High Voltage Forum
High voltage => Voltage Multipliers => Topic started by: rustedone on October 08, 2019, 05:37:49 PM
-
hello everyone.
I want to make boost convertor as power supply for induction forge ... after looking around i came up with a simple design . and
want to take a look and see it can do it's job.
-
Shows the sketch of a flyback converter, but would need a lot of work to turn into a functional circuit.
-
well the thing is while it is close to fly back convertor ... it's something different cause it doesn't use transformer . and it was doing fine in programs that explains it . but yes... it has tons of room for improvement. but what kind of problem will arise using it like this?
if i know . at least i will know what to improve in it . like : should i make feed back voltage for it . or in the first place the devise I want to use it on is not that sensitive to changes in voltage or amperage ... its all good as long as its high (induction forge) or at least that's what I believe unless it's not true.
so what is the problem of this circuit cause ... there were a lot like this when i searched and non mentioned the problem...
another thing that i just found that usually boost convertors show problems when trying to boost a lot like x10 of input or something but here it's only x2 with 50% duty cycle ... so i doubt there is really much need for delicate work ... but that's just my idea
-
Sorry, I tend to use "flyback" and "boost" converters synonymously. Perhaps flyback implies a resonant cap and transformer coupling as in typical CRT supplies.
List of issues:
Single wire from drive to gate. No common return connection.
No voltage feedback (regulation).
No input cap on bridge diode outputs.
Output cap is much too small. Perhaps you're counting on the load's bulk capacitance?
No inrush current handling.
No control to prevent Q1 turning on during the initial inrush current.
7555 chip may not have enough output current to drive Q1 gate - I haven't checked. Gate drivers are more typical between control and gate.
No snubbers (R/C camping) on Q1 and/or the diode.
-
For that power level, I would consider 3-4 phase interleave.
I've got a two-phase PFC here which is almost working; it likes to blow fuses and/or transistors so I need to do some debugging yet.
(https://www.seventransistorlabs.com/Images/1kW_PFC.jpg)
UCC28070 controller.
Over its development, I've put many weeks of labor into this design. And I'm a professional EE. This is not the domain of 555s in open-loop toys!
Tim
-
Well said Tim
Other than the 7555 being woefully inadequate to drive the gate there is no gate dV/dt control
No dV/dt control on the rectifier so even more EMI and heat
as said no snubbers so no am/fm radio bands for many tens of metres.
In fact at this power level one would have to use lossless snubbing that adds another level of complexity
We await the smoke signals
-
Well those are not even critical issues, the biggest is lack of current control.
I don't have a lot of good references handy, unfortunately, but I strongly recommend reading some introductions to switching circuits. At the very least, say, the classic Unitrode application notes -- search for them, they're free from Texas Instruments. Pick up a good book or three on the subject. Above all, test little toy circuits demonstrating your mastery of the subject -- only then should you scale up and build something that can cause injury. Exploding semiconductors and capacitors, and arc flash, are no joke!
Most of my early "study" was with small, breadboarded circuits, in the 1-20W range; here's one of my favorites (https://www.seventransistorlabs.com/Images/Discrete_Tube_Supply.png). I wouldn't use the same approach on a larger design, or at all, really -- there are better ways to control things, especially as the cost of a $5 controller pales in comparison to, say, the $20 of capacitors you need for a larger design. The key is developing the intuition and mastery of the subject. Being able to realize such a system in, well, almost seven transistors, is one way to illustrate that mastery I think.
It is absolutely possible to design and build this -- but unless you get extremely lucky on your first build, it will take many months of dedicated research, design and refinement until you have a complete, working, reliable design. (And only then will it be productive to discuss efficiency improvements and EMI reduction!)
Tim
-
thanks for all info. you see. one the problems that i encounter when trying to add feed back. and other controls is that they work for low voltage and amperage . just like one the friends mentioned most of good drivers are at low status ...
and if i use something like ir2153. then would work better i guess .
maybe i should use ntc to 1: 2 transformer . then a rectifier with some caps... that should make things simpler ...
-
and thanks TIM :) this unitrode that you recommended is super interesting . but i'm rather in hurry for now. i try making power supply with transformer while reading it. hopefully make things easier ...
-
I wonder if I add 10a ntc before 10a rectifier . it would make the current control better ?
-
i wonder if you guys know any way to actually let say 'play' with high power whether voltage or amperage ? like strong .driver. i saw one of ir21 series that had HI LI pins. so if i could send pwm or spwm with microcontroler to it then it will give a good control on switches . but wasn't very successful cause i couldn't really test it or didn't know any one to ask. maybe i should make different post for it .
-
another thing that i just remembered is that the 555 timer ic that was in circuit was not old 555. but a new ic that actually use 555 and can send to 15v signals. so i think it can do its job...