600/800W 220~230v AC Inverter SMPS from 12v DC

[szhighstar]

formular: V=N*dBS/dt, when you know V( minimum DC voltage) and B( core across section area) and t(turn on time), you can get N(turns)
V=L*di/dt, when you knowV( minimum DC voltage) and i (current) and t (turn on time), you can get L( inductance)

 

[sam781]

@Goldsmith: Actually the problem is I do not have datasheet of the core available at my hand right now. Also I do not have LCR meter. I'm planning to buy a low cost (actually lowest cost :-D) LCR meter.

 

[sam781]

I just assume flux density of my core as 1600 at 50Khz. As per the calculation, it requires 5 turns (for 22.28uH inductance) in primary with 10AWG wire. I wanted to use fourteen 22AWG wire in primary instead of a single 10AWG.
Could you let me know whether there is any difference in inductance between a single wire winding and 14 wire winding?

 

[rajudp]

are you using flyback?, don't get much worry about inductance for half and full bridge application

 

[goldsmith]

@Goldsmith: Actually the problem is I do not have datasheet of the core available at my hand right now. Also I do not have LCR meter. I'm planning to buy a low cost (actually lowest cost ) LCR meter.

Hi sam
Are you joking ? design an inductor or a transformer without referring to the datasheet or perhaps using an LCR meter ??!!
About LCR meter , i agree a low cost LCR meter would be ok for these aims . ( my first LCR meter was low cost too but it was pretty good ! and handled my aims ! ) :wink:

Could you let me know whether there is any difference in inductance between a single wire winding and 14 wire winding?

There isn't any difference between their inductances . both are the same ! but capacity of high current handling is increased .
Good luck
Goldsmith

don't get much worry about inductance for half and full bridge application

Hi Rajudp
What do you said ? are you sure about it ? i'm disagree about your statement .
Regards
Goldsmith

 

[rajudp]

i am using lcr meter only to adjust inductance in flyback design ( to adjust the gap according to inductance),

 

[sam781]

I wanted to buy this LCR meter -**broken link removed**

I was thinking about the equivalent inductance of parallel connection. If multiple wires are used for winding, will not there be parallel inductance?

 

[picgak]

I wanted to buy this LCR meter -**broken link removed**

I was thinking about the equivalent inductance of parallel connection. If multiple wires are used for winding, will not there be parallel inductance?

sam,
If you are winding multiple wires together there wont be any inductance difference . but if you are winding one wire after one separately, there will be problem.
regards ani

 

[goldsmith]

Hi sam
Yes that LCR meter is a good one to start . you'll be able to design many magnetic elements with that .
Good luck
Goldsmith

 

[MRahman]

First of all you would not used single strands round wire in the working area of over 500hz, Litz or multistrands or flat wire should be used to reduce skin effect. Because high frequency current travels on the outer surface area not the total area of the conductor, so that you need bigger surface area. The inductance will difference in every kind of stage. For instance before impregnation and after you'll get different inductance reading. See https://www.youtube.com/watch?v=3nfqBzPMknY might help you.

 

[sam781]

What about next stage? Rectification and filtering?

 

[goldsmith]

Hi Mrahman . when sam told about 500HZ ? is skinn effect important in 500HZ ? of course no ! but about litz wire i'm agree with you because at high frequencies it will increase effective diameter . thus it is a good idea ( i have used some litz wires in parallel in many of my designs . )
Best Wishes
Goldsmith


Hi sam
What do you mean by next stage ? could you design first stage ? if yes did you build it ? if yes tell me the result . thus we can continue .
Best Wishes
Goldsmith

 

[sam781]

Hi Goldsmith,
I got overall idea on designing dc-dc converter in this thread discussion with rajudp. You also helped me a lot. Now, I'll make it practically. I already mentioned that the main constraints are - (1) I do not have product number (hence, datasheet) of the ferrite core. (2) I do not have LCR meter.

Still I'm planning to make it based on some assumptions discussed in this thread. In this meantime I want to learn more. :smile:

 

[goldsmith]

In this meantime I want to learn more.

