[SOLVED] Diffrance between these two topology(going to use in induction heater)

Hello all members



i have posted this question twice but not get proper answer i hope this time i will get.:-?

i have uploaded two diagrams i am actually making induction heater.


CAN YOU TELL ME WHICH IS TECHNICALLY BEST FOR MY APPLICATION (A) OR (B). And why?



i also want to know which type of capacitors(C1,C2) i have to use in (B).Electrolytic,ceramic ???? and how much value of capacitors (uF)


i tried with 470uF electrolytic 400V but they are getting very hot.


LOOKING FOR NICE VALUABLE ANSWERS FROM YOU EXPERTS.





-THANK YOU IN ADVANCE

Hi,

I think circuit (A) can not work, because the primary side will see DC. To make it work a large capacitor in series may help. Otherwise the current may increase in a way that the transformer or the switch dies.

With circuit (B) the capacitors prevent from DC. What capacitor value to use depend on switching frequency and the primary impedance of the transformer (including load).

I'd calculate the capacitors value so that the AC across the capacitors is less than 20% of the transformers AC.
There are a lot of values missing, so we can't calculate ...

What type of capacitor depends on value, voltage rating, AC current and switching frequency.
A standard electrolytic bulk capacitor has too much ESR, that causes the heating.
Dissipated power is : P = I x I x ESR.

Btw. You work with rectangle signals with high dU/dt. This causes large current spikes in the capacitor parallel to the load.
Maybe you have to disconnect them.
(They are good with sine waveform to compensate the inductivities, or to build a resonance circuit)

Hope this helps.
Klaus

The problem with Electrolytic and Ceramic at high voltage is the leakage resistance or Dissipation Factor or Loss Tangent.

You must choose a thick film Plastic Cap such as Polyester or Polyurethane. (Epcos, Panasonic)

Induction heaters are usually fed by an inductor to a center tapped primary to avoid the DC current saturation effects when using line frequency. Another method is to use 100KHz -150kHz resonant frequency which is also ideal for Aluminum pots. But different coil using Litz wire and value of cap.is required.

Thanks for your reply


I am going with design B. I have searched many designs on internet having value of C1 and C2 of 470nF to 1uF .I have tried that all but low value of caps. Not giving output . I am at 25 KHz .primary has 25 turns.if I replace those caps with higher value machine works nice. Can you suggest me any values for C1 and C2?

F=29KHz
NP=20
Vsupply=250V

- - - Updated - - -

Again thanks for your valuable reply .thank you very much.


And ya capacitor at load side is the part of parallel resonance circuit .

- - - Updated - - -

Actually I am very confused in selection of value of this capacitors. Is it ikey if I put 5 uF?....

Unfortunately, you need to measure L and resonant loop cap C for optimal frequency and then compute impedance at resonance of either element L or C and then use C1,C2 as a virtual ground which means the impedance at resonance of the coil must be much higher than impedance of C1,C2. Then unfortunately large electrolytics have high leakage current > 1mA and high ESR which limits ripple current ( some can handle >1 Amp and are rated according to your design.

C1, C2 cannot be determined without more details such as current, L, f, Z.. This "nomograph" may be useful The intersection of any 3 known values will give the 4th unknown for F,L,C,Z(Ohms)

How does your Cap that overheated compare with these ? https://www.digikey.com/product-detail/en/LGW2G471MELB50/493-3231-ND/1966760
What part numbers did you use? Did you use Litz wire? What is the coil self resonant frequency?

Thanks for Reply

data of all parts are mentioned below

IGBT : 25N120
Frequency(Resonance): 29KHz
I am not using litz wire instead of it i am using copper strip as primary
core of transformer is COSMO ferrite Model 110


by the way i am thinking to try various values because i have some caps here.So i will post result later....



-Regards

- - - Updated - - -

and yes my cap. are exactly similar as shown on digikey

finally problem is solved.I tried some capacitors then i open and observer one industrial heater .they are uses 4uF C1 & C2 in the parallel connection of 2uF. having voltage rating of 1KV DC PP type.

what switching rate did you choose? IS that close to resonance?

Also what power can you generate with your final choices? and what efficiency do you measure?

hi

sorry for late reply at this time my exams are started so I will tell you all details after examinations ...I am thinking to go with 10Kw power .switching is almost at resonance as I am making induction heater



- regards

- - - Updated - - -

I will upload all photos of my circuit but it will take long about a month as u know examsss

Again failed .. those c1 and c2 sucks....
I used 2 uf as c1&c2.. but when I increase length of wires of secondary I must change frequency because current decreases on old frequency....

Anybody give me solution of these caps.... I have seen many sites showing such ckt. But nobody mention about selecting values of it.....


It is very good if u help me on it.......

- - - Updated - - -

I also tried to put series bulk capacitor but same result ... if I change F then current decrease from 10A to 4A .....

Hi,

give us complete information.

* What are the exact types of capacitors you used (datasheet or manufacturer and article number)
* What wires did you change in length?
* Why do you change frequency?

Maybe becauseo of changed cable length?
When you change length of cable, than you change R and L of cable resulting in another resonant frequency.


Klaus

I am using polypropelyne capacitors rated at 1kv.
I am using litzwire to eliminate asking effect.
I am changing frequency because according to change size of workpiece inductance of coil changes in induction heater chance resonance frequency also changes.....



I have seen many designs as shown in diagram B in pic. On net .. but how particularly select value of c no body mentioned.

Hi,

in post #2 i´ve given you a hint on how to calculate the capacitors.... but as long as you don´t give us voltage, currents and other tchnical information one can not calculate anything.

For calculating a C you have to know Xc and frequency
to calculate Xc you need to know current, voltage

it´s all not that difficult mathematics..
We can´t calculate, because we don´t have the necessary values.
But you surely have a DVM and a scope a can do some measurements...


Klaus

Hello KlausST

I really appreciate your old advice.but i dnt have LC meter now so can't measure inductance of winding.Actually my problems are listed below.


(1) If i tune to one perticular frequency i get high o/p current(secondary is shorted with one wire .) but if i change frequency current drops.
(2) if i change length of seconadary wires means if i connect long wires to o/p of trnsformer current drops simillarly as above.


i am also thinking to use this circuit in my welding inverter but as due to problem (2) i cnt use it....


i want following:

(1) no change in current with respect to frequency.
(2) no change in corrent wrt change in length of wires connected to secondary.


is it possible with this topology or i need to move at full bridge...




-thaks for your support all

Hi,

(1) If i tune to one perticular frequency i get high o/p current(secondary is shorted with one wire .) but if i change frequency current drops.

Click to expand...

This is typical for a resonant system. You change input frequency so you are out of resonant.

(2) if i change length of seconadary wires means if i connect long wires to o/p of trnsformer current drops simillarly as above.

Click to expand...

This is typical for a resonant system. You change circuit resonant frequency so you are out of resonant.

i want following:

(1) no change in current with respect to frequency.

Click to expand...

This is not possible with a resonant system.

(2) no change in corrent wrt change in length of wires connected to secondary.

Click to expand...

This is not possible with a resonant system.

is it possible with this topology or i need to move at full bridge...

Click to expand...

Even a full bridge doesn´t help as long as you want it to work as a resonant system.

Klaus

Ok ... i got it ... thanks.....


i will post results soon ....