Episode 711 Back to basics: Power Supply AC to DC conversion, bridge rectifier Product: www.banggood.com/EQKIT-Consta... Manual: www.icstation.com/product_docu... Be a Patron: / imsaiguy
I love your videos because you explain the purpose of the chosen components and why they are there. Over time I see the same patterns in lots of circuits and I can easily spot a voltage divider and rectifier bridge.
I have the "fake" version of this board. I'll watch your whole series... I have plenty of 5v and 12v fans, and fairly larger heat syncs salvaged from CPUs etc. I just have to source a nice AC transformer for this... As I have learned, you can not use a DC transformer. I'm sure you mention this somewhere along the way... :) Dabbling in electronics, and building kits like this is great. I wish I would have started sooner.
Even at 12v, the power dissipation in the fan regulator will become problematic. A typical 12v fan will draw over 100ma which is over 2 watts, definitely will need to add a heat sink. I used a small switching buck converter module to run a 12v fan.
Thanks for this interesting video. I am puzzled about the bleeding resistor. Is the capacitor not discharging over the diode bridge and the secondary coil of the transformer? I got triggered by the fact that the bleeding R is just generating heat during normal operation and seems to make the power supply less efficient.
Don't forget that the resistor across the filter capacitor also serves to discharge the capacitor for safety reasons. Although not really dangerous at this voltage, a capacitor can hold a high voltage for a long time unless discharged. My linear amplifier takes almost 60 seconds to discharge the capacitors from 3700V to below 100V. de WG2E.
Depending on the ratio of the input Transformer the output of the bridge will vary assuming it's a one-to-one Transformer an isolation Transformer in other words. Your filter capacitor is woefully under value voltage wise if it is a step down Transformer then perhaps a 50 volt bulk capacitor would be sufficient the chances are that's a one-to-one Transformer when you're going to have 170 volts Peak on the output note of that bridge
Wow, I've seen some poorly drawn schematics before, but this one is one of the worst !!! Thank you for translating it using the proper symbols, properly !!!
Can you do a video explaining how to use balance resistors. As often seen across diodes that are in series. . The series diodes are often used to increase the voltage handling capacity. And they use resistors across those diodes to "balance" the voltages. I would love to hear how this works and how to calculate them...
@@IMSAIGuy n example is across the series pass transistors in a linear power supply. they will have resistors across them to "balance" the voltages. I cant seem to grasp that either..
@@clems6989 I haven't seen that because I haven't worked at high voltages but my simple guess would be that you want the value of the resistors to be low enough that more current flows through them than the reverse leakage of the diodes at the maximum reverse voltage and at elevated temperature. What matters most is the variation between diodes. A simple example if you have a diode with a 1uA reverse leakage at 10v you would have to use a resistor of less than 10M so 1uA flows through the resistor, more current would distribute voltages more evenly at the cost of power dissipation in the resistors and making the entire thing act like a leakier diode.
Theoritically speaking if you really wanted to use the 7805 could you hook up a 7812 in front of it you know conect the 12 output to the 7805s input would that do anything usefull would it even be a good idea?
I have seen that done before to help with noise. dropping the voltage still requires dissipation of the power and so the heat sink will be the same size (or two), so while not a bad idea it is costly
@@IMSAIGuy so as i tought, my original train of thought would be to divide the heat between two heatsinks thus requiring two smaller ones tho as you said that probably wouldn't really save money nor space.
Be careful if you change the 7824 regulator. There can be more than 30 V on the input: Most 7812 can not handle 30V on the input, it is 24V i think, With a 24 V transformer you get about 34V.
@@IMSAIGuy The first page of the datasheet do say so. If you look a little deeper, you find that its not testet above 27 volt (according to datasheet from Fairchild Semiconductor). I am no pro and I am probably wrong but i wonder why they don't test at a higher voltage.
@@flemmingchristiansen2462 lots of games in datasheets (I've written them). There are numbers by design and numbers by test. Absolute numbers are by design and tested in design, but not on a part to part test. also many numbers are valid at 25C case temp but not at higher temps, so if it doesn't have a perfect heat sink on it then all bets are off. 35V would not make be nervous for home use.
