Hi All.

I have designed a DC/DC converter for high voltage application, which I am now designing the PCB for. The only thing I am struggling to figure out is how best to connect the inverter and rectifier boards together using a transformer, which needs to be chassis mounted. I know the best way to do this is flying leads. But how do I connect these then, to the PCB board? Since the application is high voltage, 270VDC primary and multi-kV secondary (multiple stacked windings), do I need any kind of special connector for the flying leads? If anyone has any recommendations that they can find on Mouser/Farnell/DigiKey, please do link them to me, it would be much appreciated.

Thanks!

 

For 270VDC, most house type of wiring (or salvage from electrical cords) are rated for 240VAC and would do fine. For multi-kV wires, you could use CRT TV HV cables, commonly found online. Those are usually rated 20kV to 30kV.
Without more details as to the actual kV required, that is the best guess anyone can give. Besides, putting 20kV on a PCB would require real special design, not a common thing to do...

 

For 270VDC, most house type of wiring (or salvage from electrical cords) are rated for 240VAC and would do fine. For multi-kV wires, you could use CRT TV HV cables, commonly found online. Those are usually rated 20kV to 30kV.
Without more details as to the actual kV required, that is the best guess anyone can give. Besides, putting 20kV on a PCB would require real special design, not a common thing to do...

The two output voltages are 3 and 6kV, but the transformer generates 4 voltages of 750V each which are fed into voltage doublers. So this is what is creating confusion. I believe I probably want each of the connectors of the secondary rated at the final output voltage anyway in case of faults? Would welcome input on that also.

 

Also, I was wondering more about the actual connectors to the PCB that connect the chassis mounted transformer to the flying wires. I assume I need some kind of connector rather than just soldering the wires to the PCB which probbaly would cause many issues..

 

Any connectors will cause problems with resistance over time, soldering is the best for HV also use HV cable like ignition leads for spark plugs, or TV Eht leads..

 

Any connectors will cause problems with resistance over time, soldering is the best for HV also use HV cable like ignition leads for spark plugs, or TV Eht leads..

This is for a PhD project of mine, so I care more about safety really than degradation over time. As long as the connection is safe while the connection is made during my experiments and results taking.

But for direct connection to PCB, what is the best route for doing that? I doubt it will be as simple as just soldering the flying leads into a hole in the PCB as you would a standard capacitor or something, for example? Is there extra concern?

 

Extra concern would be preventing Sparking between conductors or PCB tracks by keeping a Large gap between them, also using silicone rubber insulation on the terminals.

 

Extra concern would be preventing Sparking between conductors or PCB tracks by keeping a Large gap between them, also using silicone rubber insulation on the terminals.

So the best practice is to indeed place some vias on the PCB, that the wires can be soldered to - but to ensure that they are kept a certain clearance distance apart, and that they have some silicon potting to prevent arc-over issues etc. Have I understood correctly? The PCB Via should be sized a certain diameter, or sized so that it matches the AWG/mm of the transformer flying leads that are used to wire the transformer?

Sorry for the rookie Q's - very keen to learn about magnetics design myself as a new junior engineer, rather than end up having to source it out. Thanks for your help.

 

I would agree that soldering the wires into through-holes on the pcb would be the most reliable and safe way. Just make sure that the copper clearance distance is appropriate. The potting material could be a nice touch as well, but verify whatever potting material you use will withstand the voltages you plan to work with.

 

Yes the hole size can be the same as the wire diameter, or you can use spade terminals.

 

Here is chart for using RTV.
Max.

 

Judging by your name "SiC Engineer" I assume you are using new power MOSFETs.
Is a GaN Engineer faster than a SiC Engineer? I am a bipolar junction Engineer. (power engineer joke)

 

For the direct soldered connections to the PCB, the first requirement is cleaning off all of the flux and soldering debris.
There are suitable encapsulating adhesives around to support and protect the connection. Those should be available from electronics distributors. Spacing on the PCB is already a well understood topic, so lots of information is available.
And there are all ready a lot of high voltage circuits on small boards in copiers and printers that use powdered ink. So there are many examples to look at. No need to make new discoveries there.

 

Judging by your name "SiC Engineer" I assume you are using new power MOSFETs.
Is a GaN Engineer faster than a SiC Engineer? I am a bipolar junction Engineer. (power engineer joke)

Haha! Actually, I am using GaN devices, the "SiC" is supposed to be as in "sick" which is younger slang for "sick" meaning good. A pretty terrible joke, really.

 

Here is chart for using RTV.
Max.

View attachment 209902

I suppose RTV is the silicon potting compound, after googling. Would it be best to pot each of the transformer PCB vias with enough silicon so that they can withstand the entirety of the secondary windings combined, say 6kV? Since the inches of material required are small, would this be the best approach?

 

For the direct soldered connections to the PCB, the first requirement is cleaning off all of the flux and soldering debris.
There are suitable encapsulating adhesives around to support and protect the connection. Those should be available from electronics distributors. Spacing on the PCB is already a well understood topic, so lots of information is available.
And there are all ready a lot of high voltage circuits on small boards in copiers and printers that use powdered ink. So there are many examples to look at. No need to make new discoveries there.

Awesome! Thank you Bill, I will look at those ASAP.