500VA sine wave inverter , problem in HBridge and filter designing

Status
Not open for further replies.
i want to learn dspic30f , to generate unipolar spwm, , plz send me any link of pdf or ebook, ?
 

FvM , can u tell me any dspic30, which can be use to generate complimentary unipolar spwm , with software deadtime.??
 

Please examine first your signal from your micro controller output before you proceed setting up your output waveform. If you confirmed that your input signal drives were almost perfect then its time to focus on the output side. There may be a problem still from your signal drives. Do the best way to correct first from your inputs.
 

i am now learning DSPIC33 , to generate UNIPOLAR SPWM, as soon as i complete my design,i'll post my schematics..
 

Maybe try to use IR2104. It drive 2 MOSFETs (Low and High) alternatively, using only one input signal. So it giving synchronous PWM without unipolar output from the microcontroller. IR2104 have internal deadtime generator of course. I used it in my 1-phase ACIM inverter and it works fine.
 
Last edited:


FvM said:
In an inverter, the regular bus capacitor won't have problems to absorb the energy. But at least part of the capcitance must be placed near the output stage, otherwise current transients could cause voltage spikes at the bus.
Click to expand...
FvM, i have some confusion,,if the capacitor near the H-BRIDGE is having high ESR ,will it create problem?
what is the acceptable range of the ESR of capacitor that should be placed near the H-BRIDGE?
 

You mean the bus bypass capacitor? It's ESR is primarly a problem of current rating, I think
 

FvM said:
You mean the bus bypass capacitor? It's ESR is primarly a problem of current rating, I think
Click to expand...
ya , i am talking about the capacitor that should be placed on DC bus. i havn't measured the ESR of those capacitor, so do you think that high ESR value electrolytic capacitor was creating the problem in my design ?

and one more thing i want to ask " in my earlier design i was keeping high side mosfet of one leg permanently ON for entire 10ms duration meanwhile switching the diagonally opposite low side mosfet with SPWM. and vice versa for other daigonal pair. now i am going to to give a notion of another scheme in which for positive cycle,HIGH side mosfet of LEFT leg is kept ON for entire 10ms and opposite LOW side mosfet of RIGHT leg is SPWMed , and for negative cycle , LOW side mosfet of LEFT leg is kept ON for entire 10ms duration while HIGH side mosfet of RIGHT leg is SPWMed . i've found this scheme in an application note.
will the above scheme has any advantage over the the previous scheme that i was using?
 

I thought, others can well contribute to the thread.

I don't expect that ESR is a particular problem as long as the capacitors don't run hot. Circuit inductance may be a problem.

Essentially both schemes have the same output characteristic.
 
FvM, u had said in my earlier thread that the scheme that i am using is practical. If proper layout is made. U had also said that "try to use any synchronous scheme". Now i am working on Dspic to generate unipolar spwm.but i also want to recognize the fault in my earlier designs.
I have two questions
1) what are the pcb layout design tips for a smps based inverter?
2)which scheme should be used to remove the maximum stress on the body diode of mosfet?
 

You know that others and I couldn't but guess about the problems of the original design.

One possible explanation why the design apparently fails with inductive load refers to slow reverse recovery of body diodes and related current peaks, possible inductive overvoltage and false switching triggered by the transients in the control circuit.

It has also been said that the best way to find out about the design problems is to operate it with a variable, current limited power supply and slowly approaching the critical operation conditions, thoroughly monitoring the voltage and current waveforms. You proabably don't have the equipment to perform this kind of safe live tests.

Several working inverter projects have been published at edaboard through the years, e.g. in this serial thread https://www.edaboard.com/threads/96053/
 

FvM, in my previous designs i was switching the low side mosfets with SPWM while keeping the opposite high side mosfet ON for entire 10ms duration. Now when the low side turns OFF, the freewheeling current is injected into dc bus through the freewheeling diode of the high sidr mosfet of same side. Now i have a confusion ,if i consider a case in which i switch the high side mosfet by SPWM and keep the opposite low side mosfet fully ON for 10ms ,similarly for the other pair. So ,is the freewheeling current in this case is also injected to dc bus??
 

FvM, suppose
S1 is left high side switch
S2 is left low side switch
S3 is right high side switch
S4 is right low side switch .

Now take a case when S1 is driven with spwm and S4 is turned ON for 10ms , as S1 goes OFF the inductive current will still force to flow in same direction since it can't pass from S2 , it has two ways Ist is to continue flowing through the S4 and go to ground and 2nd is to flow through the freewheeling diode of S3. So don't you think that the amount of current flowing through the freewheeling diode in this scheme is very less as compared to earlier scheme in which all the freewheeling current was flowing from freewheeling diode of S3 and S1.??
 

The current will always commutate between high and low side switch. Why do you think that the current doesn't flow through the S2 diode?

You may want to setup a simulation.

As already said, the only difference between both schemes is that you flipped high and low side switches. How should this variation change principle behaviour?

A basic limitation of asynchronous H-bridge operation applies to both schemes, they achieve the intended duty cycle only in first and third quadrant, not during recuperation (motor working as generator) and with reactive load.
 
FvM
My circuit worked successfully. I designed the new layout . At first i tested the H-bridge with a series 220v 100watt bulb connected to DC bus. I also connected the filter and the a 60watt load connected parallel to filter capacitor.. the 60 watt bulb glowed normally .then i removed the 100 watt bulb and tested the circuit again, the 60w bulb glowed normally again but the ceramic capacitor on the DC bus exploded. I didn't know the reason . I replaced the ceramic capacitor with a new one and tested the circuit for 10 minutes with 60watt bulb at the output. Suddenly after 10 minutes of testing the ceramic capacitor on the DC bus again exploded.. what is the reason behind this and how can i overcome this problem?
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…