I’m interested in CE testing a device which includes the BBB. In order to determine if the device will pass, I’ve been shocking it with an ESD gun (both contact and discharge). I have been running into an issue where when I shock the ground plane with a 4kV contact discharge, the BBB Ethernet becomes unresponsive and does not recover until I cycle the power. I’ve tried this test with only the BBB and it reacts the same way. I see there’s a CE certification which says it passes 4kV contact discharge, but I’ve been unable to replicate it.
Are there caveats to the CE certification which the BBB has relating to ESD testing?
Is this a self certification?
Do you have any suggestions on how mitigate this problem?
I will appreciate any information you can provide on this problem.
Thanks,
Rich
Hi Guys, I am seeing the same issue as Richard above. Was a solution ever found for this ESD reset issue?
In the past I have had some success on solving this issue on other micros by adding some capacitance on the micro’s reset line to prevent the reset I/P bouncing when the ESD takes place, however looking at the BBB schematics, this is already included in the form of C24 being 1uF.
Richard, did you find a solution to this?
Gerald, I assume this family of processors must be used in many commercial products. Do you have an idea of what solutions were used to overcome this?
I am not sure what was used, other than putting it in a box or adding ESD protection devices on every point on the board likely to be touched by a person.
The certification was done by a lab. You are welcome to download the full report from the WIKI.
The beagle has been designed into a system which is in a plastic and metal enclosure. The only I/O it has to the outside (other than touchscreen) is 2 USB connections.
These connections are what are prone to the 4kv ESD. A discharge on the shell of the USB connector causes the micro to reset. This occurs even though ESD suppression has been designed in at the USB connectors.
The micro seems particularly prone to this phenomenon. ESD is an awkward problem to solve, there isn’t one solution to fit all and things such as enclosure design can have an influence…
You need to make sure that there is an earth ground on your power supply. The case of the USB connectors go to the common point ground, which is a zero ohm resistor by default to the DC ground input.
How is your esd protection designed at the usb connections, or are you
hitting the usb connectors right on the bbb? Would need a lot more
details about exactly how the system was being built and where you are
discharging to know what might be going on. I'm guessing your esd
protection may not be getting to a solid enough chassis ground close
enough to the connector that is being hit.
How is your esd protection designed at the usb connections, or are you
hitting the usb connectors right on the bbb? Would need a lot more
details about exactly how the system was being built and where you are
discharging to know what might be going on. I'm guessing your esd
protection may not be getting to a solid enough chassis ground close
enough to the connector that is being hit.
Actually, from what I recall, the USB protection isn¹t in the right place.
The correct layout is USB connector, ESD protection and then USB
transceiver. However, in the BBB, the USB connector is connected to the
USB transceiver and the ESD protection is on the other side of the USB
connector. Granted, the distance is small and may even work correctly, but
this layout doesn¹t follow good ESD layout principles.
If we are just at this theme, what about the ethernet lines?
Looking at the schematic, sheet 9, the 4 active lines are directly connected to the LAN8710 chip.
Or is there anything not visible inside the RJ45 connector?
The BBB is positioned centrally at the front of a relatively large enclosure. It is connected to a display/touchscreen which are also situated in the same area.
The USB I/O is brought out of the enclosure at the side of the unit via USB cables and another PCB. Both of the BBB USB connectors are connected within the enclosure by their own USB cable to a PCB of my own design.
This PCB sits at the internal side face of the enclosure and provides access to USB Connectors from the outside world.
On this PCB I have designed-in a Wurth Electronics ESD solution. This solution is documented in the PDF (titled: USB Port Protection) at this link:
The BBB is positioned centrally at the front of a relatively large enclosure. It is connected to a display/touchscreen which are also situated in the same area.
The USB I/O is brought out of the enclosure at the side of the unit via USB cables and another PCB. Both of the BBB USB connectors are connected within the enclosure by their own USB cable to a PCB of my own design.
This PCB sits at the internal side face of the enclosure and provides access to USB Connectors from the outside world.
On this PCB I have designed-in a Wurth Electronics ESD solution. This solution is documented in the PDF (titled: USB Port Protection) at this link:
The solution I have implemented is described in the document as “Double protection of single USB port” on page 7.
Any further comments welcome…
So what happens when you discharge the ESD probe? Are you probing the USB connector case, pins, etc? I’m guessing the board either resets or hangs? This could be a grounding issue or an antenna issue; either way, you have to use a process of elimination to determine which. Best way to determine if it is an antenna issue, start by looking for radiated signals (you need a RF probe with a spectrum analyzer). If you don’t have the required equipment, your local university is always a great resource. EMI and EMC are really two sides of the same coin. If a signal radiates at a certain frequency, then it is also susceptible to those same frequencies. For grounding, it depends on how good your ground plane is and whether you have holes in your ground plane.
When we discharged the ESD into either USB shell (positioned at the edge of the enclosure as previously described) the beagle reset (more often that not).
We believe we have now fixed that particular issue by solidly grounding the USB shell to the enclosure at the point of entry. A mechanical fix.
There are other parts of the units that are metal and would be touched by the end user. Some of these are also susceptible to the ESD. We are going through a process of grounding these but this sometimes seems to make things worse!
If by ground plane you mean on the enclosure rather than at the PCB level then yes there are holes in the ground plane, by necessity due to the mechanical design of the enclosure.
We have previously assessed the unit for emissions and are up around the limit in the region of 160MHz to 180MHz. We have not yet investigated the cause of that.
Thank you for your help John, much appreciated. Any further advice welcomed…
When we discharged the ESD into either USB shell (positioned at the edge of the enclosure as previously described) the beagle reset (more often that not).
We believe we have now fixed that particular issue by solidly grounding the USB shell to the enclosure at the point of entry. A mechanical fix.
There are other parts of the units that are metal and would be touched by the end user. Some of these are also susceptible to the ESD. We are going through a process of grounding these but this sometimes seems to make things worse!
If by ground plane you mean on the enclosure rather than at the PCB level then yes there are holes in the ground plane, by necessity due to the mechanical design of the enclosure.
We have previously assessed the unit for emissions and are up around the limit in the region of 160MHz to 180MHz. We have not yet investigated the cause of that.
Thank you for your help John, much appreciated. Any further advice welcomed…
You need to look for frequencies in the GHz region because of the sharp rise time for ESD pulse. I was talking about holes in your PCB ground plane. Always remember there is a return path to every signal and this is what newbies forget. What you want to do is prevent current from the ESD passing through your board. All these concepts and more are covered in the EMC bible:
ESD testing has to pass given the requirements you have. That is at least the understanding I had. For example, if you send an ESD shock, and it renders your unit stuck in a reset mode, that is fine as long as your requirement of passing is that you have to shutdown and restart.