringing immediately after switch-over - but relying on L3 as the smoothing inductor works fine, and emboldened me to increase C2 to 3.3uF - in fact a high voltage electrolytic would make even more sense.
SYMATTR SpiceLine Rser=0.035 Cpar=1.2p TEXT -416 760 Left 2 !.tran 0 30m 1u 10n startup TEXT -448 8 Left 2 !K1 L1 L2 L4 L6 0.99
The extra components in the gate drives are designed to kill the high-frequ ency ringing while slowing the switching transitions less than the single r esistors did. R1 and R3 damp direct drain-to-gate feedback, while the LCR n etwork is designed to block feed-back through the transformer.
I can't help thinking that a more complicated gate-drive structure, with mo re active components, such as you might buy form Linear Technology, could g ive you cleaner switching, while blocking ringing but I don't known enough about them to have any bright ideas, and in a context where everybody wants to use only off-the-shelf transformers (no matter how unsuitable for the j ob) it wouldn't make sense to learn more.
Bill Sloman, Sydney
P
Phil Hobbs
Okay, I put in a normal NMOS symbol and included a Zetex 2N7002 model, so it should just work.
I'll post this again with an attachment for all you Windows folks.
Cheers
Phil Hobbs
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
P
Phil Hobbs
J
Joerg
It still can't find the RFU diode but that is probably because I didn't update. Otherwise works fine with Windows 7 Pro.
Regards, Joerg
http://www.analogconsultants.com/
P
Phil Hobbs
It's a 450 PIV fast recovery diode. There should be another suitable one in the standard library.
Cheers
Phil Hobbs
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
J
Joerg
Must be a boutique part. Even findchips.com can't find it.
Regards, Joerg
http://www.analogconsultants.com/
P
Phil Hobbs
I just picked it out of the "choose new diode" dialogue box.
Some people are just hard to please. ;) I replaced them with DA2JF81, for which I could find a model.
So one more time--with feelin'--
Cheers
Phil Hobbs
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics
160 North State Road #203
Briarcliff Manor NY 10510
hobbs at electrooptical dot net
http://electrooptical.net
J
Joerg
Thanks, but that wasn't my intention, I could have just picked another one from my list as well. Was just wondering about such an exotic diode.
I am still in a bit of disbelief that Vladimir is gone.
Regards, Joerg
http://www.analogconsultants.com/
Join the Discussion
Have something to add? Share your thoughts — no account required.
Didn't find your answer?
Ask the community — no account required
Report Content
You are reporting this content to the moderators. They will look at it
ASAP.