3
\$\begingroup\$

I'm currently working on an electronics project. This project is intended to be portable and battery powered. I'm curious about whether or not I should use certain components on my PCB do being too ESD sensitive.

For example, I'm currently trying to use this clock on my design. After ordering it, I notice that it came packaged with a warning to only open/handle at an electrostatic-free workstation.

Now obviously I could oblige to this, however, my end design is an exposed PCB in which a user may accidentally touch. The end user (colleagues of mine, not external customers) will be handling this device to press pushbuttons, look at lights, etc. Furthermore, my design only covers the electronic portion thus I can't create an enclosure to protect during operation. That's the end user's duty.

So I'm somewhat concerned as to whether or not I should proceed with using the previously referenced IC. I'm also not sure as to what kind of ESD protection is used on portable electronics that may allow me to use the chip.

So my questions:

  • Is there ever a point in which a certain component shouldn't be selected on an open PCB design because it's too ESD sensitive? Or can you just bypass this with appropriate ESD protection?
    • What kind of ESD proection is used for portable electronics?
\$\endgroup\$
  • \$\begingroup\$ Yes, any board seen by the customer that has ESD-sensitive devices should have ESD protection on it. ESD protection is usually done using TVS diodes \$\endgroup\$ – DerStrom8 Nov 3 '16 at 19:55
  • \$\begingroup\$ Consider that these components are tied into other components with multiple, tiny protection diodes, resistors, capacitors, etc., all tied into the same nodes. It's also difficult to reach or touch certain areas, too. If you have any really high impedance nodes, even getting close can be a problem. But often, then adding static zap or efield protection to those nodes ruins their performance, too. So in some cases you'll see mental cans soldered around those areas. Use your judgment but also do some testing and make adjustments when you find problems. \$\endgroup\$ – jonk Nov 3 '16 at 20:08
  • \$\begingroup\$ @jonk: I understand it's a typo, but thinking of mental cans to protect ESD sensitive device is really hilarious! You made my day, man! :-D \$\endgroup\$ – Lorenzo Donati supports Monica Nov 3 '16 at 22:33
  • \$\begingroup\$ @LorenzoDonati ;) They are often used in conjunction with tin foil hats! \$\endgroup\$ – jonk Nov 3 '16 at 22:37
  • \$\begingroup\$ @jonk Yep! Indeed! Especially against aliens (or the NSA) using the infamous Extra-Sensory Detection technology! :-D \$\endgroup\$ – Lorenzo Donati supports Monica Nov 3 '16 at 22:50
4
\$\begingroup\$

There will always be a certain ESD level which can damage your product. The question is, how much protection do you need so that it is "good enough" ?

For (portable) consumer products the sensitive points are all the conductive / metal contacts which are exposed. So for a mobile phone, you would use TVS diodes or such to implement extra protection on the charging port and headphone socket for example. That is not to say that there would be no protection without these extra diodes ! There is always ESD protection in any modern IC but it will be a lower level protection, it can handle ESD but not too much. For example: 500 V HBM (Human Body Model).

This level of protection (500 V HBM for example) would be enough for handling that crystal oscillator you ordered in a somewhat ESD safe environment. With this I mean, precautions have been taken to keep ESD pulses at a low level. For example: a conductive and grounded ESD mat and the user grounded by wearing an ESD wrist band.

Don't draw too much of a conclusion about ESD sensitivity by looking at the packaging or warnings companies like mouser use. If you would order resistors I bet they would also come in ESD safe packaging and might also come with a warning although resistors are not ESD sensitive.

Regarding your PCB, since it is not a consumer product and you can expect the user to know that (s)he should not be poking his/her finger everywhere and generate ESD discharges. In my experience the build-in ESD protection of most ICs is sufficient and will provide enough protection for normal usage by a user who is aware of ESD.

So if you have external connections on your board: use extra protection.

For the part that the user should not touch anyway: no extra protection needed.

