It sounds like one of tantalums advantages 'low esr' is it's biggest failing point as well? Electrolytics would self regulate their current because the ESR is higher? Am I getting the rough idea here right?
I've never seen any reason to trust them?, and the lack of them now in anything I ever see seems to prove me right. Perhaps tantalums are more reliable now?, but domestic electronics manufacturers don't seem to be convinced after the previous disasters!.
I'm also somewhat dubious about using components with have a specific rating, then a warning in the datasheet not to exceed 1/3 of that rating - it seems a poor way to specify a component!.
Been telling you- their reliability is now well understood for different temps and voltages. They're used in many high end apps so their process control is quite well developed.
I'm not sure who "domestic mfgs" are nowadays since virtually everything is outsourced, but mfgs DO widely use them. They tend to be a bit more expensive and less available in throughhole pkgs. Mfgs care a LOT about cost. So big stuff (TV sets) probably won't use them. They're like gold to MP3 player, hard drive, laptop motherboard designers.
The lifespan of a tantalum cap is generally regarded longer than an aluminum electrolytic. I get the idea its window of expected lifespan and ESR rating is easier to anticipate too.
Tektronix has used tantalums in their equipment since around 1970, usually as bypass caps in an RC circuit with maybe a 10 ohm resistor between supply and cap/load and for timing caps for low-freq/long PW one-shots and timers. Tantalum caps didn't fail any more often than ICs, transistors or diodes. A company like that doesn't intentionally use unreliable components and if they do find that a component is unreliable, will quit using them and issue a modification order to swap out any that come into a service center. In my hobby work, I've never had a tantalum failure.
I wouldn't say that; it would be very bad design to rely on the ESR to limit the ripple current.
I haven't had any problems with tantalums. I thought they wer more reliable than electrolytics until I came here. Perhaps it's just that one manufracturer produced a load of poor quality components.
One guy says he doesn't like how tantalums sound in his system?
That problem has nothing to do with tantalums but rather his lack of understanding what is actually desired in his system (impedance -wise)
The context of our discussion is around the reliability aspect. Performance is what seems they are after. Impedance is impedance whether is comes from AL-elec or Tant-elec or ceramic for that matter.
It sounds like one of tantalums advantages 'low esr' is it's biggest failing point as well? Electrolytics would self regulate their current because the ESR is higher? Am I getting the rough idea here right?
Tantalum capacitor has an ESR ten times better than that of electrolytics
On a previous comment by someone else, informities that develop/part of the manufacturing process actually 'self fill' and hence they never loose capacitance, therefor its not just their shelf life, but their operating life that is considered 'infinite'
You misread my statement, electrolytics have a higher equivilant series resistance, so in pulse charging and discharging applications they will self limit their own charge/discharge rates because of the inherant resistance. A tantalum cap because of it's better (lower ESR) doesn't 'self limit' it's current. EL's and tants might have similar ripple current ratings, but a significantly lower ESR for the tant would make it more prone to self destruction from instantanious pulse heating when it was used in a low impedance situation (such as a filter cap which is usually connected via a VERY low ohmic connection to ground) Out of every post I've read so far that's the only thing that makes sense to me. If someone knows something else they better start spilling the beans =>
I wouldn't agree, if the ESR is lower then the power dissipation will also be lower and it's the job of the designer to ensure that the ripple current isn't exceeded which isn't a hazard in 99% of applications anyway.
Overcurrent is only a probelm when a capacitor is used as a mains frequency smoother in a power supply, when connected in series with a speaker in an audio amp or used in power factor correction - a 10:mu:F capacitor isn't going to have it's ripple current rating exceeded however low its ESR may be.
I just thought I'd post since this is so old and Nigel could be misleading people looking for answers about tantalums.
Maybe back when dinosaurs roamed the Earth when you were using them they were unreliable, but today they are very reliable. No less than five times have I taken apart a device certified for use in life support applications to find them littered with tantalums. I used them in college for statistics experiments to verify the claimed reliability of the manufacturers only to find every single time that it was right. Trials were conducted over several months with varying as well as extreme temperature and voltage gradients.
I even attached a picture of an RC receiver using tantalums. This module has been in service for several years in the bitter cold and blistering heat. I'd say it was a fluke if I knew as little about tantalums as you seem to but the reality is that they are better devices than their aluminum electrolytic counterparts in many applications. Your inability to get past your ego is quite irritating when all of the facts point the other direction from your opinion. That is not how science works.
I've been a board-level and chip-level hardware engineer for 15 years (including internships). For the last 13 years, I have only rarely used electrolytics and mainly used them on sensitive analog circuitry where I was not the original engineer and was afraid to replace them in case some special quality was required or where I needed very high value capacitance that I could not get from an SMT device. (One additional factor to consider, by the way, in terms of cost is the cost of manufacturing with through-hole components verses surface mount.) In these 13 years, I designed tantalums into products ranging from medical devices to computer peripheral devices to networking gear. Some of these were very high volume applications. I saw one instance in which we had some failures on prototypes. A colleague had designed in 12V rated tantalums for a 12V power supply circuit. I agree with other posters who recommend taking de-rating of tantalums very seriously. It is in their data sheets. We switched to 25V rated caps (due to the recommendation to de-rate by something like 40-50%; don't recall exactly) and had no further problems. Perhaps the one thing I agree with Nigel about is that it is misleading and silly to call a device a 25V cap if you should really not use it past 12V or 16V or whatever, but that's how it is in the era of marketing.
I became a member of this site just now so that I could post a reply to this thread because I am bothered by how much opinion there is with very little fact. I came accross this forum while doing a search for information about whether tantalums really are unreliable since I was surprised -- upon moving to the UK (Yorkshire, not far from Derbyshire where Nigel is posting from) from the USA -- to find that my new company here mainly avoids them due to reliability. So far, I have come accross no good evidence that they are less reliable, and on the contrary, lots of evidence that they should be more reliable than electrolytics. So, if you have any real evidence or your own actual experience one way or the other, please post it here. Thanks for some of the better posts already on this thread. Hopefully, the truth shall set us free (or at least provide some guidance).
PS: The following seems to be a pretty good site regarding reliability of different types of caps and lists shortcomings of both aluminum and tantalum electrolytics. (Tantalum discussion is way at the bottom so scroll down...)
I only picked up this discussion today after googleing the subject. I had a 10u 16vdc tant blow up in a 100watt rf amplifier and wonder if a higher voltage electrolytic would do the job of de-coupling the pa bias supply. The PA supply is 13.8vdc; voltage on the tant caps ~0.6vdc shouldn't blow!
I have many computer SMP they don't use tants only low ESR Electrolytics. I assume that the earlier points still apply.
I only picked up this discussion today after googleing the subject. I had a 10u 16vdc tant blow up in a 100watt rf amplifier and wonder if a higher voltage electrolytic would do the job of de-coupling the pa bias supply. The PA supply is 13.8vdc; voltage on the tant caps ~0.6vdc shouldn't blow!
I have many computer SMP they don't use tants only low ESR Electrolytics. I assume that the earlier points still apply.
16 Volt Tantalums are Too Low in voltage, for a 13.8 volt supply.
You Should use at least 25 volt ones or even better, 35 Volt Tantalums.
10uF, Low ESR Electrolytics rated at 25 volts or higher, or .1uF Disk Caps will also work, but Tantalums are better.
It Isn't just about your Supply Voltage, Its more about the RIPPLE CURRENTS that are in the circuit!