Author Topic: 1400v TO-247 creepage distances  (Read 3442 times)

0 Members and 1 Guest are viewing this topic.

Offline jusanotherTopic starter

  • Contributor
  • Posts: 43
  • Country: gb
1400v TO-247 creepage distances
« on: June 10, 2019, 06:40:29 pm »
What am I missing here. Taking this transistor as an example IRG7PK35UD1PbF. A 1400v IGBT.

The creepage distance between adjacent pins on a TO-247 is ~ 2.93mm.

If we use the attached creepage chart (Creepage.jpg) and interpolate between 1250v (4.2mm) and 1600v (5.6mm) we get 4.8mm minimum creepage distance best case.

So how the hell do you use this IGBT at or near it's rated voltage !?

Is it a case of (as per the second creepage chart Ventec Creepages.png) "For glass, mica, ceramic or similar materials it is permitted to use minimum CREEPAGE DISTANCES equal to the applicable CLEARANCES." ?

Is the TO-247 case material classed as a "similar material" ?
This exemption only exists for Pollution Degree 1, so I take it Pollution Degree 1 is mandatory for devices like this at 1400v ?

 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: 1400v TO-247 creepage distances
« Reply #1 on: June 10, 2019, 06:51:06 pm »
If we use the attached creepage chart (Creepage.jpg) and interpolate between 1250v (4.2mm) and 1600v (5.6mm) we get 4.8mm minimum creepage distance best case.
At least second chart is completely irrelevant to your question. It's for isolation safety requirements between high and low voltage parts. Nothing to do with distance between transistor legs.
« Last Edit: June 10, 2019, 06:54:27 pm by wraper »
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: 1400v TO-247 creepage distances
« Reply #2 on: June 10, 2019, 07:15:59 pm »
First chart is for RMS voltage, not peak voltage. As of how to do this, form middle leg to increase distance from two others and make cutout in PCB between collector and 2 other pins.
« Last Edit: June 10, 2019, 07:17:37 pm by wraper »
 

Offline jusanotherTopic starter

  • Contributor
  • Posts: 43
  • Country: gb
Re: 1400v TO-247 creepage distances
« Reply #3 on: June 10, 2019, 07:31:35 pm »
I understand what you mean with respect to making a tripod leg configuration but this doesn't alter the distance between the pins as they enter the device. Why is 2.9mm at the device acceptable ?
 

Offline rsjsouza

  • Super Contributor
  • ***
  • Posts: 6105
  • Country: us
  • Eternally curious
    • Vbe - vídeo blog eletrônico
Re: 1400v TO-247 creepage distances
« Reply #4 on: June 10, 2019, 07:49:35 pm »
I understand what you mean with respect to making a tripod leg configuration but this doesn't alter the distance between the pins as they enter the device. Why is 2.9mm at the device acceptable ?
I am not sure if the tables you show are relative to "air" or a PCB track creepage.

A general rule of thumb is 30kV/cm or 3kV/mm, giving enough margin for the related voltages.
Vbe - vídeo blog eletrônico http://videos.vbeletronico.com

Oh, the "whys" of the datasheets... The information is there not to be an axiomatic truth, but instead each speck of data must be slowly inhaled while carefully performing a deep search inside oneself to find the true metaphysical sense...
 

Offline jusanotherTopic starter

  • Contributor
  • Posts: 43
  • Country: gb
Re: 1400v TO-247 creepage distances
« Reply #5 on: June 10, 2019, 08:17:01 pm »
Both the tables are for creepage, the second table, the Ventec one is from a PCB manufacturer. What is also confusing me is Printed Wiring seems to get a class of it's own, why ?
 

Offline coppice

  • Super Contributor
  • ***
  • Posts: 10031
  • Country: gb
Re: 1400v TO-247 creepage distances
« Reply #6 on: June 10, 2019, 08:31:19 pm »
When you want great isolation on PCBs always remember - the slot is mightier than the board. Slots are a greatly under appreciated PCB feature.
 

Offline MagicSmoker

  • Super Contributor
  • ***
  • Posts: 1408
  • Country: us
Re: 1400v TO-247 creepage distances
« Reply #7 on: June 10, 2019, 08:48:38 pm »
When you want great isolation on PCBs always remember - the slot is mightier than the board. Slots are a greatly under appreciated PCB feature.

True, but note that as far as safety agencies are concerned, the slot has to be 1mm or greater in width before it is considered to improve the creepage distance.


