DPAK package Rant!

[Smokey]

So here is my rant for the day.  It's about the surface mount package DPAK (and D2PAK for that matter).
Now generally speaking I like this package size.  It is big enough to handle some power with the big SMD solder tab while being small enough to not take up a huge amount of board space and doesn't need to be screwed to anything. 

The issue I have with the DPAK package is the stupid worthless useless stupid (did I mention stupid) stubby middle pin!!!


So you may be asking, "But Smokey, that little pin isn't even long enough to connect to anything...  It's just hanging out minding it's own business... What's your problem???"

Well I'll tell you what my problem is.... man....  And it's voltage clearance!
Devices in this package are often some sort of power device with upwards of 600V voltage rating.
That stupid little pin is connected to one of the other pins (usually the big solder tab).  This means that the clearance between that stupid stubby pin and the adjacent pins you actually solder down is only 60mil.  What that pin does is takes a package where the pin to pin clearance would have been around 145mil and shrinks it down to 60mil ruining the voltage clearance!!!... ARGGG!!  And it's not even like there is some use for that pin that I'm just not taking advantage of.  It's too short to solder down.  The kicker is that looking at the PCB layout everything looks great since the only pads are for the outer pins and the tab.  No voltage problems there.

Now I get what they are probably doing.  Those dies are just the surface mount versions of TO-220s with the screw hole cut off and the leads crimped over.  And I think that's fine for the D2PAK which actually has the same overall body size of a TO-220.  But the DPAK has a slimmer package.  They already changed the body size, why can't they terminate that middle pin inside the package???

For high voltage but lower power circuits where a DPAK would be great without that middle pin, the next option for packages with over 90mil pin clearance is like a TO-247.  That makes me cry inside.  Having that little pin pointlessly ruin the voltage clearance of the DPAK is like.... like..... I can't even come up with an analogy that describes how frustrating that is!

And yes I know conformal coating is an option, but I have to guess that anyone that casually suggests conformal coating hasn't had to do it in production.  That or it was some military (or automotive) application where they didn't have a choice.  Conformal coating is a pain in the assssss.

So that is my rant for the day.  Stupid Stupid Stupid DPAK Pin

[Kremmen]

Yeah but why don't you tell us what you really think?

[Smokey]

Yeah but why don't you tell us what you really think?

Trust me.  If this stupid DPAK part ends up not working, it's going straight to the anvil for it's date with the Widlar-izer.

[bobdagangster]

Air has a nominal dielectric strength of 3kV/mm. At 1.5mm that gives a safety factor of 7.5 (compared to 600v max voltage), and 15 if the center pin is floating. What is the problem?

[wraper]

Transistor would die 1000 times already before voltage becomes high enough to shoot through that gap. So the isn't any problem.

[Kjelt]

If you think that is bad open up the case and see the distances on the die  But I don't see the problem either.

[Marco]

Surface tracking/creepage are the limiting factors ... 1.5 mm isn't enough for 600 Volt.

[wraper]

Surface tracking/creepage are the limiting factors ... 1.5 mm isn't enough for 600 Volt.

Don't confuse that with creepage distance needed to insulate primary and segondary sides of psu or semething similar. That isn't the same.

[M. András]

my most frustrating is this directfet mx package. http://www.irf.com/product-info/datasheets/data/irf6620.pdf particulary this eol marked fet. ordered 10 of them just to see the actual size and maybe do a prototype with it. when i saw it up close i said no way this can handle that current, there isnt enough surface area on the damn track for it under the device, i flipped the package 4 times over before i could put it back to its tape with tweezers

[Smokey]

It's not about air breakdown and arcing.  It's not about spacing on the die.

http://www.smps.us/pcbtracespacing.html

Welcome to the world of making electronics you plan on selling commercially.

[mamalala]

It's not about air breakdown and arcing.  It's not about spacing on the die.

http://www.smps.us/pcbtracespacing.html

Welcome to the world of making electronics you plan on selling commercially.

But then, the PCB trace spacing that you use has nothing to do with that little center stub there. You see, the "real art" is to shift the blame, i.e. to know where to point the finger at. You do your PCB layout and keep to the required track spacing. The part manufacturer does his package and keeps to his requirements. He tells you that his part isfit for the purpose, and that is that. If a problem arises, and it is due to the part not being up to spec, you can simply blame the manufacturer and point the finger to him instead.

In other words: it is not your job (or anyone else wo does PCB layout) to verify that the used parts _really_ do what the part manufacturer claims. There is a reason that pretty much every manufacturer puts a disclaimer in their datasheets about their stuff not to be used in medicla, avionics, or otherwise critical equipment, unless you contact them first and get some agreement. So, if you are not in any of those businesses disclaimed, you are good to go. If anything goes wrong, and you observed all the guidelines/regulations that apply to your field of work, the part manufacturer is the one who finally has to bite the bullet.

Same stuff in the EU with the CE mark, for example. Anyone can declare their product to be CE compliant. No need to have any expensive testing and certification. However, if it turns out that the stuff was not compliant, and you never did any testing and cert, you are the one to pay the bill. If instead you had some lab do the testing and cert, it would be that lab in turn who has to eat crow. Simple as that.

Greetings,

Chris

[AlfBaz]

But then, the PCB trace spacing that you use has nothing to do with that little center stub there. You see, the "real art" is to shift the blame, i.e. to know where to point the finger at. You do your PCB layout and keep to the required track spacing. The part manufacturer does his package and keeps to his requirements. He tells you that his part isfit for the purpose, and that is that. If a problem arises, and it is due to the part not being up to spec, you can simply blame the manufacturer and point the finger to him instead.

In other words: it is not your job (or anyone else wo does PCB layout) to verify that the used parts _really_ do what the part manufacturer claims. There is a reason that pretty much every manufacturer puts a disclaimer in their datasheets about their stuff not to be used in medicla, avionics, or otherwise critical equipment, unless you contact them first and get some agreement. So, if you are not in any of those businesses disclaimed, you are good to go. If anything goes wrong, and you observed all the guidelines/regulations that apply to your field of work, the part manufacturer is the one who finally has to bite the bullet.

Same stuff in the EU with the CE mark, for example. Anyone can declare their product to be CE compliant. No need to have any expensive testing and certification. However, if it turns out that the stuff was not compliant, and you never did any testing and cert, you are the one to pay the bill. If instead you had some lab do the testing and cert, it would be that lab in turn who has to eat crow. Simple as that.

Greetings,

Chris

Wrong forum dude this is an engineering forum... You may want to join one of these 
http://www.managementlogs.com/