Hello all,

I am working on an isolated gate driver bias supply with a 4-output transformer. this transformer is VP4-0075-R and it has a weird pin configuration.



Here is my attempt to layout this transformer.. and the circuit is based on this schematic..



Is my layout good? if not what could be wrong with this layout?
BTW, I need to do it on one layer since the opposite side of the board has components.

any layout suggestions are welcome..

Regards,

 

The breakdown voltage for the transformer is 500V. It is not rated for power line primary to secondary voltage.
It looks like you have 320V from winding to winding. I don't know what the primary to secondary voltage is.
I worry about the voltage between pins on the PCB. Where the lines run between pins. See blue circle. I might use the other side of the board for the return path.
Look the bottom center of the picture. I moved a diode and capacitor to keep the long wires short. The wires from the transformer have very fast edges and emit noise. Long wires have inductance which you do not want. So put the diode up against the pin of the transformer and the cap next to it. Now the power is DC and you can have long wires. You will need more cap at the gate driver also. There is a loop of transformer-diode-cap-transformer, which must be as short as possible.

 

The breakdown voltage for the transformer is 500V. It is not rated for power line primary to secondary voltage.
It looks like you have 320V from winding to winding. I don't know what the primary to secondary voltage is.
I worry about the voltage between pins on the PCB. Where the lines run between pins.

This power supply is used yo generate +15V and -4V bias rails for the gate drivers .. the transformer is not connected to the 320V in the schematic.

 

the transformer is not connected to the 320V in the schematic.

Yes it is. Look at the top IGBTs. When any one is on the transformer is connected to +320.
Look at the bottom IGBTs. When any one is on the transformer is connected to 0V.
There will be 320V between windings.

 

You need ceramic capacitors where I put blue lines.

 

The breakdown voltage for the transformer is 500V. It is not rated for power line primary to secondary voltage.
It looks like you have 320V from winding to winding. I don't know what the primary to secondary voltage is.
I worry about the voltage between pins on the PCB. Where the lines run between pins. See blue circle. I might use the other side of the board for the return path.
Look the bottom center of the picture. I moved a diode and capacitor to keep the long wires short. The wires from the transformer have very fast edges and emit noise. Long wires have inductance which you do not want. So put the diode up against the pin of the transformer and the cap next to it. Now the power is DC and you can have long wires. You will need more cap at the gate driver also. There is a loop of transformer-diode-cap-transformer, which must be as short as possible.

View attachment 327731

Yes, i understand now ...
I will put the rectifier diode close to the transformer pins ..and i will try to use return traces and the bottom layer if possible to keep the traces inductances low..and use capacitors close to transformer pins and also close to the gate driver ..
I am using a zener divider to produce +15v and -4v rails with shared ground ... I am doing this for the 4 windings ..is this gonna make problems? Knowing that each gate driver driving a mosfet have it's own power rail ...

 

I think you are OK. Can you share the schematic? It is a good idea to have other people look over your design.

 

I think you are OK. Can you share the schematic? It is a good idea to have other people look over your design.

sorry, didn't have access to my laptop .. here is the schematics.

 

When the transformer makes too much current, the TL431 turns on to load down the transformer. TL431 current will flow through 100 ohm and 3.9 Zener. I think the 100 ohms will have too much voltage across it. I think it should be 0 ohms. I would short 100 ohms and open 1k ohms.

 

When the transformer makes too much current, the TL431 turns on to load down the transformer. TL431 current will flow through 100 ohm and 3.9 Zener. I think the 100 ohms will have too much voltage across it. I think it should be 0 ohms. I would short 100 ohms and open 1k ohms.
View attachment 327830

Yes, you are right about the 100 ohm res ..
what 1k ohm resistor that should be open ?

 

what 1k ohm resistor that should be open ?

2k, sorry.
I don't see its purpose.
I think your 15V does not need to be regulated by a TL431. I know it is a good part. Why not a 5% Zener?

 

2k, sorry.
I don't see its purpose.
I think your 15V does not need to be regulated by a TL431. I know it is a good part. Why not a 5% Zener?

As far as I remember, I simulated the circuit on LTspice and I had the 15V rail increased up to 18V sometimes even more..that's why I used the TL431 to stabilize the 15V rail .. but I might remove it in the actual circuit and keep the 3.9V zener or the other way around.
I attached the simulation file if you want to look at it.

 

data sheet Depends on the diode.