[QUESTION] in one of my PC switching power supplies, the fuse literally EXPLODES. Its replacement does only lead to another magnificent explosion, reducing the fuse to dust... Obviously there's a short....Ok, now, what did cause this short ? Trying to find out why, I found and/or saw 6 elements that are broken : 1) a thermistor (SKC-2R56, diameter 13mm, R=2.5 Ohms, 6A), just after the fuse, is broken (the upper part) 2) the bridge rectifier (GBU606 mounted on a heatsink) : it's dead (a test, unsoldered from the PCB, showed only 2 diodes of the bridge are ok) 3) a N-channel MOSFET (MDP13N50, mounted on the same heatsink) seems dead as well (tested off the PCB) 4) an IC has partially exploded (I can see its guts partially), and I can only see it's 2 first ID digits : "A6...". After a research, I found that it's most probably an IC of the STR-A6000 series (Off-Line PWM Controllers with Integrated Power MOSFET, the sixth pin of the normally 8 pins is absent) 5) after a test of the surface components below the PCB, I found one diode that is dead. 6) the big filter capacitor (400V, 180µF ) is dead as well : test gave me 5.5nF I think it's not a good idea to repair it, too many broken components, but ask myself, what did cause that short-circuit in the first place ? I suspect the bridge rectifier is the culprit of that disaster, any similar experience ?
I disagree. CAD let’s me draw great schematics. As I add or alter subsections I can move whole blocks of the circuit around to make my intent more clear instead of jamming it in whatever space I have left. I always try to make voltage dividers looks like voltage dividers or inverting op amps with gain resistors look like their canonical textbook form so someone can glance at my schematic and immediately recognize the intent I had behind those resistors. The original schematic hear is either pure laziness, or someone who has no idea what any of the symbols mean copying a schematic from somewhere else into a new publication.
Nice but please, for God's sake, stop hiding everything with your hands all the time ! You always hide the only things we can see ! Use some kind of tip, it would be much nicer for your subscribers....
Наука
I love your videos because you explain the purpose of the chosen components and why they are there. Over time I see the same patterns in lots of circuits and I can easily spot a voltage divider and rectifier bridge.
Sir, I am an electronics engineering student and I must say your videos helped me out a lot. Thank you.
Excellent teaching. Even someone who has no idea about electronics can learn something from IMSAI Guy.
much better then my electronics 101 class. i am rusty on the math, and appreciate the breakdown!
I really enjoy (and appreciate) the way you cover every detail in your videos. Thank you and Keep up the great work!
BTW: I love these types of videos, you do a great job..
I love these types of videos
This is my 1st tutorial of yours I am watching and it is so helpful as I am a beginner electronics learner. Thank you! 🙏🏼
Welcome!
Thanks for the video.
Muchas gracias, amigo...
I built the original smart kit psu back in 1995, and never found any analysis of how it works. Thank you!
Thank you ❤️
Good job👍!!!
Good video
Thanks
I have the "fake" version of this board.
I'll watch your whole series...
I have plenty of 5v and 12v fans, and fairly larger heat syncs salvaged from CPUs etc.
I just have to source a nice AC transformer for this...
As I have learned, you can not use a DC transformer.
I'm sure you mention this somewhere along the way...
:)
Dabbling in electronics, and building kits like this is great.
I wish I would have started sooner.
thanks
Even at 12v, the power dissipation in the fan regulator will become problematic. A typical 12v fan will draw over 100ma which is over 2 watts, definitely will need to add a heat sink. I used a small switching buck converter module to run a 12v fan.
Thanks for this interesting video. I am puzzled about the bleeding resistor. Is the capacitor not discharging over the diode bridge and the secondary coil of the transformer? I got triggered by the fact that the bleeding R is just generating heat during normal operation and seems to make the power supply less efficient.
Nice 👍👍
The bridge diodes at 3A DC out will see about 4.5A RMS, with top-wave spikes above 20A; not just 3A as stated.
Nice
How is the value of the capacitor after the rectification is calculated
Don't forget that the resistor across the filter capacitor also serves to discharge the capacitor for safety reasons. Although not really dangerous at this voltage, a capacitor can hold a high voltage for a long time unless discharged. My linear amplifier takes almost 60 seconds to discharge the capacitors from 3700V to below 100V. de WG2E.
He covers that like 4 minutes in
@@ohmslaw6856 😂
Depending on the ratio of the input Transformer the output of the bridge will vary assuming it's a one-to-one Transformer an isolation Transformer in other words. Your filter capacitor is woefully under value voltage wise if it is a step down Transformer then perhaps a 50 volt bulk capacitor would be sufficient the chances are that's a one-to-one Transformer when you're going to have 170 volts Peak on the output note of that bridge
Wow, I've seen some poorly drawn schematics before, but this one is one of the worst !!! Thank you for translating it using the proper symbols, properly !!!
I've got this power supply...haven't got this schematic though
Can you do a video explaining how to use balance resistors. As often seen across diodes that are in series. . The series diodes are often used to increase the voltage handling capacity. And they use resistors across those diodes to "balance" the voltages. I would love to hear how this works and how to calculate them...
haven't see that and don't know anything
@@IMSAIGuy n example is across the series pass transistors in a linear power supply. they will have resistors across them to "balance" the voltages. I cant seem to grasp that either..
