Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

A varying-frequency control Flyback rectifier for constant current LED driving in Boundary-Conduction Mode

Patrick_66

Member level 3
Member level 3
Joined
Nov 18, 2023
Messages
62
Helped
0
Reputation
0
Reaction score
0
Trophy points
6
Activity points
427
Greetings everyone, there are a few questions that I'm hoping that you guys can help me to solve. What does it actually mean from the title that I had written above? Is it varying the input voltage while the output current remain the same at all times ? What is actually the use of a feedback controller is it to maintain a constant current by varying the frequency of the switch or is the controller used to ensure that the whole system is always running in BCM mode? Sorry for my ignorance guys. I hope that someone can help me. Thank you.
 
It appears to be both frequency controlled load current and BCM to minimize switching loss.
From my understanding, I can have a controller to control the current of the load by varying the frequency to change the switch's duty cycle when the input voltage changes maybe from 5 V to 10 V. But how to ensure that the system always operates in BCM mode? Can you enlighten me. Thank you in advance.
 
From my understanding, I can have a controller to control the current of the load by varying the frequency to change the switch's duty cycle when the input voltage changes maybe from 5 V to 10 V. But how to ensure that the system always operates in BCM mode? Can you enlighten me. Thank you in advance.
Where is your reference to explain that?
 
As I understand it from a patent on BCM, BCM is in between DCM & CCM or just on the edge of Continuous Conduction Mode (CCM). This is regardless of topology and modulation method using variable f, d.f. and/or pulse skipping. How the major IC companies do it may not be the same as in the patent. Your YouTube link appears to show a 13% duty cycle but I did not see the current waveform but it was possibly in DCM.

1700340614411.png


1700340571762.png

--- Updated ---

I have no experience on BCM, but it appears to have high ripple current, which may need averaging unless you simply regulate the peak current. Averaging would cause lag and add to overshoot issues.
 
Last edited:
As I understand it from a patent on BCM, BCM is in between DCM & CCM or just on the edge of Continuous Conduction Mode (CCM). This is regardless of topology and modulation method using variable f, d.f. and/or pulse skipping. How the major IC companies do it may not be the same as in the patent. Your YouTube link appears to show a 13% duty cycle but I did not see the current waveform but it was possibly in DCM.

View attachment 186286

View attachment 186285
--- Updated ---

I have no experience on BCM, but it appears to have high ripple current, which may need averaging unless you simply regulate the peak current. Averaging would cause lag and add to overshoot issues.
Sir, may I ask how the video managed to always maintain in DCM operating mode while achieving constant current output when the input voltage varies? Because based on the equation below, if the voltage and frequency vary mean that as long as I monitor the frequency of the switch I'm able to achieve a system that always maintains in DCM mode? How to do so if I want the system to be in BCM because based on the equation if I were to change the frequency of the MOSFET to regulate a constant current do I have to always change the inductance or resistance value to maintain Imin = 0? Sorry for my ignorance. Hope that you can enlighten me again. Thank you for your time.
 

Attachments

  • WhatsApp Image 2023-11-19 at 9.55.08 AM.jpeg
    WhatsApp Image 2023-11-19 at 9.55.08 AM.jpeg
    168.3 KB · Views: 90
Hi,

the quality of the video is quite high, but still it contains some mistakes. Not big ones.

for example in the power dissipation of the shunt it totally misses the waveform and the fact that it´s pulsed (duty cycle).

Also the "constant current" is by far not perfect.
And I have to agree, if one varies the load (several LEDs in a string) or if one varies the input voltage, then it´s easy to go into CCM.
This will vary the average LED current. So the current is not constant anymore.

Using a breadboard was asking for problems, especially when one tries to switch several 100mA with a 500kHz.
Spikes may easily kill the MOSFET or other devices (within seconds ... or afther months) and it sends out lots of RF dirt (EMI).

I recommend to use dedicated LED driver ICs to make the circuit more simple and more reliable.
At least I recommend to read some design notes / application notes of LED driver ICs to learn about the problems, like switching MOSFETs, regulation, reliability and EMI.

