Does my device needs an EFT test ?

Go To Last Post
12 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

 

  Hi there

 

  A client of mine has a product that is powered at 24Vdc or 24Vac. Among other features, the unit has an RS485 port.

 

  A number of units were submitted to a certification lab in China. Upon performing the EFT test, all failed with some components melted and in one case going missing from the PCB.

 

  Here is the input power circuitry:

where TVS10 is SMBJ36CA and the F116 is a PTC fuse 1.1A from Bell Fuse Inc.

 

  Here are some pictures showing the amount of destruction:

 

and for RS485, you can see F32 disapeared completely

 

  Does this unit powered at 24V needs a test with so much energy ?

 

  The unit power consumption is under 5W.

 

Last Edited: Tue. May 26, 2020 - 04:17 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

angelu wrote:
Does this unit powered at 24V needs a test with so much energy ?

So how much energy did they use?

 

We'd need to know more about your device & its intended application to know what tests would be applicable / mandatory.

Top Tips:

  1. How to properly post source code - see: https://www.avrfreaks.net/comment... - also how to properly include images/pictures
  2. "Garbage" characters on a serial terminal are (almost?) invariably due to wrong baud rate - see: https://learn.sparkfun.com/tutorials/serial-communication
  3. Wrong baud rate is usually due to not running at the speed you thought; check by blinking a LED to see if you get the speed you expected
  4. Difference between a crystal, and a crystal oscillatorhttps://www.avrfreaks.net/comment...
  5. When your question is resolved, mark the solution: https://www.avrfreaks.net/comment...
  6. Beginner's "Getting Started" tips: https://www.avrfreaks.net/comment...
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

angelu wrote:
Does this unit powered at 24V needs a test with so much energy ?
Uncertain

The context for the following is lightning instead of EFT and megaAVR in Florida (diurnal thunderstorms are common in some of the Florida Panhandle)

I posted a new version of my crc tool, CRCSNTOOL in the projects area... | AVR Freaks

Redesign Overvoltage Protection to Bolster Reliability and Safety | Electronic Design (Bourns GMOV)

Bourns - GMOV- Components

 

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Whether or not an EFT (Electric Fast Transient) test is needed depends on the application.

 

Let me give you an example of a very unexpected real-life EFT experience. I was designing electronic circuitry for CCTV systems. In this case, it involved a controller and a remote camera on a pan/tilt/zoom (PTZ) base. The PTZ unit was controlled via RS422. Almost always, these were installed in or sometimes on the exterior of buildings. BUT, some of our equipment had been installed in a perimeter security system at the Atlanta Airport. Camera+PTZ units were installed at regular points along the perimeter fence, and the cabling was routed to the control center via buried PVC conduit. Every spring, equipment would mysteriously fail. No lightning hits to the equipment or fencing, it would just fail. That said, lightning was VERY common in that area. Over-voltage protection on video and RS422 in the control center was always fried (as in the photos, above). Finally, the owner of the company went to Atlanta to see what was happening. After several days of investigation, it was found that there were arc holes at various points THROUGH the PVC conduit, and through the coax and power and control cable jackets. After some head scratching, it was discovered that each failure was associated with a NEAR-BY lightning strike. Apparently, what was  happening was that a current wave was traveling through the ground, outward from the strike zone. When the wave passed the conduit, the voltage difference between the cabling inside the conduit and the earth, outside, was so great that an arc occurred, killing the electronics at the control center; the camera and PTZ unit floated and were unharmed. 

 

The lesson: in spite of best practices in installation, the electronics in the control center were being subject to EFT conditions far in excess of what we ever imagined. The board had to be redesigned with MUCH beefier protection. My recollection was that failures ceased once the redesigned equipment was installed. But, more importantly, we never imagined that our equipment, which was mostly designed for interior or protected exterior locations, would ever be subject to lightning-type stresses. In short, you really need to test in a way that is appropriate for your equipment and YOU NEED TO HEED THE RESULTS of the testing. This may/probably will require redesigns but if you want your stuff to be reliable in the face of such assaults, then you have to do it.

