Big current ratings on really little devices

[ScotCarl]

Hello everyone.
Long time reader, first time question asker.
I apologize if this question has been asked elsewhere - I've searched, but I'm not really sure just how to phrase it, so my searches have been futile.

I'm currently designing a controller that will be using SCRs to switch some loads - pretty simple stuff really. I'm in the throws of selecting the SCRs that I will use, and I am reviewing many different brands and models.

I'm interested to know if someone can provide some insight into how it is that there are a number of PCB mount SCRs on the market that have claims of "Rated Average On-State Current" (or similar) that seems well in excess of what could be possible given their package sizes. This is not the peak current, but supposed steady state current. I've noted 2 examples here for purposes of illustration

• IXYS, CS45-16IO1, which is in a TO-247 package, and claims an average on state current of 48A.
• Vishay, VS-25TTS12SPBF, which is in a surface mount package and claims an average on state current of 16A.

(These are just 2 examples of many).

Now while I can accept that the silicon within the component is capable of handling the rated current, the part I don't understand is the connection of the component to the PCB. If we consider the IXYS example, the legs on the component only have ~1.5mm2 cross section, and are spaced 5.5mm apart. Apart from the concerns about warming up the legs of the component, the PCB tracks can be a maximum of ~9mm, and a quick check on an online PCB calculator indicates that if we were to push >40A through a 9mm wide 2oz copper trace, we would be looking at a >200 degC temperature rise on the PCB, which is not that acceptable in most designs I've ever done.

Given the vast number of devices like this on the market, I'm wondering what it is I'm missing here, because it just doesn't make sense to me. If anyone out there has any experience using these little devices for switching large currents, I'd be most grateful for some comments or feedback on this.

Thanks,
Scott.

 

[crutschow]

They do seem to be a little loose with their ratings. The maximum ratings are when mounted on a large heat sink.
If you look at the data sheet for the VS-25TTS12SPBF, for example, you will see current ratings for different types of heat sinks attached to the SCR (1st page, second table). As you can see, the SCR must be soldered to a large aluminum heat sink with a low thermal resistance to air to get the higher current ratings, and that's for a half-wave AC duty-cycle.

 

[Misterbenn]

The component temperature would need a really good heatsink to get that rating, but it's not impossible. A water cooled thermal block for example.

As for the copper traces, this is a through hole component so the traces can be on multiple layers which will reduce the resistance and heating. The weak point is usually the solder connection to the pin. With the pin connection you need to be really sure that the internal hole surface is fully plated and that you have a good solder connection.

To give this some perspective, I have a 150A continuous screw tab IGBT module that i once cut open as part of a failure investigation. Internally the screw tab connection to the IGBT went via two through hole pins on a PCB. This was a sealed module with no convection or conduction for the PCB traces.