Great video! I have one caution, as yaghiyah said below, " The Rsense resistor used needs to be, for best performance, a non-inductive type, " coiling the resistance wire is not best practice, even thought at 1kHz it may have little effect. I might have bundle 10 of those 1Ω resistors in parallel if that's all I had.
This is incredible content Robert, I'm going to pass this onto my network, I'm sure we will all learn so much from you. Thanks so much for taking the time to do this, it's an important mission. - Glenn
Thanks a lot Robert. Just what I was looking for, this method is great. I was looking for a way to determine the minimum safe operating frequency for a transformer. Just have one transformer and thoughts about saturation was haunting me. I just used an IGBT in the need for a MOSFET of that capacity and worked like a charm. Good to know there are still real engineers left.
The inductance of a transformer primary with an open secondary is just the magnetization current of the transformer primary which is an inductor. So essentially you have a series RL circuit, assuming the MOSFET Rds on can be ignored. Because we have a series RL circuit, the current through the transformer primary and the sense resistor is the same, so we can depict the current in the circuit by looking at the voltage across the sense resistor. What this means is essentially we are looking at the classic LR series circuit current characteristic which should be exponential, and is as shown on the scope. Then as the transformer primary becomes saturated, or, in other words, is unable to store more magnetic flux , the primary reverts to just looking like a resistor as you extend the pulse time. So, after saturation is reached the RL current characteristic becomes linear (a straight line) like in a resistor. Saturation then is determined by looking at the knee (inflexion point) between the exponential and linear current. The inductive kick looks a little large which is due to unclamped inductance, either in the transformer primary leads or the inductive sense resistor or both.
1:28 I would highly stress the point that.. The Rsense resistor used needs to be for best performance a non-inductive type, probably a carbon resistor 3W/5W and maybe a snubber across of 22p or so to tame spikes. That resistor too also acts a magnetic core reset.
How does voltage fit into this? Is saturation really just the consequence of cumulated current or is it really about the duration of applied _wattage_ ? E.g.: is there a difference how fast the same core saturates whether I have a high voltage and low current on the primary coil versus low voltage with high current (the product being equal)? Second question: what about timing, i.e. what are typical real-life values how fast saturation occurs in gapped vs. non-gapped cores? You mentioned 20µs here, but I don't know what kind of core you're using. For a flyback I guess this would be a gapped core, but how would - at the other end of the spectrum - e.g. non-gapped toroids compare? Are we talking nanoseconds? The reason why I'm asking this is that I wonder whether we can interrupt the circuit shortly before saturation by using a MOSFET that makes use of the voltage across the coil (which should greatly decrease near saturation) as gate voltage. Thanks!
Hi Robert, Thank you for this video. Question, in the title you mention that this is a transformer and so I am assuming that this is a ferrite material with no air gap. However, in your video you mention that this is an inductor. Is this an inductor with an air gap or a transformer?
Hi Reni, If I recall it had a gap. Adding a gap will make the inductor go into saturation at a higher voltage. If you look at a B vs H curve, the make the slope of the curve less so the saturation happens at a higher H and therefore higher current. I hope that makes sense. Have a good day. RB
The wire wound resistor is adding a decent amount of inductance. From what I just read that would only be problematic at RF. This is a very interesting lesson that has me kind of puzzled. My thinking would be current falling off with core saturation, not curving up like it does. Great lesson sir.
fla playa Hi fla, the current goes up because the core is saturated and thus the inductance start to drop exponentially. It some point it stops being an inductor and start to act like a wire. Hope this makes sense.
Ah ha I think I just put it all together with your help. Pri current rises as the indunctor of transformer becomes saturated with flux for some reason I was thinking Sec.. I'm going to have to do a lot more research on this. Sec Voltages/currents would be so cool to see as this phenomenon occurs btw. Thank you for the reply Robert, you are highly intelligent.
