Great Circuit there. Suggest adding a diode in reverse Bias to tamp down the spikes on the drain of the MOSFET. (Most MOSFETs have them built in - See datasheet) Note you need to ensure the reverse bias breakdown voltage is high enough. Check the data sheet on candidate power diodes and give yourself a 2X safety margin. Measure with an oscilloscope, but be careful to measure through a 10M ohm resistor. Voltages on those spikes could get high enough to blow the input on it. Also a good idea to use a small value, high voltage capacitor on the drain of that MOSFET as well. It will round down the peak voltage on those spikes a bit. Ceramic or Mylar Caps of .001 to .01 microfarads should do. Again, give yourself a safety margin on the voltage. [Interesting fact: If you’ve ever worked on auto ignition systems on pre-70’s cars, you’ll see they would use a Capacitor (old school folk call them condensers) across the ignition ‘points’ to keep the arcing down and keep them from burning out pre-maturely. Ignition points are the old school version of the MOSFET here.] Thanks for letting me nerd out.
Glad to see that someone else has also tried making a boost converter without any special IC's. I've made a buck converter with only some comparators and a couple transistors (for the sawtooth generation) for high-side switching a PNP transistor as fast as possible. Noise is only 1-2mV at a 2A load at 10V !!! 😄
It is relatively simple but complicated in comparison the the NE555 boost converter shown in IMSAI Guy's video #837. He uses just a single 555 timer circuit and he uses pin 5 in a feedback loop to make a kind of pulse frequency modulation.
You did good here. There are a ton of designs using IC's purpose build but enjoying to seeing your constant voltage boost converter. Those 555's have so many great uses. Good work my friend.
I keep wondering if the inductor spike affects the battery (if used). If the output were arranged around the inductor (treating it as if it were a battery, but without placing it in series with the real battery as is usually done) the output caps don't charge from the battery at all. Can you do a comparison of the two cases ? EDIT: Realized the feedback wouldn't work since ground has to be tied, and there's no way to tie it the other way (without becoming a flyback)
Voltage spikes, back EMF and noise generally do not effect batteries with some exceptions pertaining to flooded cell type but that's quite another topic..
So, for integrated circuits you use two 555 timers, and an LM358 comparator. The 7805 is a voltage regulator. So if I understand this correctly, the voltage regulator could do this job alone but isn’t nearly as efficient as a boost converter so is only used for an input voltage on the comparator. I don’t know what the efficiency loss is for this voltage regulator but I wonder if it’s substantial, if you could even make a smaller buck converter in place of the voltage regulator (the 7805)… I wonder if there is any efficiency gain by doing something crazy like that. 😅 Hope that made sense. Great video by the way, very helpful. 👍
the 7805 is only pushing 0.5ma through the 10k pot, which is actually too little for the datasheet (4.2-8ma bias), a 1k pot would be better. In short, there's not much at all to save with a non-linear regulator to produce a voltage ref.
Hi I am asking about the battery or any supply voltage of the circuit, I see there was 60 volt. across the inductance which is in some way between the + and + of the supply voltage . Will be there any problem ? Do we need to protect the Supply volatge ?
thats pretty cool man... i searched for a answer to my question which is how does a boost converter regulate its voltage accross the different amps. your solution is to change the duty cyle but not the frequency... do you think it would work if the pulse to the coil remained the same frequency and duty cycle but the number of pulses per second was the controlled thing for example if its 50 mhz maybe it only needs 100,000 pulses worth of energy to keep the output capacitor at voltage so just shoot 100,000 pulses into the inductor by disabling the signal from reaching the mosfet, would that work? maybe this could be achieved by putting the output of the comparator to the base of a transistor that would hold the gate of the mosfet low thus disabling the signal from 555 timer from turning the mosfet on. has anyone tried this?
Hi 5V Logic, thanks for the video! I built a SEPIC converter using this method with a 556 IC last year. It seems to work very well for relatively low current applications, but when I start pulling higher loads, it kills the MOSFET. I have blown about 15 MOSFETs at this point trying to figure out the issue. They aren't overheating, I am using a MOSFET driver, I have a heat-sink and check their temperature, no problems. It just runs for a few seconds and then the MOSFET fails. Have you come across anything like this with your implementation? I can't help but think (going to test soon) the 556 timer is opening the MOSFET for too long, letting the coil get too much charge, and then when it closes, there could be a huge voltage spike. I do have the second / "reset" side of the 556 tuned properly to make sure the main side resets even when a high duty cycle is needed. Any advice appreciated!
The SEPIC converter is new to me, in any case I'm not entirely sure how your switching circuit is wired. I can only suggest some tests: 1) check the inductance values with the switching frequency to be sure the current isn't too high. 2) be sure there aren't high voltage spikes on the gate. Best of luck!
Instead to use one toroi dal coil to get the power you need, use two or three parallel conection of course every single coil (L) should be drived by it's mosfet therefor the less power handle every single mosfet the less stress and overheat.I wish good evidence and resaults of your project.
@STLOGIC-s2d Thanks man, I found the problem. When drawing too much current for the battery the voltage would dip, this would result in the 556 timing circuit not triggering properly, which opened the MOSFET for far too long, leading to a huge voltage spike when it finally closed while fully energized. Thanks for the comment though!
It's just a simple voltage divider. 10k and 1k will divide the output voltage by about 10, so for 20v output, the feedback will output 2v. This is then compared to whatever voltage is set on the potentiometer by the 7805. If you want a much larger output voltage, one of the adjustments would be to increase the value of the first resistor in that divider, perhaps to something like 20k and 1k, so 100v output then becomes ~5v, and is easy to compare to the voltage from the regulator.
@@5VLogic It doesn’t look like Reset is connected on the first 555 on your breadboard (from left to right, first one being on the left). Simulating this I can’t make it work. If I connect the second 555’s Reset to Vcc the circuit doesn’t output any voltage (but does if I leave it unconnected - opposite from your breadboard configuration). I know a simulation isn’t perfect, but still confusing. Will try to build it.
Possible using 555 timer and RC network which converts the square of the 555 to a triangle wave and comparing it in another opamp to the output of the first opamp which is seen in the video , that's how PWM works , it's like one wave scans the compared wave . Summing it up : it is possible but you'll need two opamps instead of just one .
@@abov2983 Thank you! I’ve been experimenting with variations these circuits using 555. Has been eye opening (I have a math background, not an electrical engineering background). Have learned a lot. Thanks again. I’ll try this one.
@@millamulisha No problem , the output of the PWM gonna be trapezoidal (slowly rising and slowly falling) so you'll need to feed it to a schmitt trigger which will make it square before driving a MOSFET with it , that's why it's important to have a dedicated MOSFET Driver (for example TC4420) which has a schmitt trigger built in . also , if you're getting voltage spikes in your opamps or something like that , it's probably because of the quick rising and falling of the MOSFET so make sure you're using a high capacitance decoupling capacitor (for example 2200mF) at your power source because the capacitor will block those spikes and won't let them affect the rest of your circuit .
Sometimes simulators can be wrong, as long as the first 555 has a period longer than the second 555 in monostable mode, you can adjust the second 555 duty cycle with an op amp in differential amplifier (subtracting) configuration connected on pin 5.
