I recently came across an interesting story that involves the tale of a boiler system that failed to start under the close supervision of a TDK-Lambda rep, only to be solved by none other than the building maintenance technician. The story starts by the shutdown of an aging boiler in the dead of winter a British winter that required replacement of the seven-year-old radiators. After the repair and replacements took place, the boiler was switched on, but failed to start.
After hearing talk of additional cost and time, and not wanting to wait any longer for his building’s heat to return, the technician climbed up to the loft with the TDK-Lambda rep to inspect the problem. Upon being shown the suspect board, the technician brought up a small electric heating fan, directed it to the power supply, and within twenty minutes the boiler was operational.
So why was the problem remedied with such a trivial solution? The answer lies in the power supply’s small diameter electrolytic capacitor.
Often times, in higher voltage power supply applications, it is practice to rely on larger aluminum electrolytic capacitors ranging from 10mm and even surpassing 50mm (hot dog, that’s a big cap!). These larger caps are, not surprisingly, well suited for these applications due to their excellent storage capabilities, as well as their ability to reduce output voltage ripple.
Additionally, these caps provide enough storage to keep the power supply running smoothly in the even of a short AC interruption, which makes sense when the reliability of a boiler is the goal. You certainly wouldn’t want these suckers going down in the dead of a frigid British winter.
These caps, however magnificent (and large), still have their limitations. Diving into manufacturing data sheets, we can find key specs that reveal their true strengths, as well as their true weaknesses.
As many folks would, the lifetime numbers for these caps are what is most commonly taken into consideration. Over time, with these larger capacitors, the concern turns to reliability at elevated temperatures. Electrolytes are lost in diffusion of the rubber seals causing a loss in capacitance and an increase in ESR.
For these capacitors, this reduction may reduce power supply hold up times and as they continue to age, the output ripple voltage could increase to the point of an unstable power supply. Obviously, this would be a concern for any system, let alone our chilly, British, Good WIll Hunting boiler building.
The remedy for this market demand was to respond with a long-lasting cap at higher temperatures. Additionally, we often overlook the importance of the capacitors in our startup and housekeeping circuitry. So with these economically priced, longer lasting capacitors, we’re bound to see a better performing system overall, right? Not quite.
Trade offs work well when trading size for performance… sometimes.
Overlooking a few nuances in design can sometimes still result in a positively performing system. But a few too variables too many, and you could end up with faulty performance.
When we take a look into the datasheets of these long-lasting capacitors, we can see that, for example, a 12.5mm capacitor might see a lifetime stat of around 10,000-hours with temperatures ranging up to 105°C. In the same series, a 6.3mm diameter cap at the same temperature might only see a 4,000-hour lifetime stat. In our boiler building, running 24 hours a day, that’s only a 6-month lifespan. At that point, the capacitor may, at best, be worth only 75% of its original value.
Additionally, it’s not often we see these smaller caps near the main power transformer, and as well in a hot environment, hence our overlooking these data sheet values.
The start-up circuitry was very near and dear to the boiler which would allow for higher temperatures to have their way on the capacitor. This in effect would substantially lower the voltage value over the several years that this boiler was in operation.
This might not have been an issue if the boiler continued to run, however, knowing that power supplies kickstart themselves only after a certain voltage threshold is reached by the preceding capacitors (with the already lowered value of the cap at the higher temperature conditions), it would make sense that trying to restart this power supply in the dead of the winter cold wouldn’t work so well.
Enter boy-genius, Matt Damon.. err.. Mr. Building Maintenance Man with his solution to simply heat the cap up enough to where the proper voltage threshold was reached for the power supply to then find itself again and become, again, fully operational. My hero.
As fun and exciting as this mystery was, the question remains unanswered as far as what we can learn from all of this and how we can improve designs for future use. For starters, I’d recommend hiring any savant-like technicians that will be dealing with your designs. This will ensure a critically acclaimed level of ingenuity.
After that, we can be certain that overlooking small details (especially in ‘economical’, long lasting capacitors) within our power supplies and distribution networks should absolutely be taken into consideration. Check thy data sheets. Know thy environmental conditions. Don’t skimp out just to save some bucks. You may just find yourself in the depths of winter with a confused TDK-Lambda rep.
Electrolytic capacitors can sometimes be as confusing as Ph.D. level mathematics.
Additionally, utilizing power distribution network analysis (PDNA) software, we can run our power designs through the ringer of a range of testing conditions to emulate what the real world can (and will) throw our way.
Fantastic software, such as Altium’s PDN Analyzer™ , is phenomenal for analysis such as this. Giving you insights as to where potential marginal voltage values exist could just save your PCB from us writing an article about it in the near future.
To hear more examples, or learn more about how Altium can save your , talk to an Altium expert today!