I did just totally get it. Saturate/remove good efficiency of core material the inductance/reluctance fall hence current rise... Love the "Wire" analogy... Greatly appreciated!
to really drive this point home - "inductance drops once all the electron spins in the core are aligned (aka once the core is saturated)", have a look at magnetic amplifiers. [NOTE: alignment of electron spins makes the magnetic field strong enough to be detected as a magnetic field around the core; aka "aligned magnetic domains. A saturated core has reached the strongest possible magnetization, ie. all its electron spins are aligned] The magnetic amplifier actually works BECAUSE OF core saturation as follows: 1) normally, an inductor impedes current flow - it takes about five (5) L/R time constants to reach full current flow in an inductor (this is why "current lags voltage" in the inductor). "L" = the inductance value of the inductor. "R" = the resistance. You can look up "L/R time constant" for more info. 2) so you can think of the core as telling the current "look - I'm not just flipping all my electron spins into alignment just because you want to freely flow in the wire - it's gonna take a bit of time to align all my electron spins - and that means I'm impeding your flow through that wire, so hang tight" 3) after all electron spins in the core (the magnetic moment of each electron) are in full alignment, the core can no longer exert a 'damping' effect on the current flow 4) so the "inductance" is just the name for the core material exerting a brief thwarting effect, an interference, over the current freely flowing in the wire A magnetic amplifier works like this: 1) there is a core material 2) a coil is wrapped around it to host AC current flow, just like in a typical inductor or transformer 3) a second wire is also wrapped around or in the vicinity of the core but it hosts DC current 4) the AC is turned on and is THWARTED by the electron spin alignment process described above (thwarted by the inductance) 5) the DC is turned on and PRE-ALIGNS all the electron spins in the core, aka saturates the core 6) that drops the inductance because the core's electron spins are all aligned and can no longer impede current flow 7) suddenly the AC is no longer thwarted by the electron spin alignment process and is AMPLIFIED, increases in magnitude The DC current pre-aligns, aka pre-saturates, the core in order to suddenly remove the thwarting normally seen by the AC Magnetic amplifiers don't amplify a magnetic field, they amplify AC by pre-saturating the core with a DC-powered winding, and thus by removing inductance. But they do that by magnetizing the core to the max (aka saturation) hence the name 'magnetic amplifier' They are also called "Saturable Reactors" .
Hi KX36. I did not have a very accurate current sense resistor. I tried to make a current sense using magnet wire but I could not verify such a low ohms with the test equipment I have. So I relied on the Di=(V/L)*Ton to try to calculate the delta current. I have a LCR that is accurate and I rely on the LCR measurement to get accurate enough calculation. I just wanted to be in the ball park; it does not have to be that much accurate. To measure more accurate, use a precision current sense resistor. Thanks for the question! Robert
@@RobertBolanos I don't mean to measure the current accurately directly from the resistor, I mean the burden voltage of the shunt makes up a significant portion of the Vdd you used in the formula since the current is apparently so high. If this is the case, the shunt resistor is acting as a ballast to act against saturation. You can still calculate current with that formula and forego the need for an accurate shunt resistance, but you need to measure the voltage across the inductor only, not just use the supply voltage.
Thank you for the precision, but now I'm a bit confused between Bsat and Bmax. Aren't they the same, where it represent the maximum saturation level that core can handle?
Very important video Thank you.Could you make a video about about "how find SMPS transformer's most effective working frequency "? Actually could you make a video about "What can learn noname a toroide core with scope and function generator? I mean we can learn Al value and how can learn permability and other catalogue specifications? (Maybe you can make quality torode core all specs.know and you could find same specifications ?)
