I just took another look at the TI September, 2003 data sheet. (clipping below)


Notice the last sentence.

This is probably why @Alec_t made this point.

 

Well TI thinks it's an astable, with a divide-by-two output!

TI says it can be a monostable or astable. I undertood the datasheet to say the Oscillator pin has a 50% duty cycle but the Q output pin (which the TS is using) isn't guranteed to have 50%.

 

I would like to thank everybody who answered / replied to my post
I maybe haven't explained enough about my simple project
As the image shows , just a 24 v DC to 220 AC (square wave) Inverter without any modifications
I am trying to make it work at the current conditions then would try to modify/filter the output
*The transformer is EI stepdown 220 to 24 Volt , its power about 3000 VA
*Frequency used = 50-60 Hz

When I post my problem I needed your suggestions to get the best operating results and a good design
Now I am focusing on a one issue (I have asked above but looks like it ignored) which is :
Supposing I will use one of the following IC's (TL494 , SG3525 ) which are for sure much better than CD4047 at least in my case
my Question is : what is he best method to deliver the signal to a parallel (5 * IRFP150N) ?

 

It is an extremely simple squarewave inverter circuitthat is commonly used for outputs up to only 500W, not 3KW. You cannot easily filter its output.

The CD4047 uses a non-symmetrical RC oscillator that drives a digital divide-by-two that produces perfectly 50:50 duty cycle squarewave outputs.

 

Gate charge is 5 x 140nC. Output current of SG3525 is 200mA. Switching time would therefore be 3.5us. Not quick, but quick enough for a 50Hz inverter. So you could just use a 10Ω resistor in each gate circuit.

You can probably find a much better MOSFET, which would have a lot less gate charge for the same amount of Rds(on), or better Rds(on) for the same amount of gate charge.
IRFP150 is now quite old.

 

It is an extremely simple squarewave inverter circuitthat is commonly used for outputs up to only 500W

Good , So squarewave inverter's power limited to 500 W , I should build a sinewave inverter instead

Gate charge is 5 x 140nC. Output current of SG3525 is 200mA. Switching time would therefore be 3.5us. Not quick, but quick enough for a 50Hz inverter. So you could just use a 10Ω resistor in each gate circuit.

You can probably find a much better MOSFET, which would have a lot less gate charge for the same amount of Rds(on), or better Rds(on) for the same amount of gate charge.
IRFP150 is now quite old.

Its old indeed (a good design may easily get it to work very fine), its just in my hands at the moment

OK , what about using egs002 ?

 

produces perfectly 50:50 duty cycle

yes , but if you look the output latch setup , then you may notice that there is a skew in between non-inverting and inverting output ← which in turn means that the transition in between the TF's half-periods is "non-symmetrical"

 

Our electricity uses sinewaves and many new products rely on the sinewaves. What do you have that operates properly with squarewaves? Old-style incandescent light bulb/heaters?

 

I don't need squarewave for sure , I just started with simplest method to learn the basics
I'v asked above (what about using egs002 ?)
Its offer about 2.5 amp for driving FET's , Giving SPWM to produce a pure sinewave in the output of the transformer .

 

I never heard about Jingjing company and their EGS002 sinewave inverter module. Can you trust them?

 

The ICs appear to come from Yijing Microelectronics Co, www.egmicro.com
There are several sinewave inverter ASICS, of which EG8010 appears to be the most popular. There are EG8015 with built in high voltage drivers, and EG8030 which is 3-phase. The only place I have seen them for sale is on Aliexpress, which doesn’t fill me with confidence.
I would hazard a guess that they would be found in many Chinese-made inverters.

 

I have an egs002 already
Actually it's a module with alot of good features , nothing make me not to trust it .

 

(Some text removed for clarity)

*The transformer is EI stepdown 220 to 24 Volt , its power about 3000 VA
*Frequency used = 50-60 Hz

(Some text removed for clarity)
my Question is : what is he best method to deliver the signal to a parallel (5 * IRFP150N) ?

This might not be the right transformer for your application.

When a step-down transformer is used as intended the primary provides the magnetizing current (no load current) which is often substantial. Consequently the designer takes this into account and uses the necessary wire size for the primary so as to not have too much of a voltage drop in the primary. The secondary does not see any of the magnetizing current, only the load current.

When driving a step-down transformer "backwards" the secondary will have to supply a large magnetizing current through a winding that may, and in my experience did, have a high enough resistance to have a substantial voltage drop because of the magnetizing current. I

You may want to take some measurements or do some experiments to determine whether the increased voltage drop in the driven winding will be a problem with this transformer.

Another note: With the kinds of current I think you will be driving current limit/short circuit protection would be a good idea to minimize the damage should anything go wrong (and something will go wrong).

The attached white paper from Toshiba on the subject of paralleling MOSFETs might be of some interest. It is not always as easy as it might first appear. Take particular note of their recommendations for the gate resistors.

 

This might not be the right transformer for your application.

When a step-down transformer is used as intended the primary provides the magnetizing current (no load current) which is often substantial. Consequently the designer takes this into account and uses the necessary wire size for the primary so as to not have too much of a voltage drop in the primary. The secondary does not see any of the magnetizing current, only the load current.

When driving a step-down transformer "backwards" the secondary will have to supply a large magnetizing current through a winding that may, and in my experience did, have a high enough resistance to have a substantial voltage drop because of the magnetizing current. I

I am sure you are completely right , I'll take that seriously
BTW ,, Both of your attachments didn't open for me .

 

Here is the HTML article as far as it goes and the link to download the white paper directly from Toshiba's site. As originally posted there were two duplicate attachments, and that has been corrected.

https://toshiba.semicon-storage.com...innovationcentre/articles/tcm0286_mosfet.html

 

Here is the HTML article as far as it goes and the link to download the white paper directly from Toshiba's site. As originally posted there were two duplicate attachments, and that has been corrected.

https://toshiba.semicon-storage.com...innovationcentre/articles/tcm0286_mosfet.html

That's a really interesting article. Fig. 4 is intriguing - I understand that the current will eventually equalise as all devices are fully enhanced, but it surprised me that it should take as long as 4us.

 

Many thanks , I'v downloaded it .