Klaus
 
BCM is very very simple for you...it just means you use the voltage on the bias coil to tell you when the secondary coil has just "discharged". Then u switch on the pri side fet.....let me know if you want a sim of it on free ltspice.
BCM is good for emc...and good as it doesnt get reverse recovery in sec diode.
And bcm good because you get a bit of "valley switching" so less capacitive switching loss.
--- Updated ---

BCM is very very simple for you...it just means you use the voltage on the bias coil to tell you when the secondary coil has just "discharged". Then u switch on the pri side fet.....let me know if you want a sim of it on free ltspice.
BCM is good for emc...and good as it doesnt get reverse recovery in sec diode.
And bcm good because you get a bit of "valley switching" so less capacitive switching loss.
--- Updated ---

BCM is very very simple for you...it just means you use the voltage on the bias coil to tell you when the secondary coil has just "discharged". Then u switch on the pri side fet.....let me know if you want a sim of it on free ltspice.
BCM is good for emc...and good as it doesnt get reverse recovery in sec diode.
And bcm good because you get a bit of "valley switching" so less capacitive switching loss.
--- Updated ---

BCM is very very simple for you...it just means you use the voltage on the bias coil to tell you when the secondary coil has just "discharged". Then u switch on the pri side fet.....let me know if you want a sim of it on free ltspice.
BCM is good for emc...and good as it doesnt get reverse recovery in sec diode.
And bcm good because you get a bit of "valley switching" so less capacitive switching loss.
--- Updated ---

sorry, i dont know why quad post..it was slow wifi on train, maybe that
--- Updated ---

Sorry i dont know why quad..maybe slow wifi on train
 
Last edited:
BCM is very very simple for you...it just means you use the voltage on the bias coil to tell you when the secondary coil has just "discharged". Then u switch on the pri side fet.....let me know if you want a sim of it on free ltspice.
BCM is good for emc...and good as it doesnt get reverse recovery in sec diode.
And bcm good because you get a bit of "valley switching" so less capacitive switching loss.
--- Updated ---

BCM is very very simple for you...it just means you use the voltage on the bias coil to tell you when the secondary coil has just "discharged". Then u switch on the pri side fet.....let me know if you want a sim of it on free ltspice.
BCM is good for emc...and good as it doesnt get reverse recovery in sec diode.
And bcm good because you get a bit of "valley switching" so less capacitive switching loss.
--- Updated ---

BCM is very very simple for you...it just means you use the voltage on the bias coil to tell you when the secondary coil has just "discharged". Then u switch on the pri side fet.....let me know if you want a sim of it on free ltspice.
BCM is good for emc...and good as it doesnt get reverse recovery in sec diode.
And bcm good because you get a bit of "valley switching" so less capacitive switching loss.
--- Updated ---

BCM is very very simple for you...it just means you use the voltage on the bias coil to tell you when the secondary coil has just "discharged". Then u switch on the pri side fet.....let me know if you want a sim of it on free ltspice.
BCM is good for emc...and good as it doesnt get reverse recovery in sec diode.
And bcm good because you get a bit of "valley switching" so less capacitive switching loss.
--- Updated ---

sorry, i dont know why quad post..it was slow wifi on train, maybe that
--- Updated ---

Sorry i dont know why quad..maybe slow wifi on train
It would be helpful if you could share to me your simulation. Thank you in advance.
 
Hi,

the quality of the video is quite high, but still it contains some mistakes. Not big ones.

for example in the power dissipation of the shunt it totally misses the waveform and the fact that it´s pulsed (duty cycle).

Also the "constant current" is by far not perfect.
And I have to agree, if one varies the load (several LEDs in a string) or if one varies the input voltage, then it´s easy to go into CCM.
This will vary the average LED current. So the current is not constant anymore.

Using a breadboard was asking for problems, especially when one tries to switch several 100mA with a 500kHz.
Spikes may easily kill the MOSFET or other devices (within seconds ... or afther months) and it sends out lots of RF dirt (EMI).

I recommend to use dedicated LED driver ICs to make the circuit more simple and more reliable.
At least I recommend to read some design notes / application notes of LED driver ICs to learn about the problems, like switching MOSFETs, regulation, reliability and EMI.

