Sg3525 Ir2110 Smps Rar Files

Author: Topic: Current control for SMPS Arc Welding Machine (Read 21736 times) 0 Members and 1 Guest are viewing this topic.

In many situations, we need to use MOSFETs configured as high-side switches. Many a times we need to use MOSFETs configured as high-side and low-side switches. Such as in bridge circuits. In half-bridge circuits, we have 1 high-side MOSFET and 1 low-side MOSFET. In full-bridge circuits we have 2 high-side MOSFETs and 2 low-side MOSFETs. In such situations, there is a need to use high-side drive circuitry alongside low-side drive circuitry.

The most common way of driving MOSFETs in such cases is to use high-low side MOSFET drivers. Undoubtedly, the most popular such driver chip is the IR2110. And in this article/tutorial, I will talk about the IR2110.

It is common practice to use VDD = +5V. When VDD = +5V, the logic 1 input threshold is slightly higher than 3V.

Thus when VDD = +5V, the IR2110 can be used to drive loads when input “1” is higher than 3 point something volts. This means that it can be used for almost all circuits, since most circuits tend to have around 5V outputs. When you’re using microcontrollers the output voltage will be higher than 4V (when the microcontroller has VDD = +5V, which is quite common). When you’re using SG3525 or TL494 or other PWM controller, you are probably going to have them powered off greater than 10V, meaning the outputs will be higher than 8V when high. So, the IR2110 can be easily used.

HIN and LIN are the logic inputs. A high signal to HIN means that you want to drive the high-side MOSFET, meaning a high output is provided on HO.

A low signal to HIN means that you want to turn off the high-side MOSFET, meaning a low output is provided on HO. The output to HO – high or low – is not with respect to ground, but with respect to VS. We will soon see how a bootstrap circuitry (diode + capacitor) – utilizing VCC, VB and VS – is used to provide the floating supply to drive the MOSFET. VS is the high side floating supply return.

When high, the level on HO is equal to the level on VB, with respect to VS. When low, the level on HO is equal to VS, with respect to VS, effectively zero.

D1, C1 and C2 along with the IR2110 form the bootstrap circuitry. When LIN = 1 and Q2 is on, C1 and C2 get charged to the level on VB, which is one diode drop below +VCC. When LIN = 0 and HIN = 1, this charge on the C1 and C2 is used to add the extra voltage – VB in this case – above the source level of Q1 to drive the Q1 in high-side configuration. A large enough capacitance must be chosen for C1 so that it can supply the charge required to keep Q1 on for all the time.

C1 must also not be too large that charging is too slow and the voltage level does not rise sufficiently to keep the MOSFET on. The higher the on time, the higher the required capacitance. Thus, the lower the frequency, the higher the required capacitance for C1.

The higher the duty cycle, the higher the required capacitance for C1. Yes, there are formulae available for calculating the capacitance. However, there are many parameters involved, some of which we may not know – for example, the capacitor leakage current.

So, I just estimate the required capacitance. For low frequencies such as 50Hz, I use between 47 µF and 68 µF capacitance. For high frequencies like 30kHz to 50kHz, I use between 4.7 µF and 22 µF.

Since we’re using an electrolytic capacitor, a ceramic capacitor should be used in parallel with this capacitor. The ceramic capacitor is not required if the bootstrap capacitor is tantalum. Hi, P-channel MOSFETs almost always have a maximum VGS rating of -20V, but sometimes -30V. So, when the supply voltage is greater than 20V (or 30V for the ones with -30V VGS), you can't pull the gate to ground because doing so will destroy the MOSFET.

A voltage level shifter/translator is required in such cases. P-channel MOSFETs tend to have higher on-state resistances than similarly rated N-channel MOSFETs, so you will have higher losses for same/similar currents. Thus N-channel MOSFETs are preferred.

Moreover, for high voltages, finding suitable P-channel MOSFETs is quite a difficult task and the level shifter circuit is more complicated than the bootstrap circuit. Regards, Tahmid.

Hi, I have talked about the capacitor in the article/tutorial. You can find it in the paragraph right below Fig.

For simplicity I am posting the capacitor related section here: 'A large enough capacitance must be chosen for C1 so that it can supply the charge required to keep Q1 on for all the time. C1 must also not be too large that charging is too slow and the voltage level does not rise sufficiently to keep the MOSFET on. The higher the on time, the higher the required capacitance. Thus, the lower the frequency, the higher the required capacitance for C1. The higher the duty cycle, the higher the required capacitance for C1. Yes, there are formulae available for calculating the capacitance. However, there are many parameters involved, some of which we may not know – for example, the capacitor leakage current.

So, I just estimate the required capacitance. For low frequencies such as 50Hz, I use between 47µF and 68µF capacitance. For high frequencies like 30kHz to 50kHz, I use between 4.7µF and 22µF.

Since we’re using an electrolytic capacitor, a ceramic capacitor should be used in parallel with this capacitor. The ceramic capacitor is not required if the bootstrap capacitor is tantalum.'

Regards, Tahmid. In the high-side configuration, the MOSFET (I'm referring to N-channel here), the drain is connected to V+ while load is connected between source and ground. Let's say V+ is 12V. When the MOSFET is on, 12V is across the load. So, source is at +12V potential with respect to ground. So, if you were driving the MOSFET with 12V (with respect to ground), VGS is now equal to 0V. You need to drive the MOSFET with VGS >= 8V to fully turn it on.

And so, to drive the high-side MOSFET in the example above, you need to drive it with 24V (with respect to ground) or any other 12V source (12V between gate and source). This can be from an isolated supply or bootstrapping method, where the additional 12V above source is achieved using the capacitor. Hope this helps! Regards, Tahmid. Hi tahmid, But i thought that the maximum gate of MOSFET only can be support up to 15-20V only? First, im connecting 4 quadrant full bridge power mosfet as well.

I found that, a lot of theory state that ' to turn on an n-channel FET, a gate voltage must higher that the source voltage'. So, if Vpower supply =12 and VDD=15 and it surely can turn on the mosfet as well. But how about if Vpower supply=50 and VDD=15 and it sure cannot turn on the mosfet right? Because the gate voltage is not higher that the source voltage right? The gate of the MOSFET can only withstand 15V to 20V WITH RESPECT TO SOURCE.

If source is at 300V, then the gate can withstand up to 320V with respect to ground, which is 20V with respect to source. To turn the MOSFET on fully would require at least approximately 308V with respect to ground if the drive circuit is ground referenced. The IR2110 circuit uses VCC (not VDD), the bootstrap diode and the bootstrap capacitor to create the additional required voltage. The capacitor charges to (VCC - Vb)V when the high side MOSFET is on and VS is pulled to ground. When VS is no longer pulled to ground, the capacitor has a potential difference of (VCC - VB)V across it. The capacitor -ve is connected to VS which is the MOSFET source and so, the MOSFET gate is driven with a voltage higher than MOSFET source by as much as (VCC - VB)V. Of course the capacitor must be large enough that the potential difference across it doesn't drop too low for the MOSFET drive.

I hope that answers your question. If you have any doubts, feel free to ask. Regards, Tahmid. Here are three books that you can use to learn on your own: There's one book for mikroC, one for mikroBASIC and one for assembly.

Select the one depending on which language you want to use. While the boos are to be bought, the ebooks are free for viewing online.