Hi sam
It is pretty good that you want learn new things . and i realize that you are at learning process . as i'm ! still ! :wink:
Good luck
Goldsmith

 

[sam781]

Hi sam
It is pretty good that you want learn new things . and i realize that you are at learning process . as i'm ! still ! :wink:
Good luck
Goldsmith

Yes, I've already learned a lot from here from you, Mr. rajudp and many other experts. We've a long EID vacation coming up soon and I want to start practical work after the vac. So, in this meantime, I want to learn (and be clear the concepts) as much as possible with the help of you experts.

Please let us come back to learning session again. I've some confusion regarding transformer primary-secondary voltage-current calculation. Primary current, Ip= (Pout+Ploss)/Vin. Secondary current, Is= Ip/turns ratio
Is it okay? Or need to consider other things?

 

[goldsmith]

Hi sam
A non ideal transformer ( real transformer ) will have some kind of loss . the formula V1/V2=I2/I1 is for ideal transformers ( however at higher frequencies it will be nearby actual ) . in practice and real examples you can find the I1 with that formula and then summation between that and loss current !
Best Luck
Goldsmith

 

[sam781]

ok. Understood.

Now, I want to generate gate signal using microcontroller. Here is my idea -
1. Generating 50Khz square wave signal capability of varying duty cycle (d) from 70-90%
2. Measure the scaled (230v ac rms equal 5v dc) inverter output voltage using ADC.
3. Apply the following logic for terminal voltage control due to increasing load or discharging input battery voltage

if Vadc>=5
d=d-0.5
if Vadc<=4.78
d=d+0.5
else no change in duty cycle

Here,
230v is scaled as 5v (230/5=46, division factor 46)

When terminal voltage is 230v or above, Vadc=230/46=5v; in this case duty cycle will be decreased. Hence, terminal voltage will be decreased. So, terminal voltage will never cross 230v.
If terminal voltage goes down (due to increasing load or decreasing input voltage) to 220v, Vadc=220/46=4.78v; in this case duty cycle will be increased. Hence, terminal voltage will be increased.

So, as per my understanding terminal voltage will always be in between 220v to 230v. What do you think? Please provide your valuable suggestions.

[NB: I know there is a popular PWM IC (SG3525) to do the same. But I want to make it by myself as my intention is not only make it practically but also learn something new to increase my knowledge base.]

 

[goldsmith]

Hi sam
at first don't forget the out put voltage of your micro controller can't exceed from 5 volts . thus you should convert it to the 15 volts ! you can try an ICL7667 after your micro controller . or an opto coupler with internally totem pole .

[NB: I know there is a popular PWM IC (SG3525) to do the same. But I want to m

It can't give you SPWM . it is for SMPS design . you can use it for first section of inverter .

So, as per my understanding terminal voltage will always be in between 220v to 230v. What do you think? Please provide your valuable suggestions.

You should pay more attention to the peak voltage ! (325volts )

Good luck
Goldsmith

 

[sam781]

Hi sam
at first don't forget the out put voltage of your micro controller can't exceed from 5 volts . thus you should convert it to the 15 volts ! you can try an ICL7667 after your micro controller . or an opto coupler with internally totem pole .

Okay, thanks. I can use MOSFET driver IR2110 to do the same. right?

It can't give you SPWM . it is for SMPS design . you can use it for first section of inverter .

I want to use microcontroller for both the sections (converter & inverter). Here I was describing the converter section taking feedback from the inverter output. i.e the terminal voltage to be connected to the loads. In this case I considered rms voltage.

You should pay more attention to the peak voltage ! (325volts )

Peak voltage will reach 400v and higher. This peak voltage needs to be considered to chose semiconductor devices (e.g, MOSFET). Right?

 

[goldsmith]

Okay, thanks. I can use MOSFET driver IR2110 to do the same. right?

No ! a shifter needed between these sections ! ( between micro and main driver ! ) 5 to the 15 !

I want to use microcontroller for both the sections (converter & inverter). Here I was describing the converter section taking feedback from the inverter output. i.e the terminal voltage to be connected to the loads. In this case I considered rms voltage.

It is possible but it will be so harder than a PWM driver which has an internally oscillator and internally current limiting and internally dead time .

Peak voltage will reach 400v and higher. This peak voltage needs to be considered to chose semiconductor devices (e.g, MOSFET). Right?

No ! if your topology is push pull the stress across the DS will be 2vm ! 2*325 ! ( at least ) .