Re: [neonixie-l] Ceramic output capacitors on MAX1771 converter?
Thanks Martin Very detailed. Much appreciated. Ian > On 20 May 2022, at 19:00, Dekatron42 wrote: > > > If I remember correctly my colleagues spoke about the normal capacitor > reforming by applying voltage to them after long storage (1-2years or more in > storage) and also the "resetting" that they are subjected to when soldering: > https://ec.kemet.com/blog/mlcc-dielectric-differences/ > > " Ceramic Capacitor Aging > Aging is another characteristic exhibited by ferroelectric, or Class II and > III dielectrics. While manufacturing the ceramic capacitor, the dielectric is > exposed to temperatures more than 1000°C. For Barium Titanate devices, the > Curie temperature can be in the range of 130°C to 150°C, depending on the > particular formulation. When exposed to the Curie temperature, the > crystalline structure aligns to a tetragonal pattern. Once cooled, the > ceramic’s crystalline structure changes to a cubic change. As this structure > changes, so does the material’s dielectric constant. > > Over time, the capacitance will continue to decline. It is possible to reset > this aging cycle by “resetting” the material, by exposing it to its Curie > temperature this usually occurs during re-flow. Typically, you can find the > aging rate in the catalog for a particular part type. Below is an example of > aging rates:" > > Our problems started when capacitors started to break and/or fall of the > circuit boards even though our designed were approved by the capacitor > manufacturer! We also experienced the problems with degrading capacitance in > MLCC capacitors but it was even worse with self-healing polypropylene > capacitors as it affected the smart-meters power supplies - it all turned out > to be manufacturing problems of the capacitors due to moisture in the > materials. > > https://www.researchgate.net/publication/348506751_Analysis_of_Failure_Mechanism_of_Smart_Meter_Capacitance > > "In practical applications, smart meters will fail, and even have batch > quality problems. In addition to the sampling inspection of the arrival > quality of smart meters, the analysis of the failure mechanism of smart > meters has become an important and long-term task. This article mainly > analyzes the component capacitance commonly used in smart meters, and > analyses the failure of chip multilayer ceramic dielectric capacitors from > appearance observation, electrical parameter testing, metallographic slicing, > and at the same time from electrolytic appearance observation, electrical > parameter testing, X- RAY, SEM and EDS for failure analysis. The results of > failure mechanism research not only help to improve the reliability of the > domestic energy meter manufacturing level, but also can save energy meter > maintenance and transformation costs, and have very important practical > significance for the construction of smart grids." > > /Martin > > >> On Friday, 20 May 2022 at 07:27:22 UTC+2 iavine wrote: >> Could someone explain what the reforming involves ? >> >> Thanks >> Ian >> On 20 May 2022, at 06:12, SWISSNIXIE - Jonathan F. wrote: >>> I've used a MAX1771 with ceramic capacitors a while ago for a >>> pandicon-circuit and a smaller nixie project. Both consume less than >>> 15mA@170V. Not sure how the circuit would behave at more current. >> >>> On Thursday, 19 May 2022 at 20:03:18 UTC+2 Dekatron42 wrote: A few other things that I experienced at my previous employer was that larger (sizewise) MLCC capacitors easily developed cracks, came loose from the circuit board and also needed reforming after storage. /Martin > On Thursday, 19 May 2022 at 05:01:25 UTC+2 mo...@neonixie.com wrote: > Roger, > > Yes. Looks like the common dielectrics (X7T, X7R) available in the 250v > 1-3uf range suffer from a 30-80% reduction in capacitance at 180v. I > hadn't realized it was that high. > Thank you, I'll look at the polymer caps. It's my last remaining > electrolytic on my board and if possible I would like to use a solid one > instead. > > Regards, > -Moses > >> On Wednesday, May 18, 2022 at 7:54:10 PM UTC-7 Roger Brinkman wrote: >> Hi Moses, >> >> What David writes below about the capacitance reduction of ceramic >> capacitors is correct. >> I have experienced success using polymer aluminium electrolytics (solid >> electrolyte) to replace conventional low-ESR electrolytic capacitors >> that regularly fail in similar applications. >> You might like to experiment with these. >> >> Best regards >> Roger Brinkman. >> On 19 May 2022, at 12:42 pm, David Forbes wrote: >>> >> >>> I haven't tried it, but I can make some observations. The MAX1771 isn't >>> connected directly to the output circuit, so it's not likely to suffer >>> from a problem. The current f
Re: [neonixie-l] Ceramic output capacitors on MAX1771 converter?