So, just click the appropriate book and you'll get the ebook there. This is the mikroC book: mikroBASIC book: Assembly: Hope this helps! Regards, Tahmid. Please go through this to know about the benefit of the 1k resistor: ------------------------------------------------------------------------------------------------ Importance of the gate-to-source resistor: It prevents accidental turn on of the MOSFET by external noise usually at startup when the gate is floating. The MOSFET may sometimes turn on with a floating gate because of the internal drain to gate 'Miller' capacitance.

A gate to source resistor acts as a pull-down to ensure a low level for the MOSFET. I have had MOSFETs blowing up in high voltage circuits, without the resistor in place.

In most of the commercial power supplies / inverters I have seen, there is a 1k resistor used. A similar experience is narrated in Sanajaya Maniktala's 'Switching Power Supplies A to Z'. This is also talked of in Raymond Mack's 'Switching Power Supplies Demystified'.

------------------------------------------------------------------------------------------------ Using the diodes is good practice, but in most cases is unnecessary if you provide adequate heatsinking to deal with the added heating, and if you aren't switching at really high frequencies (few hundreds of kiloHertz). Usually, these diodes aren't required (frequency. Anonymous hi tahmid i built a 220V high frequency inverter operating it from a 12VDC battery however i adjusted the secondary winding of the transformer to operate from a 24VDC battery i run into problems like blowing out the fets in the DC TO DC stage could this be because i had not adjusted the primary turns of the transformer At 12VDC the primary turns is 3 turns center tapped 3 turns with secondary 90 turns and at 24VDC i have the same primary turns with secondary 45 turns could you please help me to find out where i am going wrong please help me out thanks and keep up the good work. The output of an SG3525 based H-bridge circuit will be modified sine wave. For pure sine wave, you should use microcontroller.

You can find quite a lot of tutorials on sine wave generation on my blog: No, you can't use that. The SG3525 can provide so little current. You can simply use a 7805 regulator that gets its input from the VDD powering the circuit. The circuit will obviously depend on your specifications. If you can't find IR2110, there are many other high-low side drivers available, although chances are that if you can't find IR2110 (which is probably the most popular high-low side driver), you can't find the otheres. Some alternatives: IR2113, L6385E, NCP5181, etc.

Regards, Tahmid. Anonymous Hello Tahmid, My project includes a TL494 to module a audio signal. I am using the IR2110 to drive the half MOSFETS bridge.

I use VCC=10V, VDD=5V, the signal comming from TL494 is a PWM 0-5V and 150kHz, it worked, but i have my wave on the gate of MOSFETS something like this: My output is a Capacitor with the Speaker in series to ground, and the mosfet supply is 10V. My teacher said, that i was going to have trouble with this high switching frequency, but it must be high have a better sound. Do you have and tip about something to reduce this effect? Hi Tahmid, Thank you for all the knowledge that you share with everyone. I am having issues with my inverter circuit using IGBTs with an H-bridge configuration.

It is a pure sine wave inverter with a battery bank of 24V with an out 220VAC using a step up transformer. I am using the configuration in figure 7 along with a PIC684 with the sine wave code. The output of the H-bridge seems to be good but I am having problems filtering the signal. I have tried different methods to filter the output but with very little success. I have tried using an LC filter with an inductance of 0.75mH (Choke to be more specific) and different capacitor values (ranging between 0.45uf-40uF). Inductor values higher than that are very expensive.

Is there an inductance range that you recommend using for filtering and is there a specific type of capacitor that is needed. I know they have to be non-polarized but it there a specific type you recommend (can starter capacitor be used). The output is very distorted and as I increase the capacitance the IR2110s gets damaged. The best result that I have been able to get has been using an RC filter but the power losses are tremendous and the waveform is not as clean as it should look like.

I am using 47uF(rated for 50VDC) bootstrap capacitors for the high side of the drivers. Have you experience the output the drivers getting damaged either the high side or low side and if so what was the probable cause. Does the inverter have to have a load in order to see the sine wave (I have tried both). Again thank you for everything that you do and I am putting your name on my presentation in the special thanks section. I have learned a tremedous amount about inverters from your blog. This project is going to be donated to a community in South Africa. I just want you to know that your help is not only teaching people but also helping people in need around the world.

Best regard, Pablo. Thank you for the compliments. I'm glad that I have been able top help and wish to be able to help as many people as I can.

The capacitor will usually lie in the range 1uF to 10uF. You should be able to use a starter capacitor. Avoid using too large capacitances, as, as you've seen, it damages the IR2110. The inductance will usually be much higher than the one you've used. Try in the range 1mH to 10mH in different steps. As cost is an issue, instead of buying ready made inductors, get a few toroid cores, some wire and wind the inductors yourself, adjusting the turns to adjust the inductance. Of course, you'd need an inductance meter to verify the inductance.

Or, if you know the core properties, you can calculate the number of turns. RC filters are for low power signals, not for 'power' filtering, as in the inverter, as I'm sure you already know. Here you must use LC filter. 47uF should be okay for the bootstrap capacitors.

In most cases where the drivers got damaged (in my experience), it was due to not using the 1k gate-to-source/emitter resistance for the MOSFETs/IGBTs, which prevents accidental turn on of the MOSFETs/IGBTs. Make sure you have a 1k gate-to-emitter resistance for each IGBT/IGBT combination. Read about it here: The inverter should have a minimum load to observe the output. The load doesn't have to be too high though. Few tens of watts should be okay.

I hope that answers all your questions. Feel free to ask if you have any further questions or doubt.

I would love to see your project at work and the project paper/thesis when you're done, if you don't mind sharing. I wish you success on your project.

I hope to add more to my blog to make it even better. Your comments and suggestions are welcome and will be highly appreciated! I hope that the project will benefit the community to which it will be donated. I'm glad to know that I am able to help people in need as well as teaching people. I hope I can continue helping and contributing. Best regards, Tahmid.

Anonymous hiiiiiiiiiii i had done 12v to 310v circuit using SG3525 and work good but. I used fig.8 with bootstrap as shown in this figure *** i using sG3525 to drive the H-bridge and i leave legs (1, 2,3,4,9,16) floating i used other legs with (leg 5 0.2 micro capacitance).(leg6 150k ohm resis.).(leg7 10ohm resis.).(leg8 1micro farad capacitance).(legs 10, 12 grounded). (leg11 direct to leg 10,12 of ir2110 ).

(leg14 direct to another legs 10,12 of ir2110).(leg 13,15 supply voltage 12v ).but the circuit didn't work and one of ir2110 blown!!!!! Download Game Yugioh Samsung Galaxy Y. When i use pic16f684 of your design instead of SG3525 the voltmeter read various swing readings such as ( 119,218,98.) not fixed 220v where is the problem???????

For the MOSFETs, use 1k resistor from gate to source of each MOSFET. Please go through this: That should solve the issue. If not, you will need to dig further. Don't leave all those legs floating. Go through this tutorial and then construct the correct circuit using the SG3525: Please upload your schematic to a file sharing site like rapidshare or an image hosting site like imageshack. Without looking at the schematic, it will be difficult to spot the problem.

For the PIC16F684, you must ensure that the circuit and code are both correct. Once again, you should upload the code and the schematic so that I can take a look. Regards, Tahmid. Thanks for the compliment.

I'm glad you've found it helpful. Please upload your schematic to a file sharing site like rapidshare or an image hosting site like imageshack. Without looking at the schematic, it will be difficult to spot the problem. If you're using one of the circuits I've provided, please mention which one, along with any modifications you've made. Where does the drive signal come from? Is the drive signal okay? What's the driving frequency and the duty cycle?

