Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Need help with capacitor boost charger, acting strange

Status
Not open for further replies.

jpoopdog

Member
hello,

I am at the moment trying to make a boost converter to charge a capacitor bank.

I am using this circuit diagram **broken link removed**
not shown in the circuit is that the GND pin of the lm311 is meant to be conected to 0v.

and it charges fine.
the problem is though, when its reaches a certain voltage, it should switch off, but it doesnt.
What is meant to happen is that the LED should illuminate when the target voltage is reached. also the 555 should turn off.
I am not really sure if this is happening,
the led slowly gets brighter, and remains dim. if i disconnect power then reconnect power, it turns bright, and there is no more audible noise.
however the mosfet then gets super hot, and wether or not i disconect the capacitor, it begins to discharge very quickly which it shouldnt.

If anyone knows how i can prevent this, it would be greatly apreciated.

An alternative though would be if anyone knows how i can attach a 100v digital voltmeter to the capacitor bank to read the voltage, so when it reaches 330v or therabouts, it will display 100v on the voltmeter, like a percentage completed or something.
If anyone knows how i could do this without blowing up the voltmeter, or causing the capacitor to discharge, thated be great.
 
The reason the mosfet is getting hot is because the voltage on that capacitor is higher than 550 volts and the mosfet is breaking down in avalanche.
To make the multimeter read 100v instead of 330 just add another voltage divider, just like what the circuit uses now to tap off 6 volts for the lm 311.
you'll need a resistor of something like 470K ohms instead of 22K, the exact value will depend on the impedance of your volt meter.

I couldn't tell you why the comparator isn't working, try replacing the 1 k ohm resistor with a 4.7K, the comparator may be having a problem sinking 12 mA. or replace the comparator, it may be damaged.
 
Last edited:
ill try and find that impedance value.
Understand i dont really know how a voltage divider works, this is the first circuit that uses one which ive built, to my knowledge.
 
ok, i dont know how to find out the impedance value, any suggestions on how i would measure this?

also, ive tried reducing the number of turns in the inductor, and it hasnt really solved the problem, its now 150uH, ill admit its limit now im guessing is like 350v, which i suppose is good, but it just doesnt seem right.

i might have to try testing it using my ATX psu, rather than battery bank. at the moment the charging is quite slow, just for a single capacitor.
ill keep tinkering till i get it right, and it would be nice to test it with a massive cpu heatsink so theres no fear of blowing my last fet.
 
Last edited:
first things first,

the reason the mosfet is blowing up is because it is breaking down in avalanche, you need to make sure the voltage on the drain does not go higher than 500 volts.
just because the capacitor reads 325 does not mean anything. you need to ensure a low impedance path from the drain to the diode to the capacitor. (how long are your wires, is this a circuit board or point to point wiring?)

on the other hand, the reason it is blowing up could be because the "on time" is far too long. (are you sure that 10nF capacitor is only 10 nF?)

to prevent the mosfet from blowing up, use a current limited supply. something reasonable like 1 amp is all this circuit should need.

On the impedance value of your volt meter: most digital meters are 1 M ohms. some are 10M ohms, and others could be somewhere in between.
 
actually the fet blew up when i was removing the capacitor from the breadboard.
Also the 10nF cap is measured at 9.3nF

This circuit causes little heating in the FET since i reduced the turns on the inductor.


ive been wondering though, could the cause of discharge be the fact the positive of the capacitor would have a 1mohm resistance between it and ground? or perhaps could current be leaking through the diodes? for the diode i am using 3x 1n4007's. should i try something different?
because after i turn this off in 10 seconds the voltage drops to 50v
 
Your circuit would work better with a fast diode rather than a slow 1n4007. there's two of them in every computer power supply.

how much capacitance do you have? calculate the time constant to get an idea how long it should take to drop the 300 volts down to 50.
 
i think its dropping because of the voltage comparators 1mohm resistor. perhaps if i used another MOSFET, when the thing is off, it can isolate the capacitor as well.

Also, do you know what the part number might be, for a diode in a pc power supply? or what the diodes would look like?
 
First thing is first. Because the 555 timer has a fixed on/off time (set by the 4.7k, 15k and 10nF cap) the value of your inductor is critical. When the mosfet is 'on' the current through the inductor ramps up until its core is saturated - at which time it can't put any more energy into making a magnetic field, so it effectively jsut becomes 'a piece of wire' - one with low ressitance. If you were to leave the mosfet on, it is switching the +12V fom your supply, to ground via a very low resistance. Depending on how much current your power supply can deliver, this can be many amps, which would cause your FET to get extremely hot.

Also, the 'off time'. During this phase the inductor dumps its energy into the main cap via the diode. The voltage at the diode is set by the capacitors voltage (which increases as its charged). Eventually the inductor runs out of juice so that phase is complete. If the 'of-time' is too short however, the inductor's current won't have time to reach zero, so then ext 'on time' will cause it to reach its saturation point quicker....etc.etc...

So, I would remove the inductor, mosfet, and diode, and check that the 555 is actually oscillating. (maybe bad connection, wiring etc..). Once that checks out you should check the ressitor values on the 555 (between pins 6 and 7, an also the resistor between 7, and 8) so that the 'on-time' is very short, bu the 'off time' is very long. You can use various online applets to caculate the values for on, and off times :). Then, using your inductor value (can you measure this?) you can work out how long it will take for the current to reach a certain limit - say, 1A. Inductors have a maximum saturation curent which you want to stay under.

Thats probably and overly complicated explaination, but this kind of circuit really does require specific part values for the inductor used, to help efficiency, but also to actually 'make it work'. I see no problem with the comparator part, it should turn of the 555 (and therefore the MOSFET) when the cap has reached its set voltage - but the led will most likely just go 'dim' as the capcitor discharges through the resistors, so it goes below the thrshold, turning the 555 on again - just long enough fpr the capacitor to reach the threshold voltage once again. So in that state, the whole system is quickly turning on and off as the capcitors voltage sits bang on the set voltage.

I've made a few such cap chargers, and when they work, they work well, but without a scope (I've only just got one...) it can be quite dificult to get things to run without over heating.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top