Klaus
Is there an IC that can help maintain in BCM mode and regulate the current at the output? If so, can you share with me where to get all the equations for the flyback converter to determine all the capacitor, and inductor values. Thank you for your time and help.
 
here is BCM Flyback LTspice...you can change it to do current regulation.
Or why not go for an onsemi bcm controller....ncp13xx etc.
Say NCP1340.
I think ti.com does bcm flyback controllers.
What about Innoswitch....all the design is does for you in their app....just shovel in your spec....innoswitch has output current regulation....just add the rsense
 

Attachments

  • Flyback _BCM _LT3799 _Valley Switching _85.8KHz.zip
    2.4 KB · Views: 113
here is BCM Flyback LTspice...you can change it to do current regulation.
Or why not go for an onsemi bcm controller....ncp13xx etc.
Say NCP1340.
I think ti.com does bcm flyback controllers.
What about Innoswitch....all the design is does for you in their app....just shovel in your spec....innoswitch has output current regulation....just add the rsense
Is there any website or other sources that I can learn about the equations for flyback converter with constant current regulation in BCM mode? Can you explain a little about how to maintain in BCM mode even though the input voltage changes that will tense to affect the operation mode of the system? Is it, that the controller will do all the job like maintaining in BCM mode and regulating the output current? If possible I would like to understand how the driver maintains constant current output while ensuring that the system is in BCM mode and all the necessary equations to get the inductance and capacitance values. Sorry for my ignorance but i very want to learn how to do it. By the way, thank you so much for the simulation. Hope to hear from you soon.
 
To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.

It just uses an error amplifier like all other chips, or most others....or you can add in your own error amplifier and interface to the chip...eg TL4321/opto.To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.
--- Updated ---

To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.

It just uses an error amplifier like all other chips, or most others....or you can add in your own error amplifier and interface to the chip...eg TL4321/opto.To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.
--- Updated ---

To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.

It just uses an error amplifier like all other chips, or most others....or you can add in your own error amplifier and interface to the chip...eg TL4321/opto.To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.
--- Updated ---

To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.

It just uses an error amplifier like all other chips, or most others....or you can add in your own error amplifier and interface to the chip...eg TL4321/opto.To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.
 

Attachments

  • Core Mathematics and Equations for SMPS design.zip
    121.8 KB · Views: 109
To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.

It just uses an error amplifier like all other chips, or most others....or you can add in your own error amplifier and interface to the chip...eg TL4321/opto.To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.
--- Updated ---

To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.

It just uses an error amplifier like all other chips, or most others....or you can add in your own error amplifier and interface to the chip...eg TL4321/opto.To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.
--- Updated ---

To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.

It just uses an error amplifier like all other chips, or most others....or you can add in your own error amplifier and interface to the chip...eg TL4321/opto.To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.
--- Updated ---

To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.

It just uses an error amplifier like all other chips, or most others....or you can add in your own error amplifier and interface to the chip...eg TL4321/opto.To design flyback its use of Lenz's Law. Also, (Np/Ns)^2 = Lp/Ls etc etc

It stays in BCM because the comparator trips when the aux coil input tells it that the aux coil has discharged. It changes frequency to stay in BCM....usually in light load they stop being in bcm and just go to plain DCM...otherwise the frequency would get too high.
If it changes frequency to maintain in BCM will it be able to regulate the current at the output at the same time?
 
Hi,

what information are you looking for?

When I see BCM waveforms then it´s a triangle shaped from 0 to a dedicated current I_t, then again triangle shaped down to zero ... repeating.

So the average current is half of the I_t.
So by changing I_t .. it changes I_avg = I_t/2 and it also changes frequency.

Maybe I misunderstood the problem.

Klaus
 
It regulates the current because the peak current is allowed to go higher when demand is higher......power thru'put in flyback is 1/2 x Lpri X ipk^2 x f

Where ipk is peak pri current.

higher i, higher power...its just current mode control......it doesnt matter that it does BCM and current regulation at the same time.....it sounds magic but its not.
Did you run the sim?....you can see it doing it in front of your eyes........its in BCM, the peak current can still go higher....no problem.

Do you know how peak current mode control works?...ill send you sim if you want...showing it in detail.
--- Updated ---

Here is BCM sim of led driver
LTspice
--- Updated ---

It regulates the current because the peak current is allowed to go higher when demand is higher......power thru'put in flyback is 1/2 x Lpri X ipk^2 x f

Where ipk is peak pri current.

higher i, higher power...its just current mode control......it doesnt matter that it does BCM and current regulation at the same time.....it sounds magic but its not.
Did you run the sim?....you can see it doing it in front of your eyes........its in BCM, the peak current can still go higher....no problem.

Do you know how peak current mode control works?...ill send you sim if you want...showing it in detail.
 

Attachments

  • Buck BCM _LT8312 _simple.zip
    1.9 KB · Views: 87
It regulates the current because the peak current is allowed to go higher when demand is higher......power thru'put in flyback is 1/2 x Lpri X ipk^2 x f

Where ipk is peak pri current.

higher i, higher power...its just current mode control......it doesnt matter that it does BCM and current regulation at the same time.....it sounds magic but its not.
Did you run the sim?....you can see it doing it in front of your eyes........its in BCM, the peak current can still go higher....no problem.

Do you know how peak current mode control works?...ill send you sim if you want...showing it in detail.
--- Updated ---

Here is BCM sim of led driver
LTspice
--- Updated ---

It regulates the current because the peak current is allowed to go higher when demand is higher......power thru'put in flyback is 1/2 x Lpri X ipk^2 x f

Where ipk is peak pri current.

higher i, higher power...its just current mode control......it doesnt matter that it does BCM and current regulation at the same time.....it sounds magic but its not.
Did you run the sim?....you can see it doing it in front of your eyes........its in BCM, the peak current can still go higher....no problem.

Do you know how peak current mode control works?...ill send you sim if you want...showing it in detail.
I would like to ask if we normally use a microcontroller for the feedback system. Do we have to program the microcontroller to specify it to BCM and in constant current monitoring using software or connect resistors or capacitors to do that? Sorry for my late reply.
 
It regulates the current because the peak current is allowed to go higher when demand is higher......power thru'put in flyback is 1/2 x Lpri X ipk^2 x f

Where ipk is peak pri current.

higher i, higher power...its just current mode control......it doesnt matter that it does BCM and current regulation at the same time.....it sounds magic but its not.
Did you run the sim?....you can see it doing it in front of your eyes........its in BCM, the peak current can still go higher....no problem.

Do you know how peak current mode control works?...ill send you sim if you want...showing it in detail.
--- Updated ---

Here is BCM sim of led driver
LTspice
--- Updated ---

It regulates the current because the peak current is allowed to go higher when demand is higher......power thru'put in flyback is 1/2 x Lpri X ipk^2 x f

Where ipk is peak pri current.

higher i, higher power...its just current mode control......it doesnt matter that it does BCM and current regulation at the same time.....it sounds magic but its not.
Did you run the sim?....you can see it doing it in front of your eyes........its in BCM, the peak current can still go higher....no problem.

Do you know how peak current mode control works?...ill send you sim if you want...showing it in detail.
Can you show me any webpage or youtube video that will teach me how to get all the values like resistance and inductor values for flyback converter in BCM mode for constant current control.
 
What power you want?
0.5*L*I^2*f give you the values.
You pick a suitable frequency......or get the excel open and play eith it to see.

Do it postulatively....its good excel for that....stuff in a value, see how it comes out, then change to suit....you can get a really small worksheet liek that which is good.

Get an excel worksheet open and start playing.
Otherwise give me your spec and i do for you in minutes and send back to you here...
--- Updated ---

What power you want?
0.5*L*I^2*f give you the values.
You pick a suitable frequency......or get the excel open and play eith it to see.

Do it postulatively....its good excel for that....stuff in a value, see how it comes out, then change to suit....you can get a really small worksheet liek that which is good.

Get an excel worksheet open and start playing.
Otherwise give me your spec and i do for you in minutes and send back to you here...
--- Updated ---

What power you want?
0.5*L*I^2*f give you the values.
You pick a suitable frequency......or get the excel open and play eith it to see.

Do it postulatively....its good excel for that....stuff in a value, see how it comes out, then change to suit....you can get a really small worksheet liek that which is good.

Get an excel worksheet open and start playing.
Otherwise give me your spec and i do for you in minutes and send back to you here...
 

LaTeX Commands Quick-Menu:

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top