Advertisement

Blog Signal Chain Basics

Signal Chain Basics #91: The Difference Between HBM & IEC 61000-4-2 ESD Immunity

The electrostatic discharge (ESD) ratings in component data sheets often confuse system designers by not specifying the actual standard to which a component was tested. The following explains the differences between the two most commonly applied ESD standards, the human body model (HBM) and the IEC 61000-4-2 standard.

There are two categories of ESD: component-level ESD and system-level ESD. Component-level ESD is necessary in the manufacturing environment where component assembly, packaging, and shipping can cause ESD damage to a single device. Here the HBM standard is applied, simulating a charged person’s discharge through the device under test (DUT) to ground.

Because component assembly, packaging, and shipping are performed in a controlled ESD environment through the application of ESD protective gear, the ESD stress upon a component is drastically reduced. Hence, the HBM test presents a rather soft ESD test. This is sufficient for controlled environments, but inadequate for systems in uncontrolled ESD environments.

System-level ESD is required in the uncontrolled end-user environment where, for example, a charged user can subject a handheld device to ESD levels of up to 40 kV by touching connector pins when plugging or unplugging cables.

To better rate a system’s ESD vulnerability, the IEC 61000-4-2 standard was developed. It replicates a charged person discharging via a metal object, such as a screw driver, into a grounded electrical system.

Differences between HBM and IEC 61000-4-2
The main differences between the HBM and the IEC 61000-4-2 standards are the number of strikes applied during testing and the generator models (Figure 1), which create differences in the waveforms’ rise times and peak currents (Figure 2).

Figure 1

ESD generator models according to HBM (left) and IEC 61000-4-2 (right)

ESD generator models according to HBM (left) and IEC 61000-4-2 (right)

The pulse rise times of the two waveforms differ largely. The rise time of an HBM strike can extend up to 25 ns when discharging into a 500 ohm resistor, while the rise time of an IEC 61000-4-2 strike is less than 1 ns. Hence an ESD structure designed for an HBM strike might not turn on when exposed to IEC-ESD strikes, and the circuit to be protected becomes damaged.

The amount of peak current is critical to whether or not a component survives an ESD strike. It is possible for an IC to survive an 8 kV HBM strike but fail at a much lower, 2 kV IEC-ESD strike.

The number of ESD strikes in the HBM specification is limited to a single positive and a single negative strike, whereas IEC 61000-4-2 requires a minimum of 10 positive and 10 negative strikes. Therefore, during an HBM test it is possible to survive the first strike, but fail on subsequent strikes due to damage sustained in the first strike.

Figure 2

Waveform comparison between HBM and IEC 61000-4-2 ESD strikes for 16 kV

Waveform comparison between HBM and IEC 61000-4-2 ESD strikes for 16 kV

For equal test voltages the energy content of an IEC-ESD pulse is 7.56 times higher than that of an HBM-ESD pulse, and for equal energy content, the ratio of an HBM to an IEC ESD test voltage is:

Equation 2 allows determining the maximum HBM test voltage for an ESD protection circuit passing a maximum IEC-ESD strike. For example, a protection structure successfully passing a sequence of 16 kV IEC ESD strikes is able to withstand a single HBM strike of:

However, the opposite, that a 44 kV HBM ESD structure can withstand a 16 kV IEC strike, is not likely as it won’t switch fast enough to catch the short front time of an IEC ESD pulse.

Conclusion
Because end-user applications are subjected to many ESD strikes during their lifetimes, system designers must be aware of misleading ESD specifications in component data sheets.

Devices complying with the IEC 61000-4-2 standard usually indicate this fact, while those that don’t are commonly tested to the HBM standard.

— Thomas Kugelstadt

References:

9 comments on “Signal Chain Basics #91: The Difference Between HBM & IEC 61000-4-2 ESD Immunity

  1. vasanjk
    July 18, 2014

    Thomas

     

    This post I an eye-opener for me. This means, we need to employ ESD diodes and TVS diodes with much higher specs than those specified in the datasheets. Some practical pointers on real datasheets can help even better to understand the point.

  2. samicksha
    July 19, 2014

    I guess in HBM tests are intended to ensure that integrated circuits survive the manufacturing process.

  3. jcup2306
    August 1, 2014

    I think the post is slightly misleading, although the author did allude to the key difference in the beginning; “HBM is designed for the manufacturing area where there is a lot of ESD controls in place.” Component level HBM testing that is on datasheets is the standard because manufacturing is really the only time that ESD CAN strike the device directly. 100% of the rest of the IC's life is spent on a board that has built in ESD protections such as the plastic or metal enclosure, the board, other devices, ESD components, isolation, and even the air around it. If you decided to expose every IC to system level ESD measurments like IEC61000-4 then you would break 100% of ICs before they ever left the manufacturer. 

    Simple rule: make sure the test you are using actually applies to what the device will be exposed to in real life usage. The article is very informative and definitely good information for everyone in the electronics industry, but it is just missing that critical clarification.

  4. amrutah
    August 1, 2014

    “HBM tests are intended to ensure that integrated circuits survive the manufacturing process”

    @Samicksha: I think ESD is an event and it can happen any time during the manufacturing stage or just when someone is handling a off-the-shelf IC.  What to do, how to dissipate and how to protect the IC when a charged body touches across leads.  I think HBM model standardizes how much of the charge/ volts the IC should be able to handle.

  5. BIG TOM_#1
    August 4, 2014

    Hello NEWBIE,

    you stated that if one decided to expose every IC to system level ESD measurments like IEC61000-4 then you would break 100% of ICs before they ever left the manufacturer.

    1) IEC61000-4 is an entire family of immunity standards of which there are 38.

    2) IEX-61000-4-2 is all about ESD immunity. Here Texas Instruments, and in particular the Interface department, has RS-485 transceiver that really do withstand up to 16kV IEC61000-4-2 ESD and up to 40kV HBM ESD.

    The landsacpe in ESD protection changes rapidly. I calculated that 8kV (IEC-ESD) on-chip will do fine. For more energatic pulses such as an EFT pulse train (IEC-61000-4-4) or even a 1.2/50us or 10/700us surge pulse (IEC61000-4-5) you will need external TVS diodes anyway.

     hope this helps too. Best regards, Thomas

     

     

     

  6. BIG TOM_#1
    August 4, 2014

    Hello Vasanjk,

    in many of our new data sheets (such as SN65HVD72) I have implemented TVS diodes that provide additional EFT (electrical fast transient or IEC610004-4) and Surge (IEC61000-4-5) immunity. These circuits have been tested by Bourns. Inc in combination with TI.

    If I get time before summer vacaiton I will try to create a design guide on how to determine the minimum protect one needs, what components are necessary and how to select them for a reliable surge protection.

    Regards, Thomas

     

  7. green_is_now
    October 23, 2014

    I have designed a active and passive input protection that adapts to speed of the incoming waveform to shut of input switch faster to compensate for lack of speed at static voltage switch point.

    Both dv/dt and static level have independent switching means.

    passives protect node from voltage while activce components react.

    Patent is granted.

  8. vasanjk
    October 24, 2014

    Thomas, If you could provide some case examples based on application areas such as automotive, medical etc, the application tips would be immensely useful. Thx.

  9. sreekharr
    June 10, 2015

    This post I an eye-opener for me

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.