Hi Brokenwords. That is a good question. Isat and Bsat are related. You ussaully use Bsat= (VL*Ton)/(Ae*N). Bsat is given in the core material datasheet. Bsat typically for a ferrite is around 0.3T (teslas). So you can use the same equation to calculate the Nturn and use Boperate = (VL*Ton)/(Ae*N). Per the Abraham Pressmen book, he likes to use 0.16T for Bopeate. That is about 53% of Bsat. You can probably go up to 80% of Bsat. How does that relate to Isat? (Ioperate/Isat)=Boperate/Bmax). This is how they are related. Therefore, Ioperate would be 53% to 80% of Isat. I hope that make sense. Great question and thank you for asking it . Robert Bolanos
@@RobertBolanos I think A. Pressmen used 0.16T for Boperate is because he derated it for 120C operating temperature. To be more precise Bsat = 0.28T for ferrite... that about 57% which seem right... 60%-70% of the room temp capacity seems to be the target number for high temp.
Jesse Lee thank you for your response. I think Dr. goes as high as 80 % of Bsat. I think that is where I got the 80 percent. Also keep in mind that bsat is also specified at two temperature which one of them is at 100c. So if you go 80 percent at the Bsat specified at 100c you should be Ok. The only time you would go 80 percent is if you are trying to fit many turns into a core. If that is the case, might as well use a bigger core. Again thank you for your inputs. Robert
Your explanations are really great Robert, I would love to see more circuits and explanations on the oscilloscope. Thanks
Man I am really loving your explanations. Thanks for all of the content
Thanks. This was the most practical demonstration I've seen. I need to build this rig and try it myself. Much appreciated.
-Jake
Thank you
Great video! I have one caution, as yaghiyah said below, " The Rsense resistor used needs to be, for best performance, a non-inductive type, " coiling the resistance wire is not best practice, even thought at 1kHz it may have little effect. I might have bundle 10 of those 1Ω resistors in parallel if that's all I had.
@@mavamQ You are correct. Non inductive are better. Thanks for comment! Merry Christmas to you!
This is incredible content Robert, I'm going to pass this onto my network, I'm sure we will all learn so much from you. Thanks so much for taking the time to do this, it's an important mission.
- Glenn
Thank you Glen!
@@RobertBolanos You're very welcome!
Thanks a lot Robert. Just what I was looking for, this method is great. I was looking for a way to determine the minimum safe operating frequency for a transformer. Just have one transformer and thoughts about saturation was haunting me. I just used an IGBT in the need for a MOSFET of that capacity and worked like a charm. Good to know there are still real engineers left.
Thanks for the kinds words. You made my day.
Vsat would be i_sat X Zprimary , it can also be measured with a differential probe across primary .
The inductance of a transformer primary with an open secondary is just the magnetization current of the transformer primary which is an inductor. So essentially you have a series RL circuit, assuming the MOSFET Rds on can be ignored. Because we have a series RL circuit, the current through the transformer primary and the sense resistor is the same, so we can depict the current in the circuit by looking at the voltage across the sense resistor.
What this means is essentially we are looking at the classic LR series circuit current characteristic which should be exponential, and is as shown on the scope. Then as the transformer primary becomes saturated, or, in other words, is unable to store more magnetic flux , the primary reverts to just looking like a resistor as you extend the pulse time. So, after saturation is reached the RL current characteristic becomes linear (a straight line) like in a resistor. Saturation then is determined by looking at the knee (inflexion point) between the exponential and linear current.
The inductive kick looks a little large which is due to unclamped inductance, either in the transformer primary leads or the inductive sense resistor or both.
1:28 I would highly stress the point that.. The Rsense resistor used needs to be for best performance a non-inductive type, probably a carbon resistor 3W/5W and maybe a snubber across of 22p or so to tame spikes. That resistor too also acts a magnetic core reset.
It was as good as ever. Thanks for useful tutorials videos. 🙏🙏🙏
Your channel is the best!
Love your explanations. Great educational video. Thanks.
Thank you Gazza-usa
How does voltage fit into this? Is saturation really just the consequence of cumulated current or is it really about the duration of applied _wattage_ ? E.g.: is there a difference how fast the same core saturates whether I have a high voltage and low current on the primary coil versus low voltage with high current (the product being equal)?
