A bit pricey (6 times cost factor) but would these have decent enough ESR for power supply rail noise filtering? Caps are 10uf @ 35 volts

Pete

LED lighting designers are doing all they can to eliminate electrolytics and use ceramic. Electrolytics are largest cause of failures.



https://www.edn.com/electronics-blogs/p ... LED-lights

Rich

Unfortunately there are to many characteristics that are unknown to make a solid recommendation. For example, electrolytic do not change capacitance as the voltage rises. Class 2 ceramics, or worse, vary greatly with applied voltage. Given the stated capacitance and voltage rating it's probably around an 0805 X5R, so the capacitance will be ~1uF at 30V. There are many other characteristics differences between the dielectrics. You can view these on sites like Murata and Kemet.

Darryl

That may be what some EE bloggers think, but they probably aren't the ones designing LED light bulbs. Actually, some of the very best electrolytics ever produced to date are now coming out of China specifically for ... drum roll ... LED lighting. The small uF, 450-VDC electrolytics being made for LED lighting today are physically tiny for their ratings, have extremely low dissipation and leakage, deal with the surges and spikes inherent in line operated circuitry, tolerate the elevated temperatures found in these applications, and still give 100,000 hour life expectancy. Values are in the range of 8-uF to perhaps 47-uF, making them excellent for restoration of vintage tube electronics, BTW. So electrolytics are not being eliminated from lighting, they're being improved.

As for Pete's original question, it's impossible to say without knowing the details of the application. MLCC caps are very different than electrolytics when it comes to things like resonant frequencies, self inductance, ESR, and Q. Those things may matter in some circuits, maybe not in others. In SMD boards that handle digital data at speeds above a few hundred kHz, the traces on the PC boards are often tuned to be resonant or anti-resonant at frequencies of interest, and component selection therefore becomes critical to maintaining the integrity of the original design. In other words, mess with it at your own peril.

No, actually it is what serious designers are thinking. Driver IC vendors have app notes on designing with ceramic caps and are building higher frequency drivers, so there is less need for high capacitance.

https://www.maximintegrated.com/en/app- ... vp/id/4669




https://www.analog.com/media/en/technic ... 3474fd.pdf

https://www.ti.com/lit/ds/symlink/tps61181a.pdf


https://ieeexplore.ieee.org/document/6953765



I'd welcome any documentation from "the ones designing LED light bulbs." I suppose if you are designing a $3 light bulb, electrolytics are still the norm?

Rich

As noted, the SMD caps are used on regulated power supply rails. I suspect the main purpose is digital noise filtering and isolation. The first series of Tek gear that used early SMD electrolytics are all showing signs of PC run damage due to out gassing from failing seals. When you lose a multilayer board that costs up to four hundred bucks for a board removed from a junker it can be painful. General practice, from what I have seen, is these are mostly being replaced with radial lead high quality, low ESR 105 degree Panasonic caps. It would seem monolithic ceramic caps would be ideal for this application, even though they cost ten times as much. Q (which can affect closed loops) and some of the other issues might be a bit moot for this application.

You do realize that your Linear Technoloy document is from 2005, and the Maxim one is from 2009? You are talking about parts that are probably EOL, maybe no longer available. The 2014 IEEE paper and the 2016 TI spec sheet appear to be aimed mainly at white screen LED illumination for notebooks and similar kinds of things where the higher cost of ceramics might be justified against size and the value of the equipment they go into. Hardly a defense of your blanket statement that LED lighting designers in 2019 are trying to get away from electrolytics as fast as possible.



The problem with a lot of SMD electrolytics in the late 1990s and early 2000's was that the standard flux removal and cleaning solvents used on PC boards in manufacturing would attack and weaken the seals of SMD electrolytics. Eventually the electrolyte would leak out and do lots of damage. The problem was recognized and corrected but not before a lot of capacitors were ruined and a lot of equipment failures occurred. While any electrolytic cap could in theory leak, modern SMD caps on the whole are better constructed and less likely to do so. Thus the need to find alternate types of capacitors is limited.

