Thank you for your videos. You have great explanations and logical progression through subjects. I also love the format where you talk theory -> simulate in LTspice -> build/demo. Please keep making videos!
This is such high quality content. Great explanation with demo and also comparison of simulation and real life. I really hope more people notice your channel.
This has been a super helpful video. One detail I found particularly useful was how you arranged your push-pull BJT's for driving the gate on the high side. I always put a PNP on the high side an NPN on the low side, but that inevitably resulted in a leakage when VCC would cause current to travel from emitter-base on the PNP, then from base-emitter in the NPN. Your arrangement is much better
Thank you Fesz -^\/v^--- ... when I want to know precisely what is happening in a circuit, I head straight to your channel. Sometimes I need to replay video a few times to comprehend, electronically, what's going on. Your channel = wonderful knowledge base. Using 'Synchronous Buck Converters' here lately with various tool batteries (e.g. 56-V 18650 and 18V 21700 batteries), for camping or emergency prep. Using a 9:1 ratio, hand-crank, axial-flux alternator with 3-phase Schottky diode, bridge rectifier to charge a bank of super capacitors. This, in-turn, charges my tool batteries via boost Converter. (Also have mobile 400W solar charging alternative/ backup. I'm using XT-60 connectors throughout for modular use of components. 👍
This was an amazing lesson, your style of asking the questions and answering later is very good to make you think first!! Only one thing missing, what is causing the dead time? the voltage drop across the bootstrap diode? Just saying because I noticed the low side switching is connected directly to V_stab ,high side switching goes thru bootstrap diode first. Thank you so much for taking the time to do these videos!!
I have watched a few of your videos now and I must comment on how immensely helpful these videos are! I'm working my way making detailed notes of the Art of Electronics book and these kinds of videos make for great supplement!!! Keep it coming! :D As my thanks, I always let the ads run through in full for your videos, as little of a contribution as that makes, and will now be one of your patrons!
Thanks for a great video. I am planning to build a 100 amp 48 to 24 volt buck converter for my solar system. Plan to add batteries in series to increase to 48 volts, but still want to operate a good 24 volt inverter. Right now I am ordering parts and trying to learn basics of inverters, and find your videos very instructive. Would like to see in future some tutorial on buying or making the 100 amp inductor.
Thanks, this is a very usefull explanation. Also, there are still things to be done to increase the efficiency even further. I guess in the choose of components. Demonstration is best. Still hints in that direction are interesting. :-)
These videos are educational and I already spent a good time making my own buck converter from 0 V to 24 V as a high side driver with source to ground. In this case, my PWM was a teensy 3.6. But, for something practical and without issues, just buy an Adjustable LMxxxx. Building your own buck converter with feedback is just frustrating. But the knowledge you gain is invaluable.
I'm not very familiar with the board you are referring to, but regarding feedback, is there a dedicated block in the controller to get the ADC value to control the PWM or is this done in SW? I'm asking because if the feedback is SW based, that might be the main issue.. The reaction speed is not that good and depending on implementation you might swing between max and min duty cycle every time the uC has time to measure the supply output. Maybe you can give a few more details on the implementation you are trying.
The main point of the SMPS is efficiency, especially for high current PSU's. The LM series are linear and are no where near as efficient. Both have pros and cons and their place in electronics power supplies.
Nice! Now you basically build a Class-D power amplifier ;) Next stop: dead-time adjustment. Google hint; look for XOR deadtime circuit. There is actually one where you only need 1 resistor to select the dead-time value!
Hello Piet! I don't think I will go into that much detail; at least not at the moment. I am aware you can build a complicated and accurate circuit with logic gates, but at some point its much easier to just get a dedicated driver that has everything built in. I was looking at some IC's from Infineon like 2ED2304S06 which has 2 inputs (for low and high side) but they also have IC's with a single input and an external resistor for the deadtime setting.