 

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

  Thank you for your replies.

 

awneil wrote:

We'd need to know more about your device & its intended application to know what tests would be applicable / mandatory.

 

ka7ehk wrote:

Whether or not an EFT (Electric Fast Transient) test is needed depends on the application.

 

  Where I can find the type of applications that require an EFT test ?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

  Jim, wasn't that the PVC was not burried deep enough in the ground ?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Do you have part# on the 1.1A "polyswitch"?  They are rather slow (compared to a transient burst) & only have a finite energy interruption rating.  Note also the TVS clamp voltage is pretty high (I think it was almost 60V)....so that could damage other components, not rated for that level. 

 

Was the unit operating during the applied pulse?  If so, some (or a lot) of damage can be generated due to system upset rather than the pulse of energy itself.  For example, a powered motor driver Hbridge that all of a sudden locks up into full cross-conduction & burns up the transistors.  Or a switching power supply that get knocked out of kilter, and starts frying things down the line.   Lots' of fun to investigate!

 

  The unit power consumption is under 5W.

Under normal circumstances ....be careful that doesn't cause an oversight. 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Tue. May 26, 2020 - 07:21 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I never had any information on conduit depth. So, I do not know. BUT, given where the installation was, I would expect it to have been to local code, what ever that might have been. This event was more than 10 years ago, and I have no physical records, just memory.

 

However, that really isn't the point. The point is that "it" can get you (or your equipment) even when you don't expect it. Maybe especially when you don't expect it or when you don't think that it could possibly happen. Everything was designed for ESD, and that worked. ESD protection is simply not adequate for EFT.

 

Some equipment can be reasonably expected to never experience EFT. For those, you ignore EFT, with fairly little peril. If you cannot tolerate customer replacements and the resulting loss of customer confidence (in your company or your products), then you had better pay attention to EFT. An example of "not likely to experience EFT" might be some battery powered device with no external wires; wireless ear-buds might be an example of that. Ear-bud chargers are a different story.

 

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

Last Edited: Tue. May 26, 2020 - 08:02 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Is that 24V power input connected to a plug-in AC adaptor with a standardish 1.8m lead ?If so EFT burst tests are not required. If however this power comes from a "DC power bus" or from lines that exceed either 3m or 30m I can't remember then the testing guys haul out the nasty stuff.

 

In any case your circuit is wrong. The TVS diode should be as close as physically possible to the power jack with adequate clearances to the more sensitive components. The fuse is not really for EMC type perturbations but to protect against burning under fault conditions.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Here some interesting EFT tidbits, for what they are worth...your milage may vary

 

https://www.cypress.com/file/138636/download

 

As an aside, I found fig 44 rather interesting (to see actual nH numbers)

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

That figure is really interesting. Actually, the whole appnote is a gem.

 

The rule of thumb I remember is 1nH/inch (I suppose that this was for 24 AWG connecting wire, but not sure). If you look at how the inductance varies with the connection scheme, it should be obvious that most of the inductance is in the connection, not the resistor. Remember, 1nH will generate 1V with a current pulse having 1A/ns risetime!

 

Jim

 

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

There seems to be a ground fill on the top layer - this is probably creating a favourable path for the transients rather than it going where you want.
Seems like you really do need the EFT test. Currently your circuit is a major failure! When I test for EFT, first i test to the required standard, then crank up the volts until i get a failure. Hopefully the failure level is way beyond the required standard.

I’ve found that smt polyfuses like to go up in smoke, so i tend to use through hole ones. Adding a ceramic capacitor on the power input might be beneficial - hopefully it soaks up some of the transient leaving the cleanup for the tvs. This is a common technique in automotive ecus

Last Edited: Wed. May 27, 2020 - 06:05 AM