rookie inquiry : tantalum or electro?

[plexi12000]

we all know 'lytics wear out eventually....perhaps 10-15 yrs, or whatever.  curious about tantalums and welcome any insight.  for a hobbyist, cost isn't really a huge factor, i suppose.

so, if tantalums don't wear out...and are "superior"....why use electro's at all??

[R.G.]

Tantalums *are* electrolytics, they just use tantalum foil instead of aluminum foil inside.

They offer some advantages, but are expensive, and have a habit of dying catastrophically if they're ever maltreated.

[duck_arse]

I was given a coupla bags of tag and tad tants, back when. about 2004 I used some building a theremin, and it took me some time to find it was them introducing noise/jitter to the supplies when they were meant to be bypassing. so I tested the batches, and chucked most of them.

do they have a non-catastrophic breakdown mode?

[edit:] spelling, also, we used to turn off and say "stand back" when we found the reversed clamp-mount electro in the disc-drive supply boxes. the ceiling had a few vent plugs stuck in it.

[bloxstompboxes]

Electrolytics , the amluminum foil variety, can die catastropically as well. I work in a nice factory that manufactures amps and infotainment for the autmotive industry. A coworker was analyzing a 12 channel amp and when he powered it up, something bad happened. A cap on the power line had been inserted with wrong polarity, which was the cause of the initial test failure. The tester shut down the test before anything happened. He didn't shut it down fast enough and it burst and "spewed," as he put it, its contents in his face. Luckily, it didn't get in his eyes.

[amptramp]

Tantalum electrolytics were the darling of the industry when they were invented.  There are several types that have different characteristics.  Dry slug tantalum is what you usually run into.  Wet slug tantalum is another variant that has gone almost extinct.  Tantalum-case tantalum is also rare but still used.

The high dielectric constant of tantalum pentoxide (27) meant that devices could be much smaller than aluminum electrolytics with a dielectric constant of 9.3 to 11.5.  Tantalum electrolytics are only made in voltages up to 125 volts due to the thin dielectric, but tantalum dioxide is pinhole-free (unlike aluminum oxide) so long-duration timing circuits can be made since pinhole-free means low leakage.  But there is a problem.  If sufficient energy gets stored in a defect in the tantalum pentoxide, heat will convert some of it into tantalum dioxide, which is conductive.  The short circuit this starts converts nearby pentoxide and you have a firestorm that covers the dielectric in milliseconds.  Solid tantalum capacitors should be used with a series resistance to avoid having anything more than stored charge to set off a defect site, so use as a power supply bypass capacitor is not recommended.  Unlike aluminum electrolytics, they do work well into the high-frequency region.  If you get military surplus ones, look for CS-13 or CSR-13 or MIL-C-39003.

Wet slug tantalum is OK for power supply bypassing because the liquid electrolyte does not have enough conductivity to carry enough current to detonate the capacitor in case of a dielectric fault.  But the equivalent series resistance (ESR) is higher and a brief reversal of voltage more than a few tenths of a volt is enough to short it due to silver dendrites from the case growing through the electrolyte.  Look for CLR-65 or MIL-C-39006 for military surplus.

Tantalum-case tantalum is what I used for a power supply filter at the input of a power converter for a payload on the Space Shuttle.  The Shuttle is a pain in the ass to design for - you have the minimum-weight battery with minimum-weight cabling going 90 feet to the payload and you have to ensure the filter has low enough impedance for the power converter because a regulated power converter is a negative resistance device - as the input voltage rises, the current taken falls and vice versa.  I recall having the impedance without the converter come to 0.313 ohms and with the converter, 0.413 ohms.  If the impedance of the filter exceeds the negative resistance of the converter, you have an oscillator.  If it is barely stable, it will not oscillate, but the transient performance is ugly.  Tantalum-case tantalum had to be used sparingly on the Shuttle because tantalum has a specific gravity of 16.69, so these beasts are heavy.  I used SYSCAP, a precursor of SPICE, to design a two-section filter and the finished item matched the simulation to within 1 db everywhere and 0.5 db at most frequencies.  As you can see, tantalum-case tantalum is ideal for power supplies - low ESR, good frequency response and the cost is OK on a program where an SPST toggle switch costs $650.  Look for CLR-69 for military surplus.