@@clems6989 I haven't seen that because I haven't worked at high voltages but my simple guess would be that you want the value of the resistors to be low enough that more current flows through them than the reverse leakage of the diodes at the maximum reverse voltage and at elevated temperature. What matters most is the variation between diodes.
A simple example if you have a diode with a 1uA reverse leakage at 10v you would have to use a resistor of less than 10M so 1uA flows through the resistor, more current would distribute voltages more evenly at the cost of power dissipation in the resistors and making the entire thing act like a leakier diode.
Are you going to recommend a suitable transformer and fan to buy? It seems from the comments that a 24v fan would be best.
No sorry. It will also need a very large heat sink, not the one I show. I'll cover that in a later video.
Hi, what again is the purpose of the 3300 uF capacitor? I get how the bleeder resistor works due to the existence of the capacitor.
www.nutsvolts.com/magazine/article/basic_analog_power_supply_design
Theoritically speaking if you really wanted to use the 7805 could you hook up a 7812 in front of it you know conect the 12 output to the 7805s input would that do anything usefull would it even be a good idea?
I have seen that done before to help with noise. dropping the voltage still requires dissipation of the power and so the heat sink will be the same size (or two), so while not a bad idea it is costly
@@IMSAIGuy so as i tought, my original train of thought would be to divide the heat between two heatsinks thus requiring two smaller ones tho as you said that probably wouldn't really save money nor space.
Be careful if you change the 7824 regulator. There can be more than 30 V on the input: Most 7812 can not handle 30V on the input, it is 24V i think, With a 24 V transformer you get about 34V.
the LM78xx is abs max at 35V up to 18V and 40V above that. You are right though. getting very close.
@@IMSAIGuy The first page of the datasheet do say so. If you look a little deeper, you find that its not testet above 27 volt (according to datasheet from Fairchild Semiconductor).
I am no pro and I am probably wrong but i wonder why they don't test at a higher voltage.
@@flemmingchristiansen2462 lots of games in datasheets (I've written them). There are numbers by design and numbers by test. Absolute numbers are by design and tested in design, but not on a part to part test. also many numbers are valid at 25C case temp but not at higher temps, so if it doesn't have a perfect heat sink on it then all bets are off. 35V would not make be nervous for home use.
Am I correct in assuming that you'll have to short the output to set the current limit?
yes, that is true with most supplies
[QUESTION] in one of my PC switching power supplies, the fuse literally EXPLODES. Its replacement does only lead to another magnificent explosion, reducing the fuse to dust...
Obviously there's a short....Ok, now, what did cause this short ?
Trying to find out why, I found and/or saw 6 elements that are broken :
1) a thermistor (SKC-2R56, diameter 13mm, R=2.5 Ohms, 6A), just after the fuse, is broken (the upper part)
2) the bridge rectifier (GBU606 mounted on a heatsink) : it's dead (a test, unsoldered from the PCB, showed only 2 diodes of the bridge are ok)
3) a N-channel MOSFET (MDP13N50, mounted on the same heatsink) seems dead as well (tested off the PCB)
4) an IC has partially exploded (I can see its guts partially), and I can only see it's 2 first ID digits : "A6...". After a research, I found that it's most probably an IC of the STR-A6000 series (Off-Line PWM Controllers with Integrated Power MOSFET, the sixth pin of the normally 8 pins is absent)
5) after a test of the surface components below the PCB, I found one diode that is dead.
6) the big filter capacitor (400V, 180µF ) is dead as well : test gave me 5.5nF
I think it's not a good idea to repair it, too many broken components, but ask myself, what did cause that short-circuit in the first place ?
I suspect the bridge rectifier is the culprit of that disaster, any similar experience ?
I disagree. CAD let’s me draw great schematics. As I add or alter subsections I can move whole blocks of the circuit around to make my intent more clear instead of jamming it in whatever space I have left. I always try to make voltage dividers looks like voltage dividers or inverting op amps with gain resistors look like their canonical textbook form so someone can glance at my schematic and immediately recognize the intent I had behind those resistors. The original schematic hear is either pure laziness, or someone who has no idea what any of the symbols mean copying a schematic from somewhere else into a new publication.
Sad that we are loosing the basics of simplicity
Nice but please, for God's sake, stop hiding everything with your hands all the time ! You always hide the only things we can see ! Use some kind of tip, it would be much nicer for your subscribers....
Stop moving your hands all around it's very distracting. Thank you.............