Post by thread:[EE]: Aluminum caps vs. Tantalum

If you mean electrolytic when you say aluminum, electrolytics are extremely lossy, have Higher inductance and ESR (which makes them respond badly to high frequency applications), and are very inaccurate. - end --- Jay Hanson <spam_OUTjayTakeThisOuTILHAWAII.NET> wrote: >When is it OK to use aluminum caps instead of the outrageously expensive >tantalum? > >Jay > >-- >http://www.piclist.com hint: The list server can filter out subtopics >(like ads or off topics) for you. See http://www.piclist.com/#topics _____________________________________________________________ Get your free email here at http://www.e-ha.com -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

> When is it OK to use aluminum caps instead of the outrageously expensive > tantalum? > > Jay When you do not care about high leakage, temperature dependence of the cap. value, and rapid failure at high temperature (<1000 hours), as well as the danger of leaking electrolyte (after 2-5years) and the effects of them drying out and losing capacity (and increasing ESR) after 3-5 years. In general in a consumer product you use Ta caps only for parts that need to be highly stable and low leakage, low esr. All the others are Al caps. In a higher end product you will probably be choosy about the make and type of each part, not only caps. Peter -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

From a list discussion a couple of years ago ISTR that there were two points to consider when selecting either foil or Ta caps For foil caps, choose the voltage rating that is nearest (and obviously above) your voltage rail. A 200V-rated cap on a 5V supply may not re-polarise if the circuit is not used for some time For Ta caps, allow a generous over-voltage margin (50%) as they are wont to fail if pushed anywhere near their voltage rating. For example use a 10V on a 5V supply, 25V on a 12V supply etc -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> For Ta caps, allow a generous over-voltage margin (50%) as > they are wont to fail if pushed anywhere near their voltage rating. > For example use a 10V on a 5V supply, 25V on a 12V supply etc Let me tighten that specification (although I will be flamed for it :-) ). Tantalum capacitors MUST have a rated voltage which is higher than the highest voltage or transient voltage which will EVER appear across them. When exposed to voltages not very much higher than their rated voltage they