:: anatomy of a ESD discharge
after reading much into Mr JOEQSMITH's adventures into surge testing
of multi-meters in EEVBLOG, i did a little research into energy levels
concerned with a human body ESD discharge.
there are many articles in the web, but there isnt really a good gauge of how many joules we are talking about. but this is mainly due to the fact there are just too many real world variables.
most ESD discharge models the first 30ns, which is where nearly all of the "deadly" energy is released. which is according to IEC 61000−4−2 specs. in general, i will focus on the 8kV/30A model.
equivalent energy containment @ 8kV IEC 61000−4−2
15% peak estimate - 0.6mJ = 2uF@25V / 50uF@5V
50% peak estimate - 3.6mJ = 22uF@25V / 300uF@5V
with this in mind,if simple capacitors are used, they should be of very low ESR and focused around 250Mhz to 450Mhz.
if we then consider AC circuit 50Hz 230VAC. 1 cycle = 0.02 second, @ 10A, per cycle = 0.2coulomb. energy per cycle = 0.2 x 230 = 46J.
with safety oversizing x5, 46J x 5 = 230J.
example of actual protection components :
TVS = P6KE440A x2 in series (600W peak, 840V peak, 1A peak)
MOV = ERZ-E07A431 (60J rating 275VAC, 2500A peak)
MOV = ERZ-E11A511 (190J rating 320VAC, 6000A peak)
MOV = ERZ-E14A561 (382J rating 350VAC, 10000A peak)
PTC fuse = LVR125K-240 (12A peak, 1A hold, 265VAC max operating)
the TVS diode appears to be so low in peak current handling but the proper way to use TVS diodes in high power surge containment is to stack up high current low voltage TVS onto large dissipation planes
https://www.microsemi.com/document-portal/doc_download/14625-series-stacking-tvs-devices-for-higher-current
824500401 (400W peak, 64.5V peak@6.2A surge). series string 40WV x 26 = 1040V p-p. peak handling per device = 3W, total = 78W. cost = 26x 0.252 =USD$6.6
SMBJ48CA-13TF (600W peak, 77.4V peak@7.7A surge). series string 48WV x 22 = 1056V p-p. peak handling per device = 5W, total = 110W. cost = 22x 0.27 =USD$6
SMBJ17CA-13TF (600W peak, 27.6V peak@21.7A surge). series string 17WV x 43 = 731V p-p. peak handling per device = 5W, total = 215W. cost = 43x 0.27 =USD$11.6
SMCJ40CA-TR (1500W peak, 64.5V peak@24A 10/1000us surge). series string 40WV x 18 = 720V p-p. peak handling per device = 7.3W, total = 131W. cost = 18x 0.4 =USD$7.2
SMCJ18CA-TR (1500W peak, 29.3V peak@53A 10/1000us surge). series string 18WV x 40 = 720V p-p. peak handling per device = 7.3W, total = 293W. cost = 40x 0.4 =USD$16
so the SMBJ type seems to have the best cost efficiency in terms of package power dissipation vs total configuration possible.
http://www.littelfuse.com/~/media/files/newsletter/article%20documents/top_10_circuit_protection_considerations.pdf
but as we all know, diving into the many modern cheap electronics today, we will not find much protection in them. the first line of best protection we can have is the fuse in your AC plugs. make sure to size them correctly as all come in default 13A which is too high for most things.
there are many articles in the web, but there isnt really a good gauge of how many joules we are talking about. but this is mainly due to the fact there are just too many real world variables.
most ESD discharge models the first 30ns, which is where nearly all of the "deadly" energy is released. which is according to IEC 61000−4−2 specs. in general, i will focus on the 8kV/30A model.
equivalent energy containment @ 8kV IEC 61000−4−2
15% peak estimate - 0.6mJ = 2uF@25V / 50uF@5V
50% peak estimate - 3.6mJ = 22uF@25V / 300uF@5V
with this in mind,if simple capacitors are used, they should be of very low ESR and focused around 250Mhz to 450Mhz.
if we then consider AC circuit 50Hz 230VAC. 1 cycle = 0.02 second, @ 10A, per cycle = 0.2coulomb. energy per cycle = 0.2 x 230 = 46J.
with safety oversizing x5, 46J x 5 = 230J.
example of actual protection components :
TVS = P6KE440A x2 in series (600W peak, 840V peak, 1A peak)
MOV = ERZ-E07A431 (60J rating 275VAC, 2500A peak)
MOV = ERZ-E11A511 (190J rating 320VAC, 6000A peak)
MOV = ERZ-E14A561 (382J rating 350VAC, 10000A peak)
PTC fuse = LVR125K-240 (12A peak, 1A hold, 265VAC max operating)
the TVS diode appears to be so low in peak current handling but the proper way to use TVS diodes in high power surge containment is to stack up high current low voltage TVS onto large dissipation planes
https://www.microsemi.com/document-portal/doc_download/14625-series-stacking-tvs-devices-for-higher-current
824500401 (400W peak, 64.5V peak@6.2A surge). series string 40WV x 26 = 1040V p-p. peak handling per device = 3W, total = 78W. cost = 26x 0.252 =USD$6.6
SMBJ48CA-13TF (600W peak, 77.4V peak@7.7A surge). series string 48WV x 22 = 1056V p-p. peak handling per device = 5W, total = 110W. cost = 22x 0.27 =USD$6
SMBJ17CA-13TF (600W peak, 27.6V peak@21.7A surge). series string 17WV x 43 = 731V p-p. peak handling per device = 5W, total = 215W. cost = 43x 0.27 =USD$11.6
SMCJ40CA-TR (1500W peak, 64.5V peak@24A 10/1000us surge). series string 40WV x 18 = 720V p-p. peak handling per device = 7.3W, total = 131W. cost = 18x 0.4 =USD$7.2
SMCJ18CA-TR (1500W peak, 29.3V peak@53A 10/1000us surge). series string 18WV x 40 = 720V p-p. peak handling per device = 7.3W, total = 293W. cost = 40x 0.4 =USD$16
so the SMBJ type seems to have the best cost efficiency in terms of package power dissipation vs total configuration possible.
http://www.littelfuse.com/~/media/files/newsletter/article%20documents/top_10_circuit_protection_considerations.pdf
but as we all know, diving into the many modern cheap electronics today, we will not find much protection in them. the first line of best protection we can have is the fuse in your AC plugs. make sure to size them correctly as all come in default 13A which is too high for most things.
Comments
Post a Comment
feel free to ask me anything