BTW some manufacturers have come out with niobium electrolytic capacitors that are slightly cheaper than tantalum, but they are only available up to a 15 volt rating.

[plexi12000]

man.......    (head spinning!)  but i love this forum...you guys know what the heck is going on.  i feel like a complete moron! ha!   big world out there.....  thanks for the info.

so i guess i'm hearing, especially for someone of limited knowledgge, i should just stick with the "regular" aluminum cans?

[bloxstompboxes]

so i guess i'm hearing, especially for someone of limited knowledgge, i should just stick with the "regular" aluminum cans?

Certainly not. It depends on the circuit and your opinion on the sound. Many say that electrolytics sound terrible in the signal path, but they are used. A tantalum might do well in its place. They can also, as I think it was mentioned already, be a space saver. Use whatever works and whenever it is most convenient. It can be a bit of a judgement call sometimes.

[Jdansti]

+1

If the circuit calls for an electrolytic in the signal path, I use a tantalum if I have one.

[greaser_au]

Electrolytics , the amluminum foil variety, can die catastropically as well. I work in a nice factory that manufactures amps and infotainment for the autmotive industry. A coworker was analyzing a 12 channel amp and when he powered it up, something bad happened. A cap on the power line had been inserted with wrong polarity, which was the cause of the initial test failure. The tester shut down the test before anything happened. He didn't shut it down fast enough and it burst and "spewed," as he put it, its contents in his face. Luckily, it didn't get in his eyes.

Been there, got the t-shirt!  (except it was me that put the electro in backwards... ) Pinball machine, main 50V solenoid supply - the PSU PCB was in the headboard with the capacitor at about my head height.  I was standing at the other end of the machine when I turned it on, took about 30 seconds to explode-  copped the whole lot in the face (from 2 metres (over 6') away.     Fortunately this was in the workshop and not in some pub or amusement centre...

They say the best teacher is experience...

david

[greaser_au]

Plenty of capacitor threads on here (and on other sites) with LOTS of differing opinions. Some interesting discussion and tests on Steve Bench's page 'The Sound of Capacitors'.

There are a couple of links to specialised pages  from that page on ceramics and electrolytics.

david

edit: for spelling! so embarrassed!

[slacker]

With regard to electros or tants in the signal path, most stompbox designs that used them did so because that was the only way to get that value cap cheaply enough and or physically small enough. These days there's very little need to use them because we can get film or multilayer ceramic in small enough packages, film might cost a bit more but for hobbyists not enough to worry about.

[Mark Hammer]

There was a time when:
a) aluminum electros were big and clunky
b) tants bent over on their sides more easily
c) stompswitches were much bigger
d) small pots were generally unavailable

If you look at older MXR pedals, the use of tantalums was often simply wo be able to fit the board in a 1590B, behind those bigass pots and switches.  Maybe they have a context where they "sound" better, and maybe they don't.  I find it remarkable how many desirable effects use ugly old ceramic discs, and it doesn't hurt their tone one little bit.

[bloxstompboxes]

I find it remarkable how many desirable effects use ugly old ceramic discs, and it doesn't hurt their tone one little bit.

Just look at the kitrae site, I think it is. Pictures of old big muffs had all big hunkin' ceramic caps everywhere.

[Morocotopo]

I was given a coupla bags of tag and tad tants, back when. about 2004 I used some building a theremin, and it took me some time to find it was them introducing noise/jitter to the supplies when they were meant to be bypassing. so I tested the batches, and chucked most of them.

Damn!! i just installed like, 12 of them as power supply caps... 

[duck_arse]

I hadta hang them one by one on the psu, and watch the cro. sometimes the voltage would drop slightly, then just jitter about a bit. I'd tried them all as timing caps in an oscillator circuit, then psu-ed any that were reading about right value.