Boost Converter: earth ground problem

[Bdrt]

Boost Converter : Current oscilation Problems

Hi,

I'm using a Boost converter, at a 15kHz Switching frequency. I'm using a coupled inductor with a value of 2X1600uH.
I'm using a IGBT module where i am switching the bottom IGBT and the top IGBT is used as diode ( the reverse protection Diode )
The problem is that i'm having oscilations in the current during the switches. I attached a image of it.



Can this be solved with a RC snubber placed across the IGBT? The Vce wave form doesnt show any overshoot.
Can anyone help me understande what could be the problem and how can i solve it ?

Thanks.

Best Regards
Bruno

 

[KlausST]

Re: Boost Converter : Current oscilation Problems

Hi,

hard to say. (I hope this is not from a breadboard circuit)
* maybe this is just a measurement problem. And the ringing doesn´t exist in your circuit.
* maybe it is from (stray) inductance combined with some capacitance creating a resonance.

****
For further discussion we need to see your schematic, PCB layout, measurement connections (scope).

Does it cause any problem? Or why do you want to suppress it?

Klaus

 

[Bdrt]

Re: Boost Converter : Current oscilation Problems

Hi,

hard to say. (I hope this is not from a breadboard circuit)
* maybe this is just a measurement problem. And the ringing doesn´t exist in your circuit.
* maybe it is from (stray) inductance combined with some capacitance creating a resonance.

****
For further discussion we need to see your schematic, PCB layout, measurement connections (scope).

Does it cause any problem? Or why do you want to suppress it?

Klaus

Dear Klaus,

Atached is a schematic ( sorry i dont have it properly designed ) and a image of the assembly of the componentes.
The Current in the previous picture was the ouput current. It was measure with a current probe away from the noise sources.

**broken link removed**

**broken link removed**

I'm using this converter to charge a battery pack, so im trying to improve it as i can.

Any ideas about what could be the problem or sugestions about how can i improve / correct it ?

Thanks

 

[FvM]

Re: Boost Converter : Current oscilation Problems

Why are you connecting the inductor windings like you do? The circuit will generate huge common mode swing between input and output ground, might be one reason for the oscillations. LC circuit would be formed by inductor leakage and external input-output capacitance.

Where are measuring the current waveform? There's apparently no pwm current, only ringing DC.

I guess the oscillation don't show with significant magnitude in the DC bus voltage.

Try to reconnect the inductor with both windings in series and input/output ground shorted.

 

[Bdrt]

Re: Boost Converter : Current oscilation Problems

Dear FvM,

I was trying to follow a design i had that takes advantage of the coupled inductor to reduce the riple.
So, from what i understood, you think that this oscilations can be caused by the coupled inductor ?
The current wave form was obtained in the load , while charging the batteries. It doesnt affect much the Bus voltage because the batteries absorve the oscilation.

Thanks

 

[FvM]

Re: Boost Converter : Current oscilation Problems

Reducing ripple by the present inductor circuit seems to me as an error of reasoning. You get the same boost inductance by series connecting both windings, but avoid the common mode problem.

I guess the oscillating Iout current is pure common mode, it's not realistic that a differential current bypasses the 25 µF capacitor.

 

[Bdrt]

Re: Boost Converter : Current oscilation Problems

Reducing ripple by the present inductor circuit seems to me as an error of reasoning. You get the same boost inductance by series connecting both windings, but avoid the common mode problem.

I guess the oscillating Iout current is pure common mode, it's not realistic that a differential current bypasses the 25 µF capacitor.

I did as you said ( but instead of connecting the inductance in series i connected in parallel because of the higher currents ) and it seems to improve the result.





The larger current spikes i think that can be noise from the converter. I'll try to increase the output capacitors to see if it helps.

I dont know if i should create another topic or if i should continue on this because i think its related :
I had a earth ground problem (not sure if this the correct term ).
I connected the shield of my cables to a single earth ground point. The problem was when i connected the shield cable from the drivers, the communication with my controler crashed, and eventualy, while i tryed to understand what was happening the driver burned.