\$\endgroup\$
  • 1
    \$\begingroup\$ 500V is insufficient for many parts of the world where humidity can be low so 1kV and 3kV is used for components and 15kV for system tests with ESD \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Nov 3 '16 at 21:02
  • 1
    \$\begingroup\$ The 500 V was just an example. Maybe you always use 1, 3 and 15 kV, that does not mean everyone else must do the same. \$\endgroup\$ – Bimpelrekkie Nov 3 '16 at 21:26
  • \$\begingroup\$ @FakeMoustache Many resistors are actually ESD sensitive in that an ESD pulse can cause their value to drift out of tolerance. Vishay has some good app notes on this advertising their ESD immune metal foil resistors. \$\endgroup\$ – user4574 Nov 3 '16 at 21:44
  • \$\begingroup\$ @FakeMoustache in my experience from auditing the design and test processes of dozens of large companies like the one I was in at the time. They were generally consistent with our practice for quality products. This is the error of many low tech Chinese companies and DIY designers alike yet even the Chinese have more advanced test methods in certain areas, so this is not menat to slur all chinese products as they have one of 2 highest 2GV power transmission systems in the world in R&D and longest 1.5kV I can dig up the IEC standards if you like to compare on susceptibilty \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Nov 3 '16 at 21:51
  • 1
    \$\begingroup\$ @user96037 Sure for precision resistors that could make sense. But I was talking about the cheap 1% or 5% tolerance resistors we all use when precision is not needed. \$\endgroup\$ – Bimpelrekkie Nov 3 '16 at 21:51
3
\$\begingroup\$

Let me tell you a story.

In the early 80's all keyboards would have tiny red LEDs exposed and customers found their keyboards would die for no reason. Our company did research on our computer products and tested every new product with a Schaefer 15kV to 25kV ESD generator to simulate the human finger or cart model and found the zap could be silent and not felt by a user going past the exposed 5mm LED to the leads and either anode or cathode connect to CMOS logic could trigger an SCR latchup effect which could cause the so-called ESD protected CMOS to burnup ( internal short, not flames) from an SCR crowbar in the sublayer of all CMOS.

Since then all LED's were buried well below the surface and used light pipes to prevent this or very long epoxy bodies just below a plastic surface guard.

If you do not test your product for ESD susceptibility or design it to withstand 1 hundred random discharges @ 15kV from 300pF, to all user accessible interfaces

... you can expect failures unless the user is given clear instructions how to avoid this and accept the consequences of failure.

  • I have dozens of similar stories of ESD failures at embarrassing times.

One day (1982) we were in a customer (MTS) owned "test home" with all our newly invented technology aka Project IDA, that offered digital telephony, pay-Tv , fire/burglar alarms , tele-shopping with remote keyboard, online TV opinion polling in real-time, digital weather maps, high speed serial ports in the phone jacks with dual RF TV converters outside on the poles and a whole 2 way 1.544 Mb sync modem to every home in the trial. Then the Mayor of St Louis was interested and brought his group for a demo and when the demonstrator touched the TV it conducted ESD to the motherboard in the basement and reset the unit into a fault mode, recovered by power cycling. ( so after quickly power cycling it .. dry winter in Winnipeg, the demo resumed) ( Mayor never bough it and was just trying to steal the technology to buy votes for next election) it ended up being sold off to a company that own Scientific Atlanta, Mattel, Intelivision etc. in Phili.PA and the company was bankrupt ahead of its time.

Next day we had 3 fixes for the ESD event to never happen again with improved ground shift immunity with CM choke , better grounding etc.

\$\endgroup\$
  • \$\begingroup\$ I remember using those generators to test board designs against ESD! \$\endgroup\$ – jonk Nov 3 '16 at 20:01
  • \$\begingroup\$ I designed another test with 15kV impulse to a flat sheet of metal for E-Wave impulse test at 1m into our Corporate standard while the UHF radiated immunity was too weak without modulation so I added 1kHz AM to the wide spectrum immunity tests. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Nov 3 '16 at 20:12
  • \$\begingroup\$ Yah but the Shaeffer was pretty tame compared to a Tazer these days or even better neoprene shoes in a dry winter day in a hotel room with Nylon carpets. I remember creating 3" arcs with a key to the door knob or >75mm or > 75kV.. recently when I was testing MVA distribution transformers for a client/factory owner up to 200kV , I had no fear of climbing up on top of the unit and connecting the wires to the bushing, but I would touch the wires submersed in oil with remanence as they continued to keep on giving zaps, long after disconnect.. a job for the factory workers. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Nov 3 '16 at 20:15
  • \$\begingroup\$ Wow. You can get nice, hard xrays with 75 kV. (Dental uses a mere 70 kV, typically, on the rotating tungsten target.) Which is something else that a lot of people forget when they are messing around with high voltages for huge arcs in air. \$\endgroup\$ – jonk Nov 3 '16 at 20:17
  • \$\begingroup\$ You can get Gamma rays at 15kV if the rise time in in picoseconds so it not voltage dependent , rather dI/dt. They use 1GHz detectors for PD in Transfos also UV and a few researchers in Gamma but UHF gives best results for internal ESD in oil called Partial Discharge or PD. there are about 10k thesis papers on this topic alone \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Nov 3 '16 at 20:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.