@@FesZElectronics True, I was thinking about it since, you were doing everything discrete already. (buck converters can also bought of the shelf). There are a lot of IC's that can control dead-time and even more. Really like the IRS20957S and the IRS2092S. Both are for audio, but can also be used to other things (obviously). LM5116 is also nice, as well as some other parts (can't remember the name) with actually adaptive dead-time control. When a u-controller is used I would rather control the dead-time in the code. Very interesting!
On some comments below suggesting to focus on content and less face time, I prefer some human content in technical videos. May I suggest a small continuous window in the corner for the presenter. I note that other channels use this technique very effectively.
I did resort to that in the mean time - this video is actually a few months old. If you have time maybe you can look at a more recent video and let me know what you think.
Hello @@haimt, I'm seeing the same error message but I'm not sure what you mean by changing the prefix to X. Where is that? Please clarify. I'm new to LTSpice and just beginning to learn it thru these YOuTube videos and thanks to FesZ, I'm coming along!
Hello; the problem here is that every component needs a prefix so the simulator know what type of component it is (for diodes its D, for capacitor C etc.) For non-standard components it is usually "X". So when you add a new component, if there is no prefix added, it will create problems. As @haimt1464 said, just right click on the component - to open the component attribute editor and on the first line (prefix) add "X".
For some reason I had to hold down both CTRL and ALT button and then right-click the mouse button to access the component Editor. Thank you for your help and these highly educational videos. I'm now learning LTSpice. it's a great Tool!!
Hi guys,i have buid a syncronous buck and i use ir2110 in half bridg configuration to drive the fets.All seems to be good but,my bootstrap cap is getting hot .Can anyone have this problem before?
I think, it would much better for understanding, to keep camera on the schematics when you explain them instead distraction attention by changing camera focusing from schematic to your face and vice versa frequently.
Such a great video, it will be nice if you make a video about H-Bridge High voltage Inverter Protection, I'm currently working on something like that even with having short circuit protection at the Boost side the mosfet's at the H-Bridge burn while testing short circuit I used SG3525 IC for the job
You might have issues with the voltage rating of the transistors - they should be able to handle not just the DC input but also any spikes that appear in the switching nodes (up to 10-20% of the DC). I'm not sure if and when I will cover H-bridge inverters though.. Hope this helps! Cheers!
@@FesZElectronics Thanks i fixed that, the problem was with low side mosfet's gate resistor which was too high & i also added pull down resistor and now it works just fine thanks for ur nice videos and help
Your videos are very very helpful. Nobody can explain the way u do. I want to learn ltspice, can you guide where i can get good material to start with.
Hello Harish! I'm happy you enjoyed my videos! Well I do have quite an extensive series related to LTspice starting with ( ruclips.net/video/JRcyHuyb1V0/видео.html )
Its because of the way in which current is flowing - from ground towards the switching node; for the diode to be forward biased, the anode (connected to gnd) needs to be at a higher voltage potential than the cathode (the switching node); The current is flowing from the inductor trough the output capacitor/load and closing back trough the diode.
That is usually built into the driver; you have dedicated SMPS converter IC's where this dead time is fixed (it has some tolerance but its not user settable); but you also have some IC which only perform the gate drive function where this time delay can be set; Regarding the delay length, ideally you want it as short as possible, but in practice, you will need to set it to account for transistor switching delays and also keep a bit of tolerance margin. Usually it will in the tens or hundreds of nseconds but based on specific design it could be higher or lower than this.
Thankyou for great explanation. But when we increase the source voltage to 30 then Vd becomes 30 and gate voltage be nearly 11 as fixed in that case.. So how can I get 30 voltage at source side during the turn on time of transistor? According to my thinking, the voltsges in that case at source becomes Vs=Vg-Vth which roughly bevomes 9. Need clarification from your side. Thankyou
I'm not sure I understand the question. If you are referring to the high side transistor, supplying the gate is done by the bootstrap capacitor (C8). So the gate-source voltage Vgs is not coming directly from the input supply, but rather this floating capacitor. The voltage on the upper side of C8 will always be ~ 5V more (5V because of the way the V_stab voltage is obtained) than the voltage in the source of the high side transistor - so at 30V Vin, C8 is charged by 5V and when the highside is on, then the voltage on C8 reaches 30V+5V so enough to drive the high side.