In the broadcasting neck of the woods we would just replace the old SMD caps with new ones and things would be fine again as long as the PC boards were not damaged. If you've got historical evidence that indicates Panasionic radial electrolytic caps will work successfully on the Tek boards, and those are easier to deal with, then they could be used instead. As for MLCC caps, it is very possible they could work, but there's no way to know for certain without actually trying them on a board and then running the scope through a complete calibration check to see if it really still meets all its specs. If it does, you've got another way to fix those boards.

Fair enough. Give me your list of references for the continuing preference of electrolytics! I guess designers of $3 light bulbs don't usually write papers or app notes?

Rich

https://www.psma.com/sites/default/file ... eplace.pdf

http://www.newelectronics.co.uk/electro ... st/116099/

I would expect app notes from ceramic capacitor companies to extol the virtues of their products and point out the shortcomings of others. But ceramic capacitors can and do fail too. Nothing is perfect.

As for the $3.00 LED light bulb, never knock a product that sells by the billions worldwide! Unfortunately I cannot offer you any references to look at; the high reliability electrolytics I mentioned are too new to have been extensively written about yet. There may be something in Chinese, or locked away in some obscure trade journal somewhere, but I don't have it. Nevertheless, you'll soon start seeing 450-volt electrolytics with excellent specs and much higher rated lifetimes on the market, which are the outcome of LED bulb research.

Pete is probably wondering what all this has to do with his original question of replacing SMD caps on an oscilloscope board with multilayer ceramic ones. One point is, that while chip ceramic caps do have advantages over electrolytics in some respects, they are not drop-in replacements with identical secondary characteristics. None of Rich's papers cited above suggest that they are. In order to obtain the benefits of ceramic capacitors, a circuit should be designed for them in the first place. That does not mean you can't make substitutions, only that you'd have to study and test each application on a case-by-case basis to see how well it would work out. In a piece of test equipment, specs for noise, frequency response, and sensitivity are often directly related to the ESRs of the caps in the power supplies. Substitute the caps for some other type than originally used and it might not be possible for that test instrument to meet one or more of its published specs.

Another point is that while some early SMD electrolytics were known for leaking electrolyte after a couple or a few years in service, the problems were recognized and solved. Electrolytic capacitors are not going the way of the dodo any time soon, in fact there has been considerable progress recently in making them better and more reliable. Thus it may not be worth going to an entirely different capacitor family and dealing with the possible issues. Replacing the old SMD caps with good quality new ones should provide far greater reliability while maintaining the original circuit performance.

As info, I checked the SMD electrolyic caps in my 497P and got lucky; They are a sealed, black encapsulated style that are not known to cause problems. Dodged that bullet. The aluminum can SMD caps on my 2467B A5 controller board were leaking and causing visible trace corrosion.

When I worked at Siliconix, we designed and made PWM chips for telecom and computer applications. We also made demo boards with a "reference" design to start engineers on the path to designing their own custom power supply. I learned that ESR, ground planes, trace lengths, and feedback loop compensation were all important factors and some of these related to component choices. Power supplies were designed for optimum reliability and cost for specific load conditions and component choices.

My references were not related to "how" to replace electrolytics with ceramic, rather to state that there is a huge movement to use ceramic instead of electrolytics in LED circuits. My friends who are power supply designers in Silicon Valley were the ones who got me interested in the topic. They don't work for ceramic cap suppliers. The trend is not a vast conspiracy against Chinese cap vendors.

Believe what you want to.

Rich

Well, what was seen as a wondrous new technology that was going to solve all our problems ten years ago turns out to be just another component that has its share of good points and bad points, and has to be used with due consideration for its limitations and characteristics in order to achieve maximum benefits:

https://product.tdk.com/info/en/products/capacitor/ceramic/mlcc/technote/solution/mlcc02/index.html
https://www.murata.com/en-us/products/leadmount/solution/case/film01
http://www.ieca-inc.com/images/Ceramic_capacitor_Failure_Mechanisms.pdf

This does not mean that the newest generation of LED lighting electrolytics are perfect either. They're still electrolytics after all, with many of the same quirks and limitations that you'd have with any other electrolytics. But if I was fixing up an old Tektronix scope with leaking or defective SMD electrolytics, I would go right ahead and put new SMD electrolytics in. I would not try to hack it to work with other kinds of caps.