You must ensure that the IR2110's are okay and not damaged. You might've somehow damaged one of the IR2110's (the one on the side not working). Replace this IR2110 and see if it resolves the issue.

Regards, Tahmid. Please check all connections and make sure that you have constructed the circuit correctly. Can you post an image of the waveform from the Arduino?

Have you tried replacing the MOSFETs? Maybe you have damaged MOSFETs. On the IR2110 side that doesn't work, check the voltage (with respect to ground) at pins 3, 9 and 6 with the input signals off. With the input signals present, check the voltage (with respect to ground) at pins 3 and 9. Check the signal input to the pins 10 and 12 (use an oscilloscope). Check that pins 2, 11 and 13 are connected to ground. Let me know the values you obtain.

Regards, Tahmid. Anonymous Hi Tahmid, I am now adding a post from edaboard. I added a pic that includes 4 gate signals from TL494. L1 and L2 are output of TL494,whereas H1 and H2 inverted version of TL494 output.

My problem is that when ı drive the mosfets with these 4 gate signals, high side of the mosfets become so hot. (DC Bus = 50V, it will be 311V later) I can obtain sine signal with appropriate filter. I will also ask that is this low side of the gate signals are normal? Are they supposed to be like high side of gate signals? And can ı drive whole h-bridge with using just two H1 and H2 signals, ofc Q1-Q4, Q2-Q3 will have the same gate signals.(ignore the snubbers) http://obrazki.elektroda.pl/_.jpg.

You will have to choose the switching frequency as that will depend on your requirements. For example: If the H-bridge is going to drive a step up 50Hz iron core transformer, whose output will drive AC loads, then you need to use 50Hz. If you're going to drive AC loads directly, from a high-voltage DC, then you'll need to use 50Hz. If you're going to use the H-bridge to drive a ferrite core transformer at high frequency in a DC-DC converter, the output of which will then again be converted to AC, then you need to use high frequency, typically in the range 30kHz to 100kHz.

Hope this helps! Regards, Tahmid. Anonymous Hi Thanks a lot for your quick reply. I will check it out and will let you know.

One more thing, is it necessary that Hin and Lin signals come from controller or can I use constant DC voltage too? What I am doing rite now to check my Full H-Bridge circuit is I am using constant DC voltage and I have connected Hin of 1st IR2110 to Lin of second and vice versa (meaning I am controlling my bridge with 2 inputs). So do I need PWM signal from controller or can I also use constant voltage signals?

I'm really glad that this article has helped you. It provides me tremendous happiness to know that my articles are helping many people like you.

The blissful feeling of helping others is a feeling paralleled by few other things. I would like to make my blog even better so that it can reach out to help many more people like you. Your comments and suggestions are welcome and will be appreciated. There are certainly some differences between IR2112 and IR2110, most important of which is the ability to use up to 600V (IR2112) as opposed to 500V (IR2110). However, I believe, you can use the two drivers interchangeably in your circuit and that you can use the IR2112 in place of the IR2110 in your circuit. The operation is the same.

Best regards, Tahmid. I'm not sure what you mean by: '1) Do I have to write a program in Arduino to switch between HIN and LIN or do I use an Inverter? And if so do you have a link to an example code.' The drive signals must come from the controller in the circuit. If the controller in your circuit is the Arduino, then of course you have to write the code to send the appropriate drive signals. However, I think using a PWM controller such as SG3525 will be easier than writing your own code for the Arduino.

Go through this: Hope this helps. To shed some light to the possible reason for your problem, answer the following questions.

1) Did you use 1k gate-to-source resistances for each MOSFET? 2) Did you filter the output of the power supply? 3) Did you ensure that the output of the 12V power supply does not rise too high? Did you ensure that the output stays relatively stable at 12V? 4) Did you check the inputs to the IR2110 to ensure that there is no cross-conduction? 5) Did you replace the IR2110 and MOSFETs and try again, in case the IR2110 and/or the MOSFETs were damaged from the beginning? One thing you should remember is that a damaged IR2110 can quite easily destroy the MOSFETs.

Similarly, a damaged MOSFET can quite easily damage the IR2110. So, you must ensure that everything is okay to avoid damaging components in your circuit. Regards, Tahmid. Anonymous I 've implemented the Half-Bridge Circuit using IR2112, but i found the motor not working well, it changes its state rapidly(ON-OFF-ON-OFF-ON-OFF.), it seems that its OFF state is too long. I have used 22uF capacitor, bit then i read in the article that for low freq. A 47u-68u should be used, I connect two other capacitors (22u), but still no changes. The really boring thing in this circuit is that i used three different power supplies (1 for VDD,1 for VCC, and other for motor voltage).

Another Note is that i didnt connect COM pin with the same ground(with VSS), but i connect it to motor ground. Thanks in advance. Anonymous Hi Tahmid, Superb tutorial!

Very useful.Unfortunately i think you made a mistake, on paragraph 7 after Fig.4: 'The output on LO is with respect to ground. When high, the level on LO is equal to the level of VCC, with respect to VSS, effectively ground.' I think you mean 'effectively VCC'. Anyway i have some questions: Should the ground of the MOSFET be connected to VSS and COM as well?

My Gate-to-Source Voltage is +-20V so what is best 10,12,15V for VCC? I'm running at 24khz so what do you think C1 should be 22uF? What do you think the maximum duty cycle is, for the bootstrap to stil work? Would 100% work for short periouds?(to rise the current in the inductor faster) Should i use current limiting resistors betwen uC and IR2110? Thank you, Regards The M. Hi, Thanks for the compliment.

I'm glad that you've found the tutorial useful. What I meant in that sentence was that when high, the level on LO is equal to the level of VCC with respect to VSS, which (referring to VSS and not LO) is effectively ground.

Perhaps the sentence structure gave rise to that confusion. But you've understood correctly. The source of the low side MOSFET should be connected to VSS and COM. 20V VGS is the maximum rating of the MOSFET. Use 12V or 15V for VCC. Typically, anything between 10V and 18V will work properly.

12V and 15V are good since they aren't too high to be close to the maximum/upper limit but are high enough to fully drive the MOSFET. C1 = 22uF is okay. I cannot say that with certainty. It will depend on the MOSFET as well as the bootstrap capacitor. I think that about 70% could be achievable.

You might be able to get higher, but you should still test it, since I mostly stick to around or less than 50% (since that's what my design specifications dictate). The bootstrap drive can't be used for 100% duty cycle due to inability to charge the bootstrap capacitor. For 100% duty cycle, use an isolated power supply with isolated MOSFET gate drive. No, current limiting resistors betweeen microcontroller and IR2110 are not required. Regards, Tahmid. Cezary Hi Tahmid, First of all, I have to admit, that Your articles are really usefull and clear. However, I am working on the BLDC motor controller.

I want to use the STM32F4Discovery board as the main unit.I have almost finished the electronic scheme but I have some questions. My inverter is based on IR2110 drivers so I must choose a right logic level translator - STM32(3.3V)>IR2110(5V).Can you suggest one? STM32 board generates PWM signals with a frequency of 15kHz.I have found some ICs(ST2378ETTR,ADG3300BRUZ).Is it appripriate for 15kHz signals?

