GoodSoundClub - Romy the Cat's Audio Site

In the Forum: Melquiades Amplifier
In the Thread: Single-stage Melquiades vs. DHT amps

Post Subject: Electrolytics, technology and timePosted by floobydust on: 5/17/2009

Paul,

Some good valid points... yes, they've gotten much better over the years, no doubt to technology improvements. A good example is Mallory CGS caps made in the 80's versus new ones. The newer ones are much lighter physically (same p/n), hence they have been changed internally. In any case, they still have a life span and their actual value tolerance is anything but close with a +80/-20%. I ordered qty. 100 CGS caps from a Mallory distributor in 1988. To date, virtually all have failed, many within 1-2 years sitting the box! Needless to say, all from the same batch and an obvious manufacturing defect. Problems varied from severe leakage (damaging equipment they were in), to dead shorts (causing other component failures) to completely open circuit. As for voltage ratings, I've rarely seen any electrolytic capacitor rated at more than 600 volts (Sprague Atoms). Computer grade types are generally limited to 450 volts. If you know of some that exceed that, please post which ones they are. I would also point to many designs that use them improperly.... the forums are full of scenarios where caps have popped, gone leaky or failed in multiple ways and took other circuitry out with them, electrically and physically. Note that in my design, I used 450 volt CGS caps and they never had more than 300 volts present and to date, every piece of gear I built with them needed replacement and/or repair. In any case, I avoid electrolytics for the above quality-related issues and eventual damage and/or repair requirements. Unfortunately, for DC filament supplies, you really have no other option.

As for film types, they can have issues as well, especially as you get into the larger values and voltages. You can get film caps in 630 volts and higher, I've seen upwards of 800 volts in larger values, like 40uf or higher. Exceeding this voltage could damage them as could accidently shorting them. The large physical size is also an issue and can be a problem with mechanical vibrations, etc. The better ones use high-tension on the winding so they don't exhibit mechanical excitation, but in short, not all caps are the same. Ensuring they are sealed so they don't become contaminated is also another problem. Still, I'll take my long-term chances with them over electrolytics. One thing is certain, they won't leak and they should not have a limiting lifespan and their tolerance is usually within 5%.

Oil types, motor-run and industrial lighting caps have also been used. I've seen problems with the latter as well... becoming leaky internally and getting hot leading to failure. One thing I don't like in designs is paralleled capacitors of different types (electrolytic, film, oil, etc.) to yield a "better capacitor". I consider this as the "cure is worse than the cause". In reality, components are not ideal in their function as they are not perfect in operation. If they were , there would be no reason for having so many different types optimized for specific applications. By paralleling multiple types of capacitors you effectively create small "tank circuits" which react to signal. I've actually seen some DIY designs where every capacitor in the circuit consisted of 3 individual caps of different types in parallel. It is my personal preference to keep it simple and use a single capacitor, albeit which capacitor becomes a personal choice for personal reasons.

As for break-in, I'm not too certain that perceived sonic changes in equipment are the result of film caps needing to "break-in", but everyone has their own views and experiences. My view is also from a simple prespective. If the materials used in the construction of the component are effectively inert in operation, then you really can't have a sonic signature change as a result. Electroytic capacitors do change internally over time. A film cap (by definition) should not. The same applies to other components in the equipment. Some resistor types can change from time and heat, wire and component leads can oxidize over time and of course, the tubes themselves change emission and other characteristics over their usable life. I do try and minimize this by prepping every single component used. All component leads are buffed shiny to remove all oxidiation and crud and then a very tight teflon sleeve is "spiraled" on it to seal it so it won't oxidize over time.

In any case, I've typed too much again... sorry.

Regards, KM

Rerurn to Romy the Cat's Site