Could it be related to the diferent grounds on the input / Output of the converter due to the inductor ?

Thank you so much for your help.

Best Regards.
Bruno

 

[Bdrt]

Boost Converter : earth ground problem

Hello,

I'm having a problem that somehow is related to a earth ground problem.

In my boost converter, using a half bridge IGBT module, when i connected the shield from multiple cables ( adc signals, pwm ) to the same ground point, i would lose the comunication with my microcontroller or worst ( it blew a of IGBT driver and IGBT Module ).

I'm trying to understand what could be the reason for that and how can i solve it or at least correct it. Can someone help me with this ?

Thanks

Best Regards,
Bruno

 

[FvM]

Re: Boost Converter : earth ground problem

Unfortunately the attachments in post #3 seem to be lost on the server, can you reload it?

Roughly remembering the experimental setup in the photo, I'm not too surprised about the problems you are reporting now.

 

[Bdrt]

Re: Boost Converter : earth ground problem

Hi,

I've already done some modifications since last time, due to the ressonance.
Atached is a picture of the setup and the diagram.





If you have any tips or ideas about how can i solve the those problem please share =)

Thank you.

Best regards

 

[andre_luis]

Re: Boost Converter : earth ground problem

The aerial wiring arrangement don't helps much on preventing any kind of induced EM interference. Although you took some precautions, such as twisting the driver wire pairs and grounding some components, the spread disposal of the 'modules' turns everything acting as antennas. Anyway, keeping on that layout, you could try adding several polyester capacitors close to pins of some critical devices, such as bulk capacitors at the input and output side, to see if it can decouple at least part of the ring oscillation.

 

[FvM]

Re: Boost Converter : earth ground problem

The grounding scheme isn't completely clear. The photo shows that the IGBT gate cables have a grounded shield, but the power ground seems to be floating. Doesn't fit well together.

 

[Bdrt]

Re: Boost Converter : earth ground problem

The aerial wiring arrangement don't helps much on preventing any kind of induced EM interference. Although you took some precautions, such as twisting the driver wire pairs and grounding some components, the spread disposal of the 'modules' turns everything acting as antennas. Anyway, keeping on that layout, you could try adding several polyester capacitors close to pins of some critical devices, such as bulk capacitors at the input and output side, to see if it can decouple at least part of the ring oscillation.

thanks for the tips.
I've already placed a low ESR capacitor at the output ( in parallel with the ouput eletrolitic capacitor ) and ill place another near the input.
Changing the configuration of the inductor already helped much reducing the ressonance. My main problem now is the EMI wich i think are part of the reason why i have some instability in my ADC.

The grounding scheme isn't completely clear. The photo shows that the IGBT gate cables have a grounded shield, but the power ground seems to be floating. Doesn't fit well together.

That's because when i connected the shield from the driver to my earth ground point i lost the communication with my microcontroler and other time my IGBT driver burned.

Someone told me that between the IGBT and the heatsink it would be formed a parasitc capacitor and since the heatsink and the switching of the IGBT would induce currents there. A solution to that would be place a Y2 capacitor, near the IGBT, between the GND pin of the IGBT and the heatsink, in order to allow a current to pass. Does this makes sense ? Can that help with the problem ?

Thanks

 

[andre_luis]

Re: Boost Converter : earth ground problem

I've already placed a low ESR capacitor at the output ( in parallel with the ouput eletrolitic capacitor ) and ill place another near the input

How are you taking the signals shown in the oscilloscope picture shown above ? I mean, is the scope reference of the probe near to the signal which is being acquired ?

 

[Bdrt]

Re: Boost Converter : earth ground problem

How are you taking the signals shown in the oscilloscope picture shown above ? I mean, is the scope reference of the probe near to the signal which is being acquired ?

That current waveform was taken with a current probe . Those spikes i have them in the osciloscope even without the probe in place so it is noise from the switching.
The proble was placed on the output ( blue wire ) of the setup.

Thanks.