I have downloaded the simulation and corrected the "X" in the component type as described below. However I am getting a problem where the mfets are shifted to the right and are not connected into the circuit. I just started using the program so don't know how to edit/fix this yet. I have downloaded version 17, and it seems that file is for version 4.
Hello yancy! The schematic was done in LT17 but that doesn't really matter. the point is depending on what symbol you are using for the transistors you might need to re-connect the traces, just make sure the terminals are connected and liked (between model and symbol) correctly.
FesZ Electronics Thanks for replying to my comment. I did manage to get the simulator working as I discovered the "move" tool. When I moved the fets between the wires, everything arranged itself properly. It was very good as I was really having a tough time understanding the boost voltage circuit. However after seeing it work in the sim, I am beginning to "get it".
![KSI - Thick Of It (feat. Trippie Redd) [Official Music Video]](http://i.ytimg.com/vi/At8v_Yc044Y/mqdefault.jpg)
As always, great explaination!
Glad you think so! Cheers!
Thank you for your videos. You have great explanations and logical progression through subjects. I also love the format where you talk theory -> simulate in LTspice -> build/demo. Please keep making videos!
no one ever explained a synchronous buck regulator as good as you...this is how it should be taught in colleges
This is such high quality content. Great explanation with demo and also comparison of simulation and real life. I really hope more people notice your channel.
This has been a super helpful video. One detail I found particularly useful was how you arranged your push-pull BJT's for driving the gate on the high side. I always put a PNP on the high side an NPN on the low side, but that inevitably resulted in a leakage when VCC would cause current to travel from emitter-base on the PNP, then from base-emitter in the NPN. Your arrangement is much better
Possibly the best explanation I could wish for! Thanks for sharing it such wonderfully
Glad you enjoyed it! Thank you!
Thank you Fesz -^\/v^--- ... when I want to know precisely what is happening in a circuit, I head straight to your channel.
Sometimes I need to replay video a few times to comprehend, electronically, what's going on. Your channel = wonderful knowledge base.
Using 'Synchronous Buck Converters' here lately with various tool batteries (e.g. 56-V 18650 and 18V 21700 batteries), for camping or emergency prep.
Using a 9:1 ratio, hand-crank, axial-flux alternator with 3-phase Schottky diode, bridge rectifier to charge a bank of super capacitors. This, in-turn, charges my tool batteries via boost Converter. (Also have mobile 400W solar charging alternative/ backup.
I'm using XT-60 connectors throughout for modular use of components. 👍
This was an amazing lesson, your style of asking the questions and answering later is very good to make you think first!!
Only one thing missing, what is causing the dead time? the voltage drop across the bootstrap diode? Just saying because I noticed the low side switching is connected directly to V_stab ,high side switching goes thru bootstrap diode first.
Thank you so much for taking the time to do these videos!!
Or maybe It's probably due to the cross over on the push pull drivers?
Fascinating, many thanks.
A lot of good information thanks
I have watched a few of your videos now and I must comment on how immensely helpful these videos are! I'm working my way making detailed notes of the Art of Electronics book and these kinds of videos make for great supplement!!! Keep it coming! :D As my thanks, I always let the ads run through in full for your videos, as little of a contribution as that makes, and will now be one of your patrons!
Thanks for a great video. I am planning to build a 100 amp 48 to 24 volt buck converter for my solar system. Plan to add batteries in series to increase to 48 volts, but still want to operate a good 24 volt inverter. Right now I am ordering parts and trying to learn basics of inverters, and find your videos very instructive. Would like to see in future some tutorial on buying or making the 100 amp inductor.
Thanks, this is a very usefull explanation. Also, there are still things to be done to increase the efficiency even further. I guess in the choose of components. Demonstration is best. Still hints in that direction are interesting. :-)
Very informative ..tnx a lot
Good explaination!
