Great video! Straight to the the point and without bull poo-poo. But i need a 1mHz , how do i make one of those? How can we make +3 and -3 V for opAmps with as little components?
Thanks, Do you mean a Megahertz or a millihertz. Both are on the extreme side of the spectrum. For 1MHz I think it's gonna be tough, I recommend using a different circuit, maybe if you have a fast switching comparator you can pull it off but I definitely can't guarantee... The RC value is going to be about 10^-6 so like 1kOhm and 1nF I believe. Still, I'd opt for a different circuit; crystal or LC if you're feeling adventurous. For 1mHz the RC constant would be 10^3 so like 1MOhm and 1mF. Also you'll have to wait more than 16 minutes for one period... I want to remind that these values are very approximate. For dual power supplies you need some sort of dual power supply... ;) Simplest might be two batteries in series and then regulated with two linear legulators.
The LM358 is a general purpose op-amp, not a voltage comparator. This whole circuit works gracefully because the 358 is capable of impersonating a voltage comparator, and save the day unscathed. This is a nice, simple circuit which I would like to build, but I think I'll replace the LM358 with an actual comparator, like the LM393 (dual comp). Its open collector outputs require using a pull-up resistor
For those asking about values - I built this using 47KΩ and .015uF at the input. The timing circuit consists of a 10uF capacitor with a 10K pot. I added a 91Ω fixed resistor in series with the pot to limit minimum pulse width. These values provide an adjustable pulse width of approximately 400uS to 130mS. I recommend use of a 10-turn pot for easier adjustment. All the formulas needed to ascertain values are in the NE555 data sheets, but I found it easier to model this in LTSPICE and adjust values by trial and error for the exact range needed. Hopefully the values I used will give you a starting point.
Thanks, this works well. I needed a simple pulse generator with separate frequency and pulse width controls that don't interact. I put your circuit into SPICE and massaged the values to get the range I needed, then triggered it with a separate NE555 astable oscillator. Perfect!
Isn't it possible to swap the flip-flop with another op-amp? Maybe a positive feedback circuit or something? It'd be hall cool if it can be done just on 1 quad op-amp ic
Some people can actually feel the 50/60Hz when holding the metal body of a phone or other devices plugged into a charger. Myself included, it's a weird feeling sort of like a very low constant electrocution.
@@BillAnt That's not what you're talking about. And you will not feel 50/60 Hz from chargers, there are none there, the conversion frequency is much higher. :)
@@f33netYou can actually get capacitive coupling of mains voltage into the shielding of isolated chargers - sometimes the voltage differential can even be high enough to be painful when touching the shielding.
Was the problem with triggering the circuit by interference solved? I've made a very similar thing a few years ago, and had the exact same problem. Now I think that I'd add a one way low pass filter (slow charge, fast discharge via a diode) at the comparator's input, so that very short pulse wouldn't be enough to create a clock pulse. Touching it creates very long pulse compared to the one created by interference
I tried whit a 12 v dual rail supply, all resistors 10k and a 1 mocrofarad non polarized capacitor and i was also us ig lm324 quad opamp (only one amp) but it wasent working ... Any idea why?
Thank you, it's always great to get positive feedback and to know someone actually tries to reproduce the projects! If you're trying to get a 50% duty cycle and aren't getting one, it's probably the output of the opamp that's better at sourcing current than sinking it (or viceversa). You could use a potentiometer with two diodes, like in the 555 PWM classic online/RU-vid schematics. Hope that helps!
I really don’t see how you get 24 V on the top of the capacitor. Seems to me that you were simply putting ~12 V on both sides when the upper Mosfet turns on.
Thank you With the tips you gave, the program works well There is only one problem that it reads all the inductors, but it has a problem with small smd inductors like size 1206 and it doesn't show a number! I mean, where does the problem come from?
Hi, I'm glad you were successful in building the project! Although I don't have any SMD inductors to test, I'm relatively sure it's because they have high series resistance. Being so small they have to pack a high number of turns in there, and the wire has to be very thin. This means that the energy is lost quickly before it has time to oscillate enought to give a reading. I think the program discards measurements with less than 4 oscillations maybe(?), it's been a while... Also I don't know the value of your inductors, but low values are also not read, for similar reason, not enough energy is stored in them to oscillate enough and the Q factor is also lower, creating the same result. You could measure series resistance with an ohm-meter to verify the hypothesis if you'd like.
I don't want you to change the name, I want a .hex file that can be loaded directly with the NEO programmer tool without avdure or similar, I don't use linux, I have ch341 avr and nep SPI programmer.
I don't want you to change the name, I want a .hex file that can be loaded directly with the NEO programmer tool without avdure or similar, I don't use linux, I have ch341 avr and nep SPI programmer.