Second question: what about timing, i.e. what are typical real-life values how fast saturation occurs in gapped vs. non-gapped cores? You mentioned 20µs here, but I don't know what kind of core you're using. For a flyback I guess this would be a gapped core, but how would - at the other end of the spectrum - e.g. non-gapped toroids compare? Are we talking nanoseconds? The reason why I'm asking this is that I wonder whether we can interrupt the circuit shortly before saturation by using a MOSFET that makes use of the voltage across the coil (which should greatly decrease near saturation) as gate voltage.
Thanks!
Actually love to get this answer answered
Hi Robert, Thank you for this video. Question, in the title you mention that this is a transformer and so I am assuming that this is a ferrite material with no air gap. However, in your video you mention that this is an inductor. Is this an inductor with an air gap or a transformer?
Hi Reni, If I recall it had a gap. Adding a gap will make the inductor go into saturation at a higher voltage. If you look at a B vs H curve, the make the slope of the curve less so the saturation happens at a higher H and therefore higher current. I hope that makes sense. Have a good day. RB
@@RobertBolanos Thank you, Is your Transformer spreadsheet available to dowload?
@@enochreturn I have the flyback spreadsheet that I can sent if you give me your email.
The wire wound resistor is adding a decent amount of inductance. From what I just read that would only be problematic at RF. This is a very interesting lesson that has me kind of puzzled. My thinking would be current falling off with core saturation, not curving up like it does. Great lesson sir.
fla playa Hi fla, the current goes up because the core is saturated and thus the inductance start to drop exponentially. It some point it stops being an inductor and start to act like a wire. Hope this makes sense.
Ah ha I think I just put it all together with your help. Pri current rises as the indunctor of transformer becomes saturated with flux for some reason I was thinking Sec.. I'm going to have to do a lot more research on this. Sec Voltages/currents would be so cool to see as this phenomenon occurs btw. Thank you for the reply Robert, you are highly intelligent.
I did just totally get it. Saturate/remove good efficiency of core material the inductance/reluctance fall hence current rise... Love the "Wire" analogy... Greatly appreciated!
@@flaplaya I am glad my explanation helped.
to really drive this point home - "inductance drops once all the electron spins in the core are aligned (aka once the core is saturated)", have a look at magnetic amplifiers. [NOTE: alignment of electron spins makes the magnetic field strong enough to be detected as a magnetic field around the core; aka "aligned magnetic domains. A saturated core has reached the strongest possible magnetization, ie. all its electron spins are aligned]
The magnetic amplifier actually works BECAUSE OF core saturation as follows:
1) normally, an inductor impedes current flow - it takes about five (5) L/R time constants to reach full current flow in an inductor (this is why "current lags voltage" in the inductor). "L" = the inductance value of the inductor. "R" = the resistance. You can look up "L/R time constant" for more info.
2) so you can think of the core as telling the current "look - I'm not just flipping all my electron spins into alignment just because you want to freely flow in the wire - it's gonna take a bit of time to align all my electron spins - and that means I'm impeding your flow through that wire, so hang tight"
3) after all electron spins in the core (the magnetic moment of each electron) are in full alignment, the core can no longer exert a 'damping' effect on the current flow
4) so the "inductance" is just the name for the core material exerting a brief thwarting effect, an interference, over the current freely flowing in the wire
A magnetic amplifier works like this:
1) there is a core material
2) a coil is wrapped around it to host AC current flow, just like in a typical inductor or transformer
3) a second wire is also wrapped around or in the vicinity of the core but it hosts DC current
4) the AC is turned on and is THWARTED by the electron spin alignment process described above (thwarted by the inductance)
5) the DC is turned on and PRE-ALIGNS all the electron spins in the core, aka saturates the core
6) that drops the inductance because the core's electron spins are all aligned and can no longer impede current flow
7) suddenly the AC is no longer thwarted by the electron spin alignment process and is AMPLIFIED, increases in magnitude
The DC current pre-aligns, aka pre-saturates, the core in order to suddenly remove the thwarting normally seen by the AC
Magnetic amplifiers don't amplify a magnetic field, they amplify AC by pre-saturating the core with a DC-powered winding, and thus by removing inductance. But they do that by magnetizing the core to the max (aka saturation) hence the name 'magnetic amplifier'
They are also called "Saturable Reactors"
.
great video
When you calculate Isat, you assume no voltage drop across the shunt resistors. I suspect that would be a significant source of error.