If I remember correctly my colleagues spoke about the normal capacitor reforming by applying voltage to them after long storage (1-2years or more in storage) and also the "resetting" that they are subjected to when soldering: https://ec.kemet.com/blog/mlcc-dielectric-differences/ " Ceramic Capacitor Aging Aging is another characteristic exhibited by ferroelectric, or Class II and III dielectrics. While manufacturing the ceramic capacitor, the dielectric is exposed to temperatures more than 1000°C. For Barium Titanate devices, the Curie temperature can be in the range of 130°C to 150°C, depending on the particular formulation. When exposed to the Curie temperature, the crystalline structure aligns to a tetragonal pattern. Once cooled, the ceramic’s crystalline structure changes to a cubic change. As this structure changes, so does the material’s dielectric constant. Over time, the capacitance will continue to decline. It is possible to reset this aging cycle by “resetting” the material, by exposing it to its Curie temperature this usually occurs during re-flow. Typically, you can find the aging rate in the catalog for a particular part type. Below is an example of aging rates:" Our problems started when capacitors started to break and/or fall of the circuit boards even though our designed were approved by the capacitor manufacturer! We also experienced the problems with degrading capacitance in MLCC capacitors but it was even worse with self-healing polypropylene capacitors as it affected the smart-meters power supplies - it all turned out to be manufacturing problems of the capacitors due to moisture in the materials. https://www.researchgate.net/publication/348506751_Analysis_of_Failure_Mechanism_of_Smart_Meter_Capacitance "In practical applications, smart meters will fail, and even have batch quality problems. In addition to the sampling inspection of the arrival quality of smart meters, the analysis of the failure mechanism of smart meters has become an important and long-term task. This article mainly analyzes the component capacitance commonly used in smart meters, and analyses the failure of chip multilayer ceramic dielectric capacitors from appearance observation, electrical parameter testing, metallographic slicing, and at the same time from electrolytic appearance observation, electrical parameter testing, X- RAY, SEM and EDS for failure analysis. The results of failure mechanism research not only help to improve the reliability of the domestic energy meter manufacturing level, but also can save energy meter maintenance and transformation costs, and have very important practical significance for the construction of smart grids." /Martin On Friday, 20 May 2022 at 07:27:22 UTC+2 iavine wrote: > Could someone explain what the reforming involves ? > > Thanks > Ian > > On 20 May 2022, at 06:12, SWISSNIXIE - Jonathan F. > wrote: > > I've used a MAX1771 with ceramic capacitors a while ago for a > pandicon-circuit and a smaller nixie project. Both consume less than > 15mA@170V. Not sure how the circuit would behave at more current. > > > On Thursday, 19 May 2022 at 20:03:18 UTC+2 Dekatron42 wrote: > >> A few other things that I experienced at my previous employer was that >> larger (sizewise) MLCC capacitors easily developed cracks, came loose from >> the circuit board and also needed reforming after storage. >> >> /Martin >> >> On Thursday, 19 May 2022 at 05:01:25 UTC+2 mo...@neonixie.com wrote: >> >>> Roger, >>> >>> Yes. Looks like the common dielectrics (X7T, X7R) available in the 250v >>> 1-3uf range suffer from a 30-80% reduction in capacitance at 180v. I hadn't >>> realized it was that high. >>> Thank you, I'll look at the polymer caps. It's my last remaining >>> electrolytic on my board and if possible I would like to use a solid one >>> instead. >>> >>> Regards, >>> -Moses >>> >>> On Wednesday, May 18, 2022 at 7:54:10 PM UTC-7 Roger Brinkman wrote: >>> Hi Moses, What David writes below about the capacitance reduction of ceramic capacitors is correct. I have experienced success using polymer aluminium electrolytics (solid electrolyte) to replace conventional low-ESR electrolytic capacitors that regularly fail in similar applications. You might like to experiment with these. Best regards Roger Brinkman. On 19 May 2022, at 12:42 pm, David Forbes wrote: I haven't tried it, but I can make some observations. The MAX1771 isn't connected directly to the output circuit, so it's not likely to suffer from a problem. The current flows through the inductor which will accommodate a momentary short circuit caused by the capacitor. In short, it shouldn't be a problem. Bear in mind that the effective capacitance of a modern ceramic capacitor is much lower with a DC bias near its rated voltage, so you would need to
