Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Problems with my inverter at high voltage... can't go higher than 3.3v!!

Status
Not open for further replies.
Y

yokahu

Newbie level 6
Joined
Jan 25, 2010
Messages
14
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,400
Hi,
First of all, this is my first post here, but i've been reading these forums for quite some time now.

I'm doing a 9 level Flying Capacitor Multilevel Inverter (50Hz) and the topology for my schematic is:

micro (propeller) (3.3v) ----> optocoupler (5v) ----> ir2110 (Vcc=12v) (Vdd=5v) ----> irfp450 (x16) -- (16 caps from 10mF to 47mF 35v)

general info on multilevels: Discussions

The inverter works fine (see output waveform attachment below) until I increase the voltage from my power supply above 3.3votls, after that the output voltage doesn't increase anymore, it stays at ~3.3v with a poor waveform.

output at ~3.0v: View attachment tek00008.bmp

BTW, for the ir2110 im using bootstrap capacitor of 2.2uF

Any ideas why my inverter doesn't work at higher voltages?

Thanks!
 

Full schematic showing all the gate drive interconnects needed to answer your question. Regards Orson Cart.
 
Hi,

this is the schematic for the gate driver circuit. I didn't have access to PSpice or Cadence so i drew it.

Hope this helps and thanks for the replies!

IMG_0829[1].JPG
 

Why are you stacking FETs in the H-bridge? Could that be making
problems such as nowhere for bootstrap current charging to return,
when the center point of the leg is high-Z?
 
Well, for the stacked FET it is part of the topology.

The problem must be around the ir2110 and the bootstrap cap (maybe?)...

i connected the COM pin to the source of the low side FETs and the Vs pin to the source of the high side FETs because i wasn't sure where to connect them since in the datasheet of the ir2110 they don't have stacking FETs and the source of the high side is the load and the source of the low side is ground.

maybe i should connect all the Vs pins to the load and all the COM pins to the ground instead of the sources...

any ideas?
 

VS has to be tied to the individual FET source and VB has to fly with it.
But I have not seen the 2110 applied in this stacked way. COM I
expect is going to be ground for all of them.

You might elect to start with the simpler H-bridge form and see
if that works, as a waypoint.

You should be seeing full-swing (~12V low side, ~11V high side)
on all gates regardless of V+ level, and they ought to have the
proper timing / phase relation. Might be good to verify that, and
see if changing V+ does something to that aspect around the point
where things go bad. Like, maybe your V+ supply is not stiff
enough for some startup current (maybe, to do with gate drive
or timing, shoot-through / cross conduction or just too heavy
a load).
 
  • Like
Reactions: yokahu

    Y

    yokahu

    Points: 2
    Helpful Answer Positive Rating
Looking at the topology, I don't think bootstrapped gate drivers will work out. In order to maintain the bootstrap voltage, the source voltage of each switch (Vs on the driver) must periodically drop near Vss. I don't think you can satisfy the condition for the upper switches, at least not while maintaining the desired output waveform.

Also what you're doing with the COM pins definitely won't work. The COM and Vss pins should be at the same potential always.
 
  • Like
Reactions: yokahu

    Y

    yokahu

    Points: 2
    Helpful Answer Positive Rating
With a multi-lever converter you will need opto gate drivers giong to each FET and an isolated 100kHz square wave supply going to each fet source that can be rectified for the local gate drive power - this is due to the long on times you get on some devices.
A central simple square wave switcher 100kHz (H bridge at 49% duty, DC blocking cap) from a 15V DC supply on a suitable ferrite core, can be used to provide 16 secondaries for the 16 fets, ideally double or triple ins wire for the sec's - phasing not important. Regards, Orson Cart.
 
OK,

so as far as i understand, the gate driver ir2110 is the source of my problem and i'm going to need one gate driver for each FET to make it work?

if that is the case, can I use the low side of the ir2110 for each FET? I'm still trying to find a solution using ir2110 as the gate drivers since I've read papers in the IEEE from people that have used ir2110 on multilevels but they don't explain how :/

What u mentioned, Orson Cart, sounds like a solution but I'm not that advance in Power Electronics and i didn't understand a few things or how to implement it.

More advises are welcome! thx
 

IR2110 can be made to work - but you still need the isolated 12-15V supplies going to all the fets not connected to the 0V rail, also you can only use the high side part of the 2110 for the upper devices (apart from the two on the 0V rail - you can use the low side part for the 2 low side fets - i.e. those connected to the 0V rail) - Regards, Orson Cart.
 
In a short, the IR2110 (an similar device's) low side driver can only work for a bottom switch (source connected to GND respectively Vbus-).

The high side driver with bootstrap circuit can only work for switches, that have their source node periodically returned to GND respectively Vbus-. I your multilevel cicruit, this may be the case for several devices, but probably not guaranteed under all operation conditions. Bootstrap time constants and duty cacles have to be considered.

A high side driver with auxilary voltage supply (not relying on boostrap operation) can be used for any switch in the multilevel converter.
 
  • Like
Reactions: yokahu

    Y

    yokahu

    Points: 2
    Helpful Answer Positive Rating
in a multilevel converter the switches can be ON or OFF for long periods - so bootstrapped gate drive supplies generally do not work unless the switching regime is modified to allow bootstrap cap charging (messy, complex and makes converter less efficient) - quite often direct transformer drive is used (500kHz - 1MHz) where the high freq AC is rectified and applied directly to the gate for ON and removed for OFF. Regards, Orson Cart.
 
OK,

I now understand why the ir2110 is not working for my multilevel inverter, because the high side fet's sources are not returning to ground freuqnetly and the bootstrap capacitor doesn't do its work right.

Now, please help me explore and understand the auxiliary power supply option...
do i need a new gate driver (to replace the ir2110) or the isolated supply can be integrated to the ir2110?

Thanks for the replies and patience!
 

Hi,
That driver looks great and simple enough for me :)

So, can i use that driver for my low side FETs too?!?

is there a thru-hole version of the driver? looks really small
 
Last edited:

With the FOD3150 gate driver i can drive my 16 FETs (high and low) without the use of an isolated supply?
 

  • Like
Reactions: yokahu

    Y

    yokahu

    Points: 2
    Helpful Answer Positive Rating
Thank you all for the replies,

Now that i know that the individual isolated supply for each FET is a must...

i need help with a simple design of an isolated supply for my FETs to finally finish my little project :)
btw i need the components to be though-hole since i'm doing my control circuit on a bread board,

Thanks!
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top