I want to use my controller witk 500 Wats BLDC motor.I have attached my scheme here(inverter and gate drivers). I will be very grateful if you check my scheme and give some advices. Thank you a lot. Hi thamid.really appreciate the knowledge that you have shared. Just wanted to know what are the criteria to calculate the current limiting resistor for the gate in the full bridge circuit.R1 and R2. Secondly i am not very well versed in electronics.could you please explain what you mean in your reply to anonymous dated 26th april, that 100% duty cycle cannot be used with the boot strap and that u stick to around 50%.to state my problem.

I wanted to use your above schematic for driving a 24V volt motor using all n channel bridge using irf540.incase i do use irf2110.do u mean to say that i wont be able to effectively drive my motor at 24volts and in actuality it would be getting 12V(50% duty cyle) and at max 16-17V(70%). Incase i want to go for 100% duty cycle what is the alternative tlp250???

I choose a value of the current limiting resistor depending on the MOSFET used and the frequency. Usually, 10R is okay for up to about 50kHz. Above that you might use 6.8R resistor. There's an application note that talks about current limiting resistors. I'll attach the link soon.

At high duty cycles, the bootstrap capacitor can't (quickly enough) fully charge and can't drive the high side MOSFET fully. You might be able to go to up to 70% duty cycle or a little more using a large enough bootstrap capacitor. However, for higher, you should look at other drive methods. I suggest you use a separate isolated power supply along with an isolated driver such as TLP250, TLP350, HCPL3120, etc. Regards, Tahmid. Hi Tahnid, Your blog is very impressive and helpful for innovation.

I am trying to make inverter for home appliances. My circuit is as similar as you have shown in fig-8. Primary section of my step up transformer is connected to out pins as shown in fig-8. I am successfully getting 240 V AC at the primary by varying duty cycles. My transformer rating is 12V AC to 300V AC. But now if we think about the reverse direction means if I give I/P of 230V AC to secondary of transformer I am getting about 24V DC across drain of higher side MOSFETs and source of lower side MOSFETs while all the MOSFETs are in off state. I am using IRFZ48N MOSFETs.

Now I want to charge my 12V DC battery which is connected across drain of higher side MOSFETs and source of lower side MOSFETs. Means battery is connected across 24V DC. While I connect battery to charge that O/P goes down to about 10.5V and battery is not getting enough current for charging. Let me know the proper direction to get required voltage and current to charge the battery. First of all, thanks for your great help and advice. Since the starter is manufacture connected to the ground, and p-mostet will not be able to handle such a large current to drive the motor to drive, that is why I am trying to make this n-channel mosfet to be connected to the motor to drive properly.

But can I know the near current of gate for each mosfet. So that I will start to use the current limiting circuits for it. Well I have not described that I am using for my bike starter motor in which the relay refuse to return to the original position (off/no connection) after releasing the push button switch, thus draining all the current to the motor resulting to wastage of fuel on the run and damaged the battery. Hi Tahmid, reading your blogs, I am quite impressed with the way you present and explain things. I would need some help regarding IR2113 and PWM. I'm working on a grid tie inverter project where I need to supply PWM signals to HIN/LIN inputs of the 2113s driving H-bridge mosfets. I had a look at your SG3525 blog.

Would you recommend using SG3525 to generate PWM signals from 50Hz grid signal. If so, how would you go about it. The grid has lots of distortion, noise and other rubbish that somehow needs to be filtered out before a PWM could be generated (I think). So far, I got the IR2113s and H bridge running from the grid's 50Hz using a dead-band circuit.

This only generates square waves at 50 Hz in sync and uploading efficiency is questionable. I prefer a more or less good looking PWMd sine wave in sync with the grid for uploading. Any suggestions or ideas are very much appreciated, rods, Selim. Anonymous Hi Tahmid First of all, thank you very much for helpful blog. I want to test the circuit of ir2110 before connected the MOSFET (H-bridge).I've followed the circuit in this post,but doesn't connect the MOSFET. I got the signal from LO, but the HO was 10-12 constant. May I ask you for my understand.

If I didn't connect the MOSFET, the Vs(pin5) of ir2110 have to connected to ground,that right? I can connect Vs to ground directly or have to connect the resistance before ground. Thank you very much, Regards, Song. Anonymous Thankyou for sharing your experience/knowledge. I am building a full wave bridge controlled rectifier with power factor correction.

The two drivers I used are IR2110 and the four mosfets are IR640. I used the figure 8 diagram for full bridge. This figure shows the leg of the H bridge as the output branch. I use this leg as input, so the two Vs pins are tied to the floating input. I have followed your recomendations closely but the circuit does not working.

I am guessing that the IR2110 cannot be used for rectifier applications or I am missing something or doing something wrong. Can you give me a clue where is the problema. Thank you again. I am Syed Tahmid Mahbub, from Dhaka, Bangladesh, born on August 1, 1994. Electronics is my passion and from class V, I have been learning electronics.

I learnt and worked mostly on SMPS, power electronics, microcontrollers and integration of microcontrollers with SMPS and power electronics. I've used PIC and AVR microcontrollers - PIC 10F, 12F, 16F, 18F, 24F, dsPIC 30F, 33F, PIC32, ATmega and ATtiny, integrating them with various SMPS and power electronics circuits. I have completed my Bachelor's degree from Cornell University (Class of 2017) in Ithaca, New York, USA, majoring in Electrical and Computer Engineering (ECE). I am a member of the forum www.edaboard.com, where I am an 'Advanced Member Level 5' (the highest level attainable) and also the forum allaboutcircuits.com, where I am a 'Senior Member'. I post to help solve electronics-related problems of engineers and engineering students from all over the world. I love watching and playing cricket and football (soccer), and listening to music. I am now a hardware engineer at Apple in Silicon Valley, California, USA.

Hello today i`m build new smps with ir2153, this schematic and pcb files are from forosdeelectronica and credits go to mniclauo. This smps was up and running but i have little problem when im feed it with high load. Smps details, 2 x irf740, ir2153, run freq is 100khz, protection use current sense trafo, 4x 1n4148 + mcr100, output voltage +43.5 and -43.4, output diode mur420 x4, output inductor 25+25 laps awg 18. Today im testing this smps with 200wrms discrete class D and solidstate with 8 output transistor (4 x 2SA1302 + 4 x 3281). Testing result: 1. With class D amp play loud at max, mosfet irf540 and 9540 only warm about 38 Celsius with psu heatsink.

Irf740 remain cool just warm a little, 220v bridge rectifier heat raise, ntc heat raise, main trafo about 45 Celsius, mur420 heat raise, output inductor warm. With solid state amp play loud at max, 4 x 2sa1302 and 4 x 2sc3281 is hot and need bigger heatsink, best if we place fan with heatsink, when play normal about half volume temp is stable. Irf740 remain cool just warm a little, 220v bridge rectifier heat raise, ntc heat raise, main trafo about 45 Celsius, mur420 heat raise, output inductor warm.

Regarding smps output voltage, +vcc drop 40.4 and -vee increase -45.6, there is the problem. For both test class D and with solid state amp will face same problem. This smps have no feedback and its run at 50% duty cycle all the time. From what i read from the net, smps without fb no need output filter because it will hurt output voltage. So i need opinion and suggestion from you all in diysmps and microsim to solve this problem, because i dont face this situation with ir2110 with fb there is no imbalance voltage appear.

I`ll post some picture, maybe 2-3 hours more becuase now here is about 6:36am, hehehehe:D early morning and no sunrise yet. I`ll post complete include schematics and pcb like usual:D.

Best regards, Azmi. Problem solved, it was inductor. So i make new inductor with 2 t106-26 seperate on each rails, it works ok now, both rails have same voltage drop and raise is together now.

