Full Bridge Inverter from bits and pieces

[Thunderstruck]

Originally I wanted to build Profdc9’s full bridge inverter, I also wanted to build it using parts listed in the BOM, I even filled a shopping cart at Mouser with all the necessary parts - about $120 worth...

Then the power and gas bills came together... 

So, because I have to act as a responsible adult (BORING), hobby has to wait for funding to be available again.

Then I remebered that I have some bits n pieces left over from my first SSTC attempt that might work together to make a functioning bridge using Profdc9’s full bridge PCB.

I have:
- 4 IRFP460 FETs
- 4 MUR860 Fast Diodes
- 4 WIMA MKP10 0.68uF 630V Polypropylene Caps

I think that there is enough parts here to make a circuit based on Steve Ward’s mini SSTC half bridge inverter, but full bridge instead and using above mentioned PCB.
Maybe my thinking is wrong, but to me it looks like just a matter of doubling the number of the components, and of course driving the FETS in the correct sequence.

Drive would be provided by Profdc9’s SSTC Driver.

So far I mounted the parts together, not exacly as one would expect, but if everything works well I reckon it will make a neater build.

I cannot figure a couple things, one is due to inexperience in interpreting component data sheets:

1. What is the maximum Gate voltage on IRFP460 - I gather its about 10V but not sure
2. GDT from the driver outputs about 34V peak to peak, which is way too high for IRFP460,   
    how do you bring the voltage down so it can be handled by the gate ? Only way I can think of is
    changing primary to secondary winding ratio.

I’d appreciate if someone could help with this - even if you tell me that this contraption will never work the way I imagined.

Image 1 Snip from Steve Ward’s schematic that i want to duplicate in a full bridge configuration
Image 2 Parts on hand
Image 3 Build so far - TVS diodes are bent like that because I wanted to put them under the board, but later realised it wouldn’t work 
Image 4 This way FETS get a huge heatsink that can later serve as a base for the coil - of course keeping in mind that the transistors must be isolated from the heatsink, otherwise there will be a lot of angry Pixies in that sink, waiting to bite.

[Fumeaux]

I cannot figure a couple things, one is due to inexperience in interpreting component data sheets:

1. What is the maximum Gate voltage on IRFP460 - I gather its about 10V but not sure
2. GDT from the driver outputs about 34V peak to peak, which is way too high for IRFP460, how do you bring the voltage down so it can be handled by the gate ? Only way I can think of is changing primary to secondary winding ratio.

1) The gatevoltage for the IRFP460 is +-20V. (+- mean you can put everything between 20V and -20V on the gate)
91237.pdf
2) 34Vpp (Peak to Peak) means 17Vp (one Peak) or +-17V which is smaller than +-20V. So you are fine as it is.


You could bring it down either by lowering the secondary turns or lowering the gdt driver voltage (which may or may not be possible).

[Thunderstruck]

I cannot figure a couple things, one is due to inexperience in interpreting component data sheets:

1. What is the maximum Gate voltage on IRFP460 - I gather its about 10V but not sure
2. GDT from the driver outputs about 34V peak to peak, which is way too high for IRFP460, how do you bring the voltage down so it can be handled by the gate ? Only way I can think of is changing primary to secondary winding ratio.

1) The gatevoltage for the IRFP460 is +-20V. (+- mean you can put everything between 20V and -20V on the gate)
91237.pdf
2) 34Vpp (Peak to Peak) means 17Vp (one Peak) or +-17V which is smaller than +-20V. So you are fine as it is.

You could bring it down either by lowering the secondary turns or lowering the gdt driver voltage (which may or may not be possible).

Two birds with one stone ! Thanks Fumeaux !
So, Vgs is what I needed... well I suppose now I know 

I tried lowering GDT driver voltage - it doesn’t like it, gets emotional very quickly
But, GDT output voltage is not an issue any more.

[Laci]

MUR860s are not TVS diodes,but ultrafast recovery and works as freewheeling diodes.

[Thunderstruck]

MUR860s are not TVS diodes,but ultrafast recovery and works as freewheeling diodes.

Yes, that... what you said 

But they deal with the same thing, right ? Feedback from an inductive load ?

[Laci]

TVS diodes are more likely used in high inductance bridge design,where high voltages are generated due to parasitic inductances,which they clamp to a safe level.

[Fumeaux]

MUR860s are not TVS diodes,but ultrafast recovery and works as freewheeling diodes.

Yes, that... what you said 

But they deal with the same thing, right ? Feedback from an inductive load ?

Point is: they are not ideal, but I do think that they are good enough for a SSTC. TVS diodes are essential for DRSSTCs or I guess when you want a 3kW SSTC, but most schematics online for normal SSTCs don't feature TVS diodes.

[Teravolt]

the diode conducts only one direction and a tvs conducts in two directions like a big Zenar diode. The reason for them is to protect the internal diode in the mosfet during times of hard switching. If your switching ZCS timing is in phase and or on time they will never conduct.