Hi KX36. I did not have a very accurate current sense resistor. I tried to make a current sense using magnet wire but I could not verify such a low ohms with the test equipment I have. So I relied on the Di=(V/L)*Ton to try to calculate the delta current. I have a LCR that is accurate and I rely on the LCR measurement to get accurate enough calculation. I just wanted to be in the ball park; it does not have to be that much accurate. To measure more accurate, use a precision current sense resistor. Thanks for the question! Robert
@@RobertBolanos I don't mean to measure the current accurately directly from the resistor, I mean the burden voltage of the shunt makes up a significant portion of the Vdd you used in the formula since the current is apparently so high. If this is the case, the shunt resistor is acting as a ballast to act against saturation.
You can still calculate current with that formula and forego the need for an accurate shunt resistance, but you need to measure the voltage across the inductor only, not just use the supply voltage.
Would that be Bmax (or Bsat)?
Bsat
Would you be able to make a video on how to find bsat of an unknown ferrite core?
I meant Bmax
@@coucouj2781 You can use the same procedure I used here and calculate it
Thank you for the precision, but now I'm a bit confused between Bsat and Bmax. Aren't they the same, where it represent the maximum saturation level that core can handle?
Very important video Thank you.Could you make a video about about "how find SMPS transformer's most effective working frequency "?
Actually could you make a video about "What can learn noname a toroide core with scope and function generator?
I mean we can learn Al value and how can learn permability and other catalogue specifications?
(Maybe you can make quality torode core all specs.know and you could find same specifications ?)
Great suggestion!
What percentage should the safety gap be between the saturation and operating current?
Thanks for video.
Hi Brokenwords. That is a good question. Isat and Bsat are related. You ussaully use Bsat= (VL*Ton)/(Ae*N). Bsat is given in the core material datasheet. Bsat typically for a ferrite is around 0.3T (teslas). So you can use the same equation to calculate the Nturn and use Boperate = (VL*Ton)/(Ae*N). Per the Abraham Pressmen book, he likes to use 0.16T for Bopeate. That is about 53% of Bsat. You can probably go up to 80% of Bsat. How does that relate to Isat? (Ioperate/Isat)=Boperate/Bmax). This is how they are related. Therefore, Ioperate would be 53% to 80% of Isat. I hope that make sense. Great question and thank you for asking it . Robert Bolanos
@@RobertBolanos
I think A. Pressmen used 0.16T for Boperate is because he derated it for 120C operating temperature. To be more precise Bsat = 0.28T for ferrite... that about 57% which seem right... 60%-70% of the room temp capacity seems to be the target number for high temp.
Jesse Lee thank you for your response. I think Dr. goes as high as 80 % of Bsat. I think that is where I got the 80 percent. Also keep in mind that bsat is also specified at two temperature which one of them is at 100c. So if you go 80 percent at the Bsat specified at 100c you should be Ok. The only time you would go 80 percent is if you are trying to fit many turns into a core. If that is the case, might as well use a bigger core. Again thank you for your inputs. Robert
Sir what is “E” in your equation?
E is the voltage. Sorry for the confusion.
Hi sır. Can you create PFC Circuit video tutorial? Your videos very important for me. Thanks
Hi Mesut, I intend to do a couple of videos on that subject. Robert
@@RobertBolanos Waiting Impatiently. Thanks sir
❤❤❤❤ thank you ❤❤❤
Thank you Ahmed!
Please add schematic
👍 analyze
Youre great, but i suggest u go sleep for a while, its seen from your voice tone
yes, it was late at night when I did the video.
Only good