I think i need to short ntc with relay and add 4 more output diode with mur420 x 8. Also bridge rectifier need to change from 8A to 16A and filter caps from 2 x 330uf to 4 x 330 uf. Will make new one will be more bigger than this size and with ETD49 wanna see the power can deliver.

One thing i notice after change inductor, before this i use 1 toroid t106-26 inductor for each rails wind together at one core, and i`ve got little noise coming out from solid state amp, but no noise with class D amp. After change the inductor there is no noise at all with solid state amp. Hi stewin, yes its t106-26, you can use any ferrite torodial core or you can simply take out EE16 core from old psu or EE19 which is suiteable for you, i`ll instruct you for the current sense trafo winding if u want to build it and its easy. The output watts is depend on your primary transformer, the large core u used then more power u will get. Estimated irf740 will go 500-600 watts and irfp460 will be more maybe 1kw. To reach this we must use at least etd49 core with multi strand wire awg 22-29.:D. For standard winding i`ve use 12 + 12 for the primary with 6 strand AWG29.

Because this smps is no feedback and always run at 50% duty cycle, the output voltage estimated about 6vdc per turn. So if we wound 6 turn with AWG29 ( i use 8 strand) we will have approx +- 36vdc at output. For the auxiliary i use AWG24 1 strand 4 laps. And for the +- 15vdc supply AWG24 1 strand 4 + 4 laps. For the current sense trafo: 25 turns for each branch.

Are performed in parallel (both arms at once), a branch between 'a' and 'c', the other between 'c' and 'b'. Regards, Azmi.

Yes, it can work without buffer, or u can simply add buffer your choice. You can do just replace it no problem. The protection section work on both side, i`ve test it. It will also shutdown if there is something short at 220 vac supply maybe cause by something else.

And at the output side will work, to push the power supply at maximum, proper 12v aux supply is needed and more caps bank at input after rectified, also u can make additional relay for the NTC, to by pass it after few second after power on. Renotrend, if u have high power resistor will do, or u can use hi power bulb such as 200-500 watts rated. There r many way to test em, or u can drive power amp with low bass freq see if ur supply hickup or not. If not enouh aux supply 12v to IR2153 u can hear some voice from it and seems IR2153 try to shutdown the pwm. If u wanna go for hi-power please use external aux supply for IR2153 and more caps bank at input (220 vac or 110 vac). Ok, thanks Norazmi, you have circuit protect for this smps?

Can you share to me thanks, reno. Hi Norazmi, After to collect some diagrams from the internet, I decided to start to play with the IR2153 and IRF740. Well, I had lots of problem to put the circuit to work without blow up the mosfets even without load.

The main problem was to design a decent and stable soft start circuit, especially when you use big caps on the output filter. The protection circuit was not difficult to put to work. That’s my conclusion: This is a cheaper solution but not the best, if by compared with a regular smps. Now the circuit is working fine, but I don’t trust too much! What do you think about it? I dont know what u mean by the best, there is no the best but working with high freq we need to watch the efficiency how much power it can provide. Why we need to put big caps at output with smps, its not needed and its only for linear supply to put big caps.

Everything in the world have the weakness and we cant avoid it 100%. Ur smps blown up at start up with big caps at output because they feed with inrush current at startup u can use relay or something to stablelize it first at initial startup after that, relay will contact with big caps bank at output i think it should be ok. Btw for initial start with smps, its good to start with ir2153 and irf740, they r cheap and easy to make for medium power range.

With good calculation and winding of trafo, will provide good power supply to feed ur amp or else. Its good to learn and of course we need to sacrifice something and after success with few tries u will happy with it:D Bettecher what u gonna to feed with 3.2kw smps? Norazmi, You’re right, the efficiency and the cost is the key for any smps. Of course I know we don’t need big caps at the filter, but when you feed a power amp. We need a extra peak current.

I’m talking about 5000 uF total, 2500 each rail. Sure it will produce more inrush current at the start up.

My circuit already have a NTC and also a relay at the ac input but still need a good soft start circuit. See my solution below.

By using that circuit, I didn’t have any problems at all. By the way a 1kW smps is enough for me! Norazmi, I’m here just to share my experiences and your comments are always welcome. Nice, oh., i did test experiment in real, using clamp supply ir2153 by resistor feed to zener it will hickup at peak current when u push hard. Even im using irf740 so if using irfp460 will be more demanding supply at ir2153 vcc. I think its because high voltage also drop and it will effect ir2153 supply since we take the power from same source. So what i did, there will be 2 solution, 1.

Using additional supply from main trafo and limit with LM7812, 2. Using external supply.

Yeah please post some picture here:D. Btw irf740 is good up to 600 wrms and 800wrms peak with proper trafo winding. Norazmi, my new trafo is not already but, I did run a test today using an old trafo that was made before rated to 500w and believe or not, I successfully extracted 630 watts! 70+70 x 4.5 amps. Everything warm up a little bit but, not too much. By using my formulas, here are the results for my new trafo: Fs = 80 Khz Min. Ac input = 90vca = (2.28 * 90) ' for 90 vca = ( 205.2 VDC ) worst case!

Vo = 50 + 50 Io = 10 amp. Ps = 1000Total power B =1500Gauss J =319 Kj =397 Ap =5,1Min. Area product required I’ll use core EE - 42/21/20 Ap =6,14( Aw * Ae ) cm4 Ae =2,4(Ae) cm2 AL5800(Al) cm2 Main results are: Ip rms = 8,186 amp. Is rms = 14,14 amp.

Apri = 0,0256 - Acu Prim. (Cm2) Asec = 0,0443 - Acu Sec (Cm2) Np = 16 - Prim. Turns - wires = 40 x 29 awg - (Skin effect) Ns = 7 + 7 – Sec. Turns - wires = 34 + 34 x 29 awg. What do you think about that? Please, let me know.

Here are some pics from my last 1kw spms based on IRF460, IR2110, SG3525, TL431 and optocoupler PC817. As you can see, there are too many components, that’s why I decide tried the IR2153. The design is working very well and one channel of my D class amp is included. Hope you like! Very nice job friend Bettecher. It would be great to see some testing results of your SMPS, as I can see that's a regulated version? How much voltage drop you have at 500W load?

Do you have an oscilliscope to inspect some waves? Very nice job friend Bettecher. It would be great to see some testing results of your SMPS, as I can see that's a regulated version? How much voltage drop you have at 500W load? Do you have an oscilliscope to inspect some waves? Regards:UP: Thanks MicrosiM, Answer your question: The circuit is based on TL431 and I don’t have any voltage dropped until 800w when the short protection circuit starts to limit the current smoothly. ( I do not use the shutdown pin.) It’s done by reducing the duty cycle.

Yes, I have an old 100 mhz four channels oscilloscope and if you want I can post here any test, just give me time for that. I also can provide de schematic if you need! Do you have pcb for it? Yes, but It was the first pcb made and needs some changes.

If you want I can send to you by e-mail. I don’t think is a good idea to post a corrupted pcb here. By the way, I finished my new trafo today. It works and has a good voltage regulation but, I having a saturation problem. ( It’s warming ). I think the problem is gauss value that used.

I decided to increase the gauss a little bit in order to improve the voltage regulation but I think it was too much! Do you have some suggestion or a reasonable value? If ur trafo warming while working with 80 Khz it does not mean saturation.

