Help on Inverter problem

[Ogu Reginald]

(TIP41 and TIP42) AND(BC547 and BC557), which one should I use? Should I connect resistors between the two base of the transistors?
If yes, should the resistors come before the two base are connected together or should the resistor be connected at the individual base after they have been connected together,

 

[patelhirenh]

Problem is nothing but you check out put voltage and HZ, if HZ is high then mosfet will brunt. and removed all mosfet and first try to only 1 mosfet per bank.

 

[pradeep.g.belchada]

use totem pole ouput use bc327 & bc37 pnp npn transistors

 

[Ogu Reginald]

I have soldered another power circuit but I am afraid of loosing the MOSFETs. Below is the circuit diagram of the power circuit which I soldered, Please I need corrections, suggestions and modifications where neccessary so that I can test the inverter. https://obrazki.elektroda.pl/5743287800_1380467058.jpg
and the thumbnail
**broken link removed**

 

[pradeep.g.belchada]

ckt is ok . **broken link removed** this link is not working

to safe guard mosfet use 12v 50watt bulb in series with the battery. at this time dont conect load check o/p voltage it should be 230v. if very high current flows from battery to inverter then again mosfet will blow, ir2110 gate driver ic is better solution

 

[Ogu Reginald]

If the thumbnail is not opening then check the main diagram. I just opened it now.

 

[E-design]

It is not a good idea to add the resistors you have on the gate driver. With those values, you will slow down and limit your gate drive to a maximum of about 4V. It is better to get the gate driver to switch as fast as possible with good levels and tailor the gate charge/discharge times with resistor and diode combinations at the gates itself. At 4V your MOSFET will not even turn on. You need at least 9V.

You don't need TIP 41/2 for the gate drive. They are big and slow and an overkill. Just use something like 2N2222/2N2907.

 

[Ogu Reginald]

Should I remove the resistors? Also what can I do to make the voltage at the gate of the MOSFETs be >10V.


Check this out:
https://obrazki.elektroda.pl/3244904800_1380545349.jpg

 

[E-design]

Yes, remove the resistors. You gain very little by having gate drive levels higher than 12V, since the device is already fully turned on (Rds-on at a minimum with 10V).

If you use the IRF gate driver, then follow the recommendations in the application note.

 

[Ogu Reginald]

I can't open the image you uploaded, please is it two diodes that I am seeing?

- - - Updated - - -

I have looked at the figures, from your waveform in figure 1, the amplitude of your pulse is approximately 10V but my main problem is that the voltage output of SG3524 is about 5V. I just re-measured my oscillator this morning to discover that the voltage at each output is 5V (please if it is abnormal do tell me). I have consulyed the datasheet of SG3524 to discover that the actual voltage output is 5V.
In this case what do I do to use it to trigger the gate of the MOSFETs.

- - - Updated - - -

I have looked at the figures, from your waveform in figure 1, the amplitude of your pulse is approximately 10V but my main problem is that the voltage output of SG3524 is about 5V. I just re-measured my oscillator this morning to discover that the voltage at each output is 5V (please if it is abnormal do tell me). I have consulyed the datasheet of SG3524 to discover that the actual voltage output is 5V.
In this case what do I do to use it to trigger the gate of the MOSFETs.

 

[ALERTLINKS]

the actual voltage output is 5V.

Multimeters only show average voltage. This is the reason you see a 0-11V 45% duty cycle pulse showig 5V on multimeter.

 

[--BawA--]

there can be two possible reasons.
1)you both mosfet banks might be ON at same time resulting in short circuit.
2)if your switching frequency is too high, mosfet's gate will require more charge to fully turn on.
solution
try to increase the deadtime
it is good to u use a dedicated mosfet driver like TC427.
also u can use a snubber circuit between the drains of your mosfet banks.

 

[E-design]

You should rather use the SG3525 as it is perfect for driving gates. https://www.onsemi.com/pub_link/Collateral/SG3525A-D.PDF

 

[Ogu Reginald]

I have tried all that have been suggested but I tested it this morning only to see the MOSFETs burn.
At first, I used a 6.3A fuse to connect the positive terminal of the battery to the center tap of the transformer and at switch ON the fuse got burnt.
I decided to remove the fuse and immediately I removed the fuse the MOSFET got burnt.
I decided to take resistance measurement at the terminals of the transformer and the transformer resistance gave 1.2 Ohm. I don't know whether the resistance readings are normal, should it be higher or lower?

 

[ALERTLINKS]

You should not have removed the fuse when there was an obvious fault. The current drawn will be around 500mA whe there is no load applied. It may rise max upto 4-Amps when there is a filter capacitor on output and the duty cycle is close to 50%. Another smart way for checking UPSes is to put two 100 watts car headlight bulbs in series with ups. You can check the response with load change by watching glow of bulbs. In short-circuit condition, maximum 20A will pass through fets which won't destroy them. When testing is over, even then use a fuse accoding to maximum allowed load (fuses used in cars are easily available for high current), check and observe its response at full load. Don't kill your FETs anymore. There should be o problem when load is fastly changed. Test with different ypes of load.
If you supply transformr with mains using a series lamp in between, there will be no glow in the lamp. If you short circuit output terminals of the transformer, it will fully glow it. If it is center tapped, check voltage on individual windings. They shoud match. Be assure, your transformer is good if any of the winding is not shorted with body.
To check your driver circuit with mutimeter, you will get around 5V on each drive pin output on DC range but if you check AC voltage between the two driver output You will get near 12V AC. Now after interchanging the meter leads, you should get the same AC voltage. An AC capacitor can also be used in series with one of meter lead to avoid DC content. This tests and indicate good anti-phase drive signal.
It may be rare but do you get genuine parts, or you may end-up thinking that some genie is not letting your gadget to work properly! :twisted:

 

[Ogu Reginald]

Concerning the resistance measurement at the terminals of the transformer, is it normal or abnormal for the resistance of a transformer to read 1.2 Ohm?

- - - Updated - - -

I just isolated my oscillator circuit and powered it for testing now and discovered that the 7812 regulator is heating up and I touched the SG3524 ic and also noticed that it is hot. As soon as i disconnect the power and removed the SG3524 ic and the regulator stopped heating up. What can be the cause?

 

[pradeep.g.belchada]

sg3524 may be short

- - - Updated - - -

or its ouput may be sinking

 

[Ogu Reginald]

I have tested the transformer with a 200watt bulb in series with the input of the transformer. The bulb glows small but as soon as I short the output, the bulb glow fully and with this I believe that the transformer is in good condition (please if I am wrong do correct me).
The inverter under discussion is designed to convert 48V DC to 220V. Based on the specification above, the transformer is a center tap (48V-0V-48V). In view of this I have been using four 12V batteries in series and the MOSFET bank has been burning. I just used two batteries (24V) and the MOSFETs didn’t burn again. But the output of the transformer now is 110V instead of 220V. Please can anyone tell me what the problem is and how to solve it?
Don’t forget that the inverter was working before.

 

[pradeep.g.belchada]

in series with the battery 48v (+ve wire of inverter to battery) conect 50watt 12v motorbike bulb. and give full 48v to the inverter. at this time dont conect the load . see if their is a 230v o/p. this method will safeguard the mosfet

 

[ALERTLINKS]

I found some issues with your design.
1) You did not drop voltage on SG3524 to 12V. 7812 voltage regulator can not work beyond 36V. A series transistor regulator ahead of 7812 will drop voltage to 24V for input of 7812 regulator . For example a setup like this will also provide supply for 12V relays (which are more common and cheap) and any other circuit such as status display.

2) Maximum operating voltage of IRF150 is 100V. 12V battery when fully charge can reach upto 14.4V. That is four batteries in series will have a potential of 57V. As this is a pushpull circuit, th voltage i doubled reching near 114V. When a load is attached another spike is also generatd worstening the situation. As there is a margin some FETs may work and some don't. You should put a snubber network to absorb the spikes also.
Replace them with FETs which can withsand higher voltage e.g. IRF250 which works upto 200V.
**broken link removed**
Initially use a single pair as early sugested. This will be good for upto 1000 watts of load. There will be less loss in case of damage.
TIP41 and TIP42 are ovekill. Small TO92 transistors are good in application.