Good explanation sir.
Keep going on
Great content
Expecting more videos from you in field of converters drive.... Great content...
These videos are educational and I already spent a good time making my own buck converter from 0 V to 24 V as a high side driver with source to ground. In this case, my PWM was a teensy 3.6. But, for something practical and without issues, just buy an Adjustable LMxxxx. Building your own buck converter with feedback is just frustrating. But the knowledge you gain is invaluable.
I'm not very familiar with the board you are referring to, but regarding feedback, is there a dedicated block in the controller to get the ADC value to control the PWM or is this done in SW? I'm asking because if the feedback is SW based, that might be the main issue.. The reaction speed is not that good and depending on implementation you might swing between max and min duty cycle every time the uC has time to measure the supply output. Maybe you can give a few more details on the implementation you are trying.
The main point of the SMPS is efficiency, especially for high current PSU's. The LM series are linear and are no where near as efficient. Both have pros and cons and their place in electronics power supplies.
Great quality content.
Great video, great explaination! Keep on going!
Great!
Convincing video.... commendable explaination.
Awesome
Nice! Now you basically build a Class-D power amplifier ;)
Next stop: dead-time adjustment. Google hint; look for XOR deadtime circuit.
There is actually one where you only need 1 resistor to select the dead-time value!
Hello Piet! I don't think I will go into that much detail; at least not at the moment. I am aware you can build a complicated and accurate circuit with logic gates, but at some point its much easier to just get a dedicated driver that has everything built in. I was looking at some IC's from Infineon like 2ED2304S06 which has 2 inputs (for low and high side) but they also have IC's with a single input and an external resistor for the deadtime setting.
@@FesZElectronics True, I was thinking about it since, you were doing everything discrete already. (buck converters can also bought of the shelf).
There are a lot of IC's that can control dead-time and even more. Really like the IRS20957S and the IRS2092S. Both are for audio, but can also be used to other things (obviously). LM5116 is also nice, as well as some other parts (can't remember the name) with actually adaptive dead-time control. When a u-controller is used I would rather control the dead-time in the code.
Very interesting!
On some comments below suggesting to focus on content and less face time, I prefer some human content in technical videos. May I suggest a small continuous window in the corner for the presenter. I note that other channels use this technique very effectively.
I did resort to that in the mean time - this video is actually a few months old. If you have time maybe you can look at a more recent video and let me know what you think.
last diode was great !
Great explanation. Will be something about floating gate driver of high side N-channel mosfet in forward 2T topology?
please make a vid about Continuous Conduction Mode and Discontinuous conduction mode
Why do you not upload your ltspice files? Your circuit designs could be helpful!
I did plan to add the simulation but I guess I forgot. I added it to the description now.
Hi FesZ, I getting an error at running ltSpice Sim "can't find definition of model 'ntd18n06l'?
I found a fix for that , need to change the prefix to X (CTRL+ right-mouse-click) , thanks to Helmut.
Hello @@haimt, I'm seeing the same error message but I'm not sure what you mean by changing the prefix to X. Where is that? Please clarify. I'm new to LTSpice and just beginning to learn it thru these YOuTube videos and thanks to FesZ, I'm coming along!
Hello; the problem here is that every component needs a prefix so the simulator know what type of component it is (for diodes its D, for capacitor C etc.) For non-standard components it is usually "X". So when you add a new component, if there is no prefix added, it will create problems. As @haimt1464 said, just right click on the component - to open the component attribute editor and on the first line (prefix) add "X".
For some reason I had to hold down both CTRL and ALT button and then right-click the mouse button to access the component Editor. Thank you for your help and these highly educational videos. I'm now learning LTSpice. it's a great Tool!!
Hi guys,i have buid a syncronous buck and i use ir2110 in half bridg configuration to drive the fets.All seems to be good but,my bootstrap cap is getting hot .Can anyone have this problem before?