To the OP: it is easy to get confused by what is required to merely achieve "functional insulation" in a device - literally, the minimum creepage (distance along a surface) or clearance (distance through the air) that is required to hold off a given potential difference - and what is required to meet basic or reinforced insulation requirements, which are more influenced by regulatory/safety agency factors than anything else.

 

Offline jusanotherTopic starter

  • Contributor
  • Posts: 43
  • Country: gb
Re: 1400v TO-247 creepage distances
« Reply #8 on: June 10, 2019, 09:14:34 pm »
OK, so lets limit this to functional insulation. How does the TO-247 package meet that @ 1400v ?
 

Offline MagicSmoker

  • Super Contributor
  • ***
  • Posts: 1408
  • Country: us
Re: 1400v TO-247 creepage distances
« Reply #9 on: June 10, 2019, 09:25:44 pm »
OK, so lets limit this to functional insulation. How does the TO-247 package meet that @ 1400v ?

Shortest distance between pins is ~3mm, which is easily capable of holding off 1400V even for "Pollution Degree 2" (the most common category which basically assumes that condensation and/or non-conductive dust will occasionally be present).

 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: 1400v TO-247 creepage distances
« Reply #10 on: June 10, 2019, 09:31:13 pm »
OK, so lets limit this to functional insulation. How does the TO-247 package meet that @ 1400v ?
Easily, because minimum creepage distances depend on the comparative tracking index and the package material will be better, than most PCB materials.
https://en.wikipedia.org/wiki/Comparative_Tracking_Index

In any case, this is irrelevant since the device should be protected against over-current by a fuse. Arcing over should present no safety/fire hazard, as the fuse shall fail first, before any flames appear.
 

Offline jusanotherTopic starter

  • Contributor
  • Posts: 43
  • Country: gb
Re: 1400v TO-247 creepage distances
« Reply #11 on: June 10, 2019, 10:44:33 pm »
Shortest distance between pins is ~3mm, which is easily capable of holding off 1400V even for "Pollution Degree 2"

So assuming the case has a CTI >= 600v putting it in material class 1, and using pollution degree 2, even at 1000v you need 5mm creepage spacing according to this table that stipulates it is for functional insulation. http://download.wecogroup.com/docs/en/TechInfo_F4_Creepage_distance.pdf

How do I prove ~3mm  is easily capable of holding off 1400V even for "Pollution Degree 2" ?

In any case, this is irrelevant since the device should be protected against over-current by a fuse. Arcing over should present no safety/fire hazard.

I haven't got as far as safety concerns, I can't understand how it reaches functional requirements.
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: 1400v TO-247 creepage distances
« Reply #12 on: June 11, 2019, 04:15:54 am »
I wonder if a Conformal Coating would be required here.  I don't have any hands-on experience with that package at those voltages.  I did design an assembly that used a Silgard product to solve a creepage problem with wirebonds to a flex circuit operating at about 100 VAC.   The pitch was much tighter, so the voltage gradient was higher.  The device is proprietary and I can't really go into too much detail.
 

Offline MagicSmoker

  • Super Contributor
  • ***
  • Posts: 1408
  • Country: us
Re: 1400v TO-247 creepage distances
« Reply #13 on: June 11, 2019, 10:01:05 am »
So assuming the case has a CTI >= 600v putting it in material class 1, and using pollution degree 2, even at 1000v you need 5mm creepage spacing according to this table that stipulates it is for functional insulation.
...

Ahh yes, I see your point. You will likely have to apply an approved conformal coating or silicone to the lead area to meet creepage requirements UNLESS the overmolding epoxy can be considered as the same as glass, mica, or other materials that do not experience surface breakdown/carbonization. Note that such materials are not put into material group 1 (CTI >= 600), rather, they effectively have "infinite" CTI so only have to meet clearance, not creepage, distance requirements.

But if you think the example you gave in the OP is bad, check out the IXYS IXBH12N300 3000V rated IGBT in a standard TO-247 package:

https://www.digikey.com/product-detail/en/ixys/IXBH12N300/IXBH12N300-ND/2183282

Talk about brazen!
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: 1400v TO-247 creepage distances
« Reply #14 on: June 11, 2019, 12:38:41 pm »
Molding compound seems suspect, but it looks like high-CTI compounds are indeed available, up to 600 but maybe not infinity.  The dominant component is silica fume, and it appears a wide variety of epoxy and phenolic resins are also involved.  Basically if an arc forms, it can burn away the resin, leaving a huge amount of silica "ash" that hopefully tends to insulate things.  This does assume the resin doesn't carbonize under all that silica, which I expect is why it's not usually more than 600?

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf