Low Capacitance Tranzorb in the 10-30V range
in [Design]
|
Prev: motor driver boards
Next: Soundcard with 3rd octave eq
From: mook Johnson on 30 Jan 2010 19:05 On 1/30/2010 8:24 AM, Nemo wrote: > Keep in mind the C varies with applied V across the tranzorb. They > sometimes quote C at the breakdown voltage, where it's much lower. these are rated at 0 volts. usually gets lower with higher voltage.
From: mook Johnson on 30 Jan 2010 19:06 On 1/29/2010 2:21 AM, Archimedes' Lever wrote: > On Thu, 28 Jan 2010 21:52:20 -0600, mook Johnson<mook(a)mook.net> wrote: > >> On 1/26/2010 8:34 PM, Joerg wrote: >>> mook johnson wrote: >>>> Gents. >>>> >>>> I'm looking for a tranzorb or equivilent that can discharge a 45mH >>>> inductor with 1A DC flowing through it if one end gets disconnected. >>>> This tranzorb must have less than 100pF of capacitance. The tranzorb >>>> will be in parallel with the inductor. >>>> >>>> any thoughts? All if the 1.5KE like parts have 800pF and up. >>>> >>> >>> Not sure what you want to do but if the inductor is pulled down, held >>> and then flies back until it leans into the TVS maybe the trick in >>> figure 1b works: >>> >>> http://www.microsemi.com/micnotes/122.pdf >>> >>> But keep in mind that TVS devices are meant to snuff the occasional >>> spike. Regular load dumps out of 45mH might be responded to with a loud >>> *PHOOMP* and some debris flying about. The last time I did a >>> (transistorized) load dump for a roughly 20mH coil at a few amps the >>> dump part needed to be heat-sinked. Without that it would have soldered >>> itself right off the board. >>> >> >> Found some of the SA18 parts at digikey. Looks like the trick they use >> is to put a diode junction in series with the high capacitance zener >> junction. As long as the zener is not conducting the diode junction >> looks like a small capacitance in series with a large one so the net is >> a small capacitance. >> >> The downside of course is you now need two devices do bidirectional. >> >> no free lunch again but this will work since I only need to do this in >> one place. :) >> >> >> Thanks for the input guys. >> >> >> > You can buy the dual device in a single package. Do you have a part number handy?
From: Joerg on 30 Jan 2010 19:09
mook Johnson wrote: > On 1/29/2010 4:22 PM, Joerg wrote: >> mook Johnson wrote: >>> On 1/26/2010 8:34 PM, Joerg wrote: >>>> mook johnson wrote: >>>>> Gents. >>>>> >>>>> I'm looking for a tranzorb or equivilent that can discharge a 45mH >>>>> inductor with 1A DC flowing through it if one end gets disconnected. >>>>> This tranzorb must have less than 100pF of capacitance. The tranzorb >>>>> will be in parallel with the inductor. >>>>> >>>>> any thoughts? All if the 1.5KE like parts have 800pF and up. >>>>> >>>> >>>> Not sure what you want to do but if the inductor is pulled down, held >>>> and then flies back until it leans into the TVS maybe the trick in >>>> figure 1b works: >>>> >>>> http://www.microsemi.com/micnotes/122.pdf >>>> >>>> But keep in mind that TVS devices are meant to snuff the occasional >>>> spike. Regular load dumps out of 45mH might be responded to with a loud >>>> *PHOOMP* and some debris flying about. The last time I did a >>>> (transistorized) load dump for a roughly 20mH coil at a few amps the >>>> dump part needed to be heat-sinked. Without that it would have soldered >>>> itself right off the board. >>>> >>> >>> Found some of the SA18 parts at digikey. Looks like the trick they use >>> is to put a diode junction in series with the high capacitance zener >>> junction. As long as the zener is not conducting the diode junction >>> looks like a small capacitance in series with a large one so the net >>> is a small capacitance. >>> >> >> The capacitance looks huge, figure 4: >> >> http://www.fairchildsemi.com/ds/SA%2FSA170CA.pdf >> >> [...] >> > > > My bad its a SAC15 from litelfuse. > Ah, with diode built in. But you've got to use two per line, or a diode in the other direction if unipolar. http://www.littelfuse.com/data/en/Data_Sheets/Littelfuse_TVS-Diodes_SAC.pdf -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM. |