Its normal trafo warm up a little but not going to overheat. I have with me ir2153 working with 100Khz exactly i built what i post here, and remember snubber will help much to keep trafo stable and not overheat, without them trafo will go heat like crazy. Take time to look at my circuit, maybe u need to put snubber like mine. Because i`ve test with ucd discrete amp with totem pole 4 pcs irfp250N to push 2 ohm load audiobahn Q series with IR2153 smps and voltage drop only +- 3vdc both rails when play hard. I use only ETD35 trafo and its enough to pull out about 600 wrms for 100Khz.

When u smell something from speaker because of coils heats up and when u touch the speaker cone hot, u know that the power deliver to the speaker is more than enough and smps done their job very well. I use main caps filter 680uf/250V x 4. P/s - i dont use output inductor because its not regulated, with inductor it will drop more and main trafo will be hot, but when play with IR2110 + SG3525 its ok to have output inductor because the supply is regulated. Thats what i`ve done experiment in real life. Keep going Bettecher we will try to solve your problem together:D. Marco, u can use bettecher pcb if possible, or u can use mine at main post, few components need to be change to get proper 1kw out.

Trafo edt49 will do the job and with 100 Khz switching u can get more than 1kw by using irfp460. Norazmi, if I can find IR2153 in my place I'd love to give this SMPS a try,:UP: it seems though that there are modifications done from the original circuit (that is you put 2 inductors in the +/-rail one to + and the other to - ) have you made an updated schematic to come-up with 1KW output power? And where is bettecher pdf files?

Is it being shared for DIY?:D Regards! Hi there, Michelle: Its not really difficult to make 1kw offline smps. With Irfp460 will be more easy as rds-on will be more faster than irf740. Also if you can get etd49 will be more great.

With switching 100Khz you can get more than 1kw, add more caps bank at hi-voltage, at least 15A diode at hi-voltage. At this time i dont have pcb yet with IR2153 to go with 1kwrms power. But i can make one if u want.

Just give me some time to have it done. Abet: Ok to be clear it up, Actually i did modification with remove the output inductor. I did put 2 inductor at the second test but without them the smps will be more efficient. Because of this smps is not regulate, so inductor just make this smps will be more suffer and ineffective. I`ll take a look again with bettecher design and he already mention that the design is almost same at page 1 project. And i already mention regarding aux supply to IR2153 need to be stable or use external aux for good result. Also you can see bettecher use relay to short the NTC few secs after smps power up.

Bettecher, Any plans on sharing your PCB for DIY? Have you replaced the other version with the relay? Though I have not constructed an SMPS, I thought the use of an added relay on board adds up to better protection of the whole circuit.

Abet, The pcb is not updated. I made lots of changes on the circuit but I didn’t have time to make the corrections. I don’t want to share a corrupted pcb here. My schematic doesn’t need a relay because already has a soft start and short circuit protection. The diagram is working fine and you can build if you want.

The circuit is based on the ICs SG3525 and IR2110. You can find all info about on this forum also a good pcb. Choose one and make the little changes. Marcos Bettecher.

Stewin, nice work! Notes: Your schematic is missing the output filter coils. In my case I used a 2 x 22 uH. Don’t forget!

You can use a EI33 from an old PC smps. Should have a 0,1 mm gap. ( pieces of paper sheet). TR1 should be 1:50 turns on iron powder core. (Small green core from a old computer board). Winding should be in the right phase otherwise Q1 will get hot really fast.

(Just invert the wires on the circuit if Q1 is get hot) You can not extract 1Kw from the IRF740 maybe 700 W in a better case! Better if you use the IRF460LC. Give me a little time to recalculate the right number of turns for your ETD44.

Regards Marcos Bettecher. Thanks again. I use irf 740 because i wanted a small smps, but i will add irfp460lc for the larger version. Is it a must to use output filter coil? Because in LabGruppen-FP3400 pwr amp user manual, they use none except 2.2ohms resistor and a 4.7nf on the two rectifying diodes used. I have used the same and i had no problem.

The protect small former i have used one from an old ups.As i see it is a small EE core and has 3turns on primary side and many turns the sec side will it work or i need to coil the 50:1 iron powder core? Stewin, The 2.2ohms resistor and a 4.7nf on the two rectifying diodes are used as snubber. When a smps is working in pwm mode (controlling the voltage output) yes, it must have the filters!

But, if works without control ( 50% duty cycle) you don’t need it. The only problem about the current transform is the resonance. Since you’re working in a high frequency, a good sense is to reduce the max is possible the primary inductance.

(less turns). Remember it should works like a current sense shunt resistor. (less inductance is better) Regards Marcos Bettecher. Norazmi, my new trafo is not already but, I did run a test today using an old trafo that was made before rated to 500w and believe or not, I successfully extracted 630 watts! 70+70 x 4.5 amps. Everything warm up a little bit but, not too much. By using my formulas, here are the results for my new trafo: Fs = 80 Khz Min.

Ac input = 90vca = (2.28 * 90) ' for 90 vca = ( 205.2 VDC ) worst case! Vo = 50 + 50 Io = 10 amp. Ps = 1000Total power B =1500Gauss J =319 Kj =397 Ap =5,1Min.

Area product required I’ll use core EE - 42/21/20 Ap =6,14( Aw * Ae ) cm4 Ae =2,4(Ae) cm2 AL5800(Al) cm2 Main results are: Ip rms = 8,186 amp. Is rms = 14,14 amp. Apri = 0,0256 - Acu Prim. (Cm2) Asec = 0,0443 - Acu Sec (Cm2) Np = 16 - Prim. Turns - wires = 40 x 29 awg - (Skin effect) Ns = 7 + 7 – Sec. Turns - wires = 34 + 34 x 29 awg. What do you think about that?

Please, let me know. Bettecher, for your calculations here, are they for a full bridge you were talking about earlier or for a half bridge you posted the schematic for a little later in thread. Bettecher, I have been studying how to make switching power supplies and i always try to check my calculations with others when enough information is given. The only information that you left out was the ferrite material used in your EE42/21/20 core. This makes it hard to confirm the best flux level, but since you gave the flux level you used i could still calculate the primary turns. Please take a look at my work and compare it to yours.

I belive you should have used about 33 primary turns and you used 16 turns. Np = (Vmax * 10 ^ 8) / (X * F * B * Ae) The constant X used in the primary turns formula can be any of the following. 1.11 - compensates for the sine wave ratio of RMS to average voltage. Used for standard sine wave transformer design. 1 - Used for half wave square wave operation of a transformer. A flyback converter uses this constant.

2 - This is the square wave constant above multiplied by 2. There are two semi cycles in each full cycle (Hertz refers to full cycles). A half bridge uses this constant. 4 - This is the constant 2 above multiplied by two for topographies that have a flux swing in all 4 quadrants, like a full bridge circuit. These are the only constants there are for this formula. UNITS IN FORMULA Ae = Area of center leg of transformer in CM squared B = Flux in Gauss, common values of 1400 to 1600 F = Frequency the transformer is switched at in Hz Vmax = Voltage in volts the transformer sees GENERAL RULES For a half bridge driven off of a split capacitor supply the transformer only sees half of the bus voltage.