Hello I do the circuit according to the circuit I closed the video and checked it several times, I also used the hex file that you put in the new one It only says "open" on the LCD But the problem is that when connecting the inductor in the circuit, the board makes a connection and draws 1 ampere! Please guide 😢
@@erfanblaze4524 Hi, it's very difficult helping people through the internet debug circuits, but what I can do is give a few pointers and recommendations: 1) Check your transistor, if it's truly a BD137 and not a PNP. 2) Check the overall wiring. If you have and oscilloscope, probe all voltages around the transistor (the base in particular). 3) See if compiling your code and using my .ihex file gives the same result, if not there could be a problem in the code (potentially mine as well, although I don't think so). 4) Check the integrity of the arduino and it's pins, try an other GPIO to be sure, or test the one you're using with an LED to confirm it's working properly. Ideally check that it's capable of both sourcing and draining current. 5) Try swapping the transistor, in case it was bad or it got damaged. A MOSFET should work fine, but use a series resistor to limit the drain-source current. Place it on the drain side, not the source because it won't turn on properly. Unfortunately if the problem is hard for someone working on it in person, it's REALLY hard for someone over the internet. The general approach I recommend is to progressively excluding all possible faults and faulty components, one at a time by testing or swapping them. I really hope that helps, good debugging!
My circuit worked correctly and the problem was using the software >avrdudess< to put the hex code on the arduino nano, when I connected it directly to the atmega328 with the stk500 programmer and programmed the hex code, the problem was solved and the circuit worked correctly. he does Anyway, thank you very much for your advice 🥰🙏
@@5VLogicMy circuit worked correctly and the problem was using the software >avrdudess< to put the hex code on the arduino nano, when I connected it directly to the atmega328 with the stk500 programmer and programmed the hex code, the problem was solved and the circuit worked correctly. he does Anyway, thank you very much for your advice
There is only one problem that it reads all the inductors, but it has a problem with small smd inductors like size 1206 and it doesn't show a number! I mean, where does the problem come from?
Hi, I use avr-gcc as a compiler and avr-objcopy for the hex file. My makefile (and the whole project) is on GitHub and you can check it out and/or download everything there (link in description). In my case I run everything on Linux, but I think it should work on all OSs if the necessary utilities are installed.
@@5VLogic thank you I compiled with avr -stdio and put the hex into the arduino nuno And I closed the circuit and it says open on the LCD, and when I connect the inductor to the circuit, the board makes a connection! Maybe the problem is with my compiler, please upload hex file 🙏
Finally got it done, hopefully it's correct because I stopped working on this project since I turned it into the LC meter in my following video, but it should work. I'll upload the hex file for that one too in case you'd like to try it.
It's just a simple opinion - it's not a tip - what if you completely eliminate what's in the middle - and the part of the wiring that is outside (those external outputs that can be seen in this video) should be completely covered by the magnetic core - in theory - that voltage transfer from the primary to the secondary is done by magnetizing the core as strongly as possible - by making that hole in the middle - you make a "magnetic field obligation" towards the outside of the coil - it's just a simple opinion.
Use a 4 layer PCB or even better use copper sheet only, 1 turn per layer, increase the frequency to 350kHz. I made this kind o transformers professionally the size you are using is good for over 300W.
Sir, I designed a 24V half-bridge SMPS with an IR2153 IC. I'm getting a 24V output, but when I connect a load, the voltage drops to around half of 24V. What is the problem, and how can I fix it?
The voltage across the primary is always half of the input supply so I can assume 12V is correct. The over-voltage with no load can be caused by rectification of the ringing oscillations from switching transients. It's just me taking a guess though.
Yes definitely, although one should keep in mind that the inner layers are more prone to overheating because they're not in contact with the air. Also, they're more expensive.
Yes I forgot to mention it but the other reason I made the single layer PCB was to reduce the parasitic capacitance across the primary. In general also between primary and secondary it can be somewhat high depending on the distance they have between them.
Also! Instead of a single trace you can draw double trace to decouple by a little the parasitic capacitance, but i don't know if this technique worth the effort.
Nice! Keep in mind that flyback transformes cores have an air gap, and that's no good for forward converters, it may be influencing the behavior of your circuit and influence efficiency for sure.
I forgot to mention that I ground down the ferrites to remove the gap, it's not the most beautiful solution but it works well. The important thing is that the primary inductance is sufficient, which is what the gap effects.
@@5VLogicIt doesn't need to be pretty, it only needs to work :) the only issue possible with sanding is it may not be 100% gapless due to flatness imperfections, but it wont make a huge diference anyway.
Hi 5 Volt, thanks for the video. please, share your gerber file to github. btw, can you make video how to write code on Vim and compile it. i have been using arduio ide, so i want to try learn something new. thanks before
Hi, Sorry for the super late reply, I guess your comment flew under the radar. I'll keep you suggestion in mind, but since I'm not that knowledgeable on this topic it might have to wait. Also not many people use vim and Linux unfortunately so I'm not sure the video would do well. Still, thanks
Good job! I wonder if instead of trial and error method for choosing the range, you could use an ADC to track the very first points on a charge curve. Since it is linear in this region, you would instantly have a decent approximation