I think, it would much better for understanding, to keep camera on the schematics when you explain them instead distraction attention by changing camera focusing from schematic to your face and vice versa frequently.
Such a great video, it will be nice if you make a video about H-Bridge High voltage Inverter Protection, I'm currently working on something like that even with having short circuit protection at the Boost side the mosfet's at the H-Bridge burn while testing short circuit I used SG3525 IC for the job
You might have issues with the voltage rating of the transistors - they should be able to handle not just the DC input but also any spikes that appear in the switching nodes (up to 10-20% of the DC). I'm not sure if and when I will cover H-bridge inverters though..
Hope this helps! Cheers!
@@FesZElectronics Thanks i fixed that, the problem was with low side mosfet's gate resistor which was too high & i also added pull down resistor and now it works just fine thanks for ur nice videos and help
Your videos are very very helpful. Nobody can explain the way u do. I want to learn ltspice, can you guide where i can get good material to start with.
Hello Harish! I'm happy you enjoyed my videos! Well I do have quite an extensive series related to LTspice starting with ( ruclips.net/video/JRcyHuyb1V0/видео.html )
Why voltage drop of the diode is negative w.r.t ground....
Its because of the way in which current is flowing - from ground towards the switching node; for the diode to be forward biased, the anode (connected to gnd) needs to be at a higher voltage potential than the cathode (the switching node); The current is flowing from the inductor trough the output capacitor/load and closing back trough the diode.
GREAT...., how to generate that signal gap (all transistor off), and how long time ideally?
That is usually built into the driver; you have dedicated SMPS converter IC's where this dead time is fixed (it has some tolerance but its not user settable); but you also have some IC which only perform the gate drive function where this time delay can be set;
Regarding the delay length, ideally you want it as short as possible, but in practice, you will need to set it to account for transistor switching delays and also keep a bit of tolerance margin. Usually it will in the tens or hundreds of nseconds but based on specific design it could be higher or lower than this.
@@FesZElectronics great, thankyou
From where did you take your reference when you measured -0.4V (Ge) and -0.7V (Si) due to diode drop?
👍
Thankyou for great explanation. But when we increase the source voltage to 30 then Vd becomes 30 and gate voltage be nearly 11 as fixed in that case.. So how can I get 30 voltage at source side during the turn on time of transistor? According to my thinking, the voltsges in that case at source becomes Vs=Vg-Vth which roughly bevomes 9.
Need clarification from your side. Thankyou
I'm not sure I understand the question. If you are referring to the high side transistor, supplying the gate is done by the bootstrap capacitor (C8). So the gate-source voltage Vgs is not coming directly from the input supply, but rather this floating capacitor. The voltage on the upper side of C8 will always be ~ 5V more (5V because of the way the V_stab voltage is obtained) than the voltage in the source of the high side transistor - so at 30V Vin, C8 is charged by 5V and when the highside is on, then the voltage on C8 reaches 30V+5V so enough to drive the high side.
@@FesZElectronics Thankyou for your response. My query got cleared. It's Vgs not Vg which should not excees 20V.
I have downloaded the simulation and corrected the "X" in the component type as described below. However I am getting a problem where the mfets are shifted to the right and are not connected into the circuit. I just started using the program so don't know how to edit/fix this yet. I have downloaded version 17, and it seems that file is for version 4.
Hello yancy! The schematic was done in LT17 but that doesn't really matter. the point is depending on what symbol you are using for the transistors you might need to re-connect the traces, just make sure the terminals are connected and liked (between model and symbol) correctly.
FesZ Electronics Thanks for replying to my comment. I did manage to get the simulator working as I discovered the "move" tool. When I moved the fets between the wires, everything arranged itself properly. It was very good as I was really having a tough time understanding the boost voltage circuit. However after seeing it work in the sim, I am beginning to "get it".
You changed the music!
Yeah, I tried to upgrade the intro also. What do you think?
FesZ Electronics it is definitely more energetic.
Make video on synchronous boost converter.