Vpri minV gives you the minimum turns to use on the primary, this sets the flux number you used in the original calculations as a minimum. As the voltage rises to Vpri maxV the flux rises, possibly to the point of saturation, if you started to high in flux density to begin with. Vpri maxV gives you the maximum turns to use on the primary, this limits the flux to a maximum of your original number used in the formula. The flux will go down when the voltage goes to Vpri minV. This will make a safe design. Increasing the Gauss decreases the turns. As the turns go down the Gauss goes up NUMBERS TO USE FOR YOUR DESIGN Ae = 2.4 CM squared B = 1500 Gauss The constant 2 as defined above Frequency 80,000 Hz Vmax, from schematic, 380Vdc bus / 2 for driving transformer with split capacitor power supply = 190Vdc Np = (Vmax * 10 ^ 8) / (X * F * B * Ae) Np = 190Vdc * 100,000,000 / 2 * 80,000Hz * 1500G * 2.4CM squared Np = 19,000,000,000 / 576,000,000 Np = 32.986 = 33 turns.

Your calculations showed 16 turns. So if i am right you are actually running 3000 Gauss on your transformer. Let me know what you think. OK guys, I found my mistake. In this particular topology used by Bettecher, it turns out that the flux does swing in all 4 quadrants even though it is a half bridge. Therefore this will multiply the flux x 2 because it swings through zero flux both ways 1500Gauss, for a total of 3000 Gauss. I found this out by looking in Keith Billings book Switchmode Power Supply Handbook, Chapter 12, Half bridge Push-Pull Duty Ratio Controlled Converters.

The exact same power circuit Bettecher used is in there with an explanation of how it works with an example calculating transformer turns. So in the formula above in my calculations i should have used the 4 constant and not the 2 constant. Hi guys, sorry for the delay. I got a health problem (sinusitis). Now I'm fine.

Ok, I had the same doubts in the past and I had to make a compilation of various formulas in order to choose the best ones. After to read some books like: Transformer and inductor design - Colonel W.M. Mclyman Switching Power Supply Design Third Edition - Abraham I. Pressman Power Supply Cookbook 2E - Marty Brown And others from my country, I finally found some workable equations and was possible to make some successfully smps designs. I’d like to say thanks to everybody for the help and the comments what are extremely welcome. I’m here to learn and share.

Regards Marcos Bettecher. I've finally completed the pcb of the smps. I chose ir2110 because it has protection features and here where i am they both are sold at the same price as ir2153;). I like the protection because if the amp gets over loaded or the amp transistor fails the power supply will switch itself off hopefully:( Stewin, By looking your schematic I found some errors. Please, consider to make some changes: 1.

LED2 needs to be inverted. LED3 needs to be inverted. Better 4k7 4. R24 - 33K – the valor should be increased for 80 + 80 volts. ( the original design was 50 + 50). Minimal load resistor not found. Output filter – Inductors not found.( don’t forget!) Regards Marcos Bettecher.

Norazmi and others who wrote on this treat. I want to build this SMPS and this will be my FIRST SMPS. I used schematic from first page that norazmi posted but I want 1KW of power for my class D amp. I will use ETD49 transformer, IRFP460 and unit with IRF730 for IR2153 supply with zeners 15V (you can see that on board.

Board is small (real size is in PDF) and I found 1000uF 200V caps with 15mm diameter and 60mm high. I have few questions: 1) Is my board good for this smps?

2) I can find in my country ETD49 core 3C90 and ETD49 core 3F3, which of these can I use? 3) Do I need air gap on central leg? 4) I dont know number of turns and wire ticknes for primary and secondary (I need +-50V) 5) Number of turns for EE16 current sens transformer and wire tickens? 6) And one realy stupid question:) What means when you say for primary turns on example 11+11, is this means 22 turns on primary side?

When you say on example 6+6 on secondary side I know that means that seconadary side has 3 pins (~ 0 ~ voltage), but primary side doesnt has 3 pins, only two, that I dont understand. Thanks and Ragards Nikola. You need amp who can produce min 150W at 4 ohms and 300W at 2 ohms. To avoide cliping and distorsion I recomanded you to build AMP with +-40V. At 4 ohms thats aprox 180W and on 2 ohms that would be aprox 350W without cliping but if you push it hard you can get 380W. 400-450W power supplay will be good (2 x 40V 5A per rail) That is for on example for Class D AMP, but if you wanna build AB class amp, voltage will be +-45V for same power because of voltage trops on emitter/source resistors and VAS stage.

This will be theory. You need 300W on 2 ohms: 40V / 1.41 = ~28.3Vout 28.3V / 2ohms = 14.15Aout 14.15A * 28.3V = 400W @ 2ohms and 200W @ 4ohms But without distorsion and cliping you will get pure 350W @ 2 ohms to drive thats subs:) Regards. Dzony: no gaps for half bridge or with gaps u will lost lots of power inside the core. Your pcb is good. Nice and clean track with good gaps between each track i like it.

U can use edt49 ferrite core, look at datasheet between 3C90 and 3F3 which one have high freq is better. No need air gap for half bridge topology. Its easy to calculate the number of turn with tools here at diysmps already post how to calculate the primary and secondary turn with range of freq. Say 100Khz, primary u need 13 turn so make it 14. So each primary will have 7+7 turn.

And for the secondary u need 10 turn which mean 5+5 turn for each side. Wire thickness is depend what u got.

Parallel more wire will have good efficiency make it awg29 x 7 for primary and awg29 x 10 for secondary because u have enough space for the edt49 trafo. You can use iron powder core T106-26(yellow) for current sense no need ee16 which is too small and hard to wind. U can use awg24 25 + 25 each. Thats not stupid question and i think u already got the answer:D I can see u made the aux supply using irf730, did u simulate it yet with high supply like >300 vdc it can stand for long?

Not overheat? Why not make another small linear supply to provide aux supply for ir2153. Ur going to make high power smps try to avoid taking aux supply from main rails psu. When smps try to get more power main power rails will have high tension which can cause it will drop, there u may have problem with aux supply also will drop. Regards Azmi. Azmy, thanks for remanding me, I totaly forgat for snubers.

One problem came out. My chasis for amp is Aluminium and 1H high (44mm), I planed to use chasis like big heatsink for IRFP460, if I mount 2 of this SMPS's and 10n 1KV between GND and chasis, can I couse some problem? If that can be good, I can insert separate heastink for each smps without contact with chasis. Tell me, how can I calculate current transformer, how many turnns on secondary and primary side. I understand how it works: second pin from main transformer (who goes to 1uF and float) is primary of current transformer. That few turns inducate voltage and curent in T106-26 core and on secondary of that core, with diodes 1N4148 we get positiv voltage who is the trigger on MCR100. If current accros primary is low, we dont have enough voltage to trigger MCR100, when current is high, MCR short CT on gnd and IR2153 is 'shut down'.

I ask you that because I wanna to learn and use that idea for D class amp like overcurrent protect (secondary can start some elctronic for SD pin) Thanks again for everything:). 10nf 1kv it must be close to smps pcb and the other one working like an rf antena to avoid noise coming from swithing device at power mosfet. You can insert separate heatsink no problem with that. For power amp its different application this current sense trafo cannot be applied successful within class D or AB amp, u need to trigger with resistor example 0.5 ohm 5W or any sense resistor to feed the mcr. Short circuit protection already include with smps and u need additional amp output short circuit protection which u can use mr bruno design using opto and ne555 to trigger it. That protection can be applied to any amp and any class. That protection is good to be applied into your power amp.

Hi azmi, it's working now. IN4007 not soldering perfect, sometimes not connected. When i touch it with test pen it's make disconneted rapidly, then the mosfet, ir2153 blown. I replace them, and it's working.

One question, can i remove the bohlam now?? It's normal to hear 'click' when start up? Playing with SMPS while its connected to the 220VAC line?

X-( that sounds like bad soldering, bad PCB etc. If you continue this way, you may face your end!:x: Its normal to get SMPS blown from what you have described! I am sorry for you. Please, check every thing and show us clear pictures, of PCB, we will try to figure out whats going on. Hi all Microsim, sorry for what i have done, i will not do that again.

Azmi, i am not remove the aux when it blown. I am thinking about the voltage that AUX applied at VCC pin. With Secondary 5 turn, i get 29,6 VDC. With 4 turn of Aux, it's around 24 VDC, limiting current resistor is 33O ohm, will provide about 72ma with 24vdc to Vcc pin. This is to much, and kill the IR2153, make Mosfet short to ground and blown.

I compare this with start up current from 47K/5W resistor that only apply 12vdc/ 6.3ma to vcc pin. The smps working when series with the bulb (40w), as the bulb act as current limiter for the smps.

I have check and check again, main caps not blown, but the plastic was burned when mosfet blown. The trace clean, no short at all. Compared with the original schema, the different is only Input Filter. Pls give me some advice. Regards, Glo.

Hi glo, the problem is come from aux supply 24 vdc is too much. Originally i state using ei33 and 4 turn, with ei35 is diffrent, thats why i`m asking u before to remove aux winding for testing and i would like suggest u to put voltage regulator for aux supply if u want use aux supply from main trafo. For the main cap 2200 x 2 is ok no problem with that, NTC should be change to 10 ohm and recommended fuse for testing 2 A. This is clearly the problem is come from your aux supply and overload the ir2153 and failed to operation.

Yes i have remove it. Primary seems was burn, i tried to measure, give me 99 ohm form CT to first half. Need to winding new trafo. I need to understanding this smps before make 2KW from core below. Gloriel, look at his PDF and give us the dimensions for the transformer you are using and the one you want to use.

I will tell you what transformers you have. Need dimensions for A, B, C, D, E, F, H. Also give us 2 pictures one of each transformer full apart so we can see exactly what type it is. We need to see both half's of the transformer.

The schema high power with fan control but the input caps that's what to be increased to about 200vlts x 1000uf. I've also corrected the voltage adjustment, this works for me in real life. Sadly, i cant use a potential meter only zeener diodes, but works perfectly and is stable with or without load. In order for you to change voltage adjust the zeener diodes e.g. If you want +/-40v you use zeeners which will sum up to 80v e.g. 60v and 20v or 70v and 10v etc.

I've tried the version with irf740 and it works perfectly, I want to build the one with irfp460 for a bit more power. Here are the schems, I will post the artwork later. The current protection is working perfectly just adjusting the preset potential meter vr1 will help you choose the current you want.

For my testing I loaded two 100wts bulbs and the protection responded even to that load. My transormer turns ratio are >>primary 21 - 0 -13turns, >>secondary is 9 turns 0-9 turns >>Aux for fan and IC control 4turns The other small former >>Current protect former small from an old ups i see the primary turns are 3turns, the secondary are many, i assume they are 100 turns It works perfectly with more turns on the primary side, my previous experiments with transformer gate-drive showed that it used less current on the mains if i used more primary turns but you guys can simulate and tell me it i'm wrong. The schema high power with fan control but the input caps that's what to be increased to about 200vlts x 1000uf. I've also corrected the voltage adjustment, this works for me in real life. Sadly, i cant use a potential meter only zeener diodes, but works perfectly and is stable with or without load. In order for you to change voltage adjust the zeener diodes e.g.

If you want +/-40v you use zeeners which will sum up to 80v e.g. 60v and 20v or 70v and 10v etc. I've tried the version with irf740 and it works perfectly, I want to build the one with irfp460 for a bit more power. Here are the schems, I will post the artwork later. The current protection is working perfectly just adjusting the preset potential meter vr1 will help you choose the current you want.

For my testing I loaded two 100wts bulbs and the protection responded even to that load. This cannot be true because there is no output differential inductor used in secondary after diodes to hold energy and perform regulation. You are sending feedback thru opto+zener to error amp of SG3525 for regulating output by varying dutycycle, but where is the output inductor????????????? Flyback operation is totally different than what you are doing as a halfbridge stage driving primary of transformer. In flyback you dont need full bridge rectification because the supply is working in one quadrant only, not like in half bridge which is using 2 quadrants. Flyback is not a real transformer, it is two coupled inductors, using the first as a energy storage, then the energy is transfered to the 'secondary' by magnetic coupling. It still behaves as an inductor, hence no need for an inductor on the secondary circuit.

I'm building Azmi's SMPS and i have some questions: 1. The 2.7K 1w resistances must be 2.7K or I can put 3.3K or 4.7K for example? (the resistances between plus - gnd and minus - gnd on power output) 2. The value of zenner diode for IR2153 is 13V, can I put a 12V zenner? Instead of MCR100 I put BT151. I think it's ok, right?

On output I put 2 inductors, it is ok? (see photos) 5. How about sense trafo? I make it but I have problems with connection on PCB. How it must be connected on A, B, C points on PCB? I put all this questions for my safe, i saw that a SMPS is not a joke, it may explode very easy. After this, I want to build the 200w class D amp:D This is my first SMPS-) Thanks, Vali.

2.7k 1W is for +/- 35vdc at output, u can vary the resistance value depends on output voltage, ie: 3.3k 2W good for +/- 50vdc and lower, +/- 60-65 vdc use 4.7k 2w, and etc. The value of zener i use 13V because i dont use external or build on aux supply. 13V will produce about 12.7 vdc after clamping at high supply, ie: use 15K 3W x 2 in series means equal to 30k 6W. When u push smps at peak, with 12V clamping zener the smps might be hickup because of not enough aux supply of ir2153. So to avoid it, u can use more caps bank at main supply, or increase clamping diode to 13V.

BT151 it should work, and if incorrect value for the MCR the smps will not turn on. And it will stay at protection mode. Thats ok with DC filter u put and its good. Thats not inductor, and its dc filter to avoid noise coming from smps and unwanted noise. Refer to back page, i think i already explain how to connect current sense trafo. Something for smps diyers warm regards michelle Good day to all, it's my 1st post in this forum I hope I'm welcome here!

Hi Michelle, I'm a bit interested on the schematic you posted somewhere here with self-oscillating GDT. The operation is quite familiar to me because its just an adaptation of popular CFL inverter circuit which we can find easily on the net, am I right? The starting oscillation relies on Diac just like any other 12V/50W halogen lamp inverter driver does. You mentioned previously that you are going to buy one of that smps for cloning, do you have it already? I observed that there's no fuses on the secondary, do you think the primary fuse can handle the entire protection for it? My main query to you is, how's the entire performance of that smps? Sorry for my bit lengthy question, I'm just curious because I'm also wishing to have a high power SMPS sooner.

The components used in oscillating circuit which was build by many diyer's here can't find easily in my place and maybe this self-oscillating trafo is the answer to my needs. Thanks in advance, 576Pinoy_Tech. Have you resolved your problem regarding overvoltage of aux supply? What ntc did you use? You may also try using external softstart circuit just to isolate if the problem is caused by inrush current.

BTW, I've successfully built this SMPS with the great help of this forum and Bettecher. I solved aux supply overvoltage by using lm7815 regulator., what resistor value u use for start up and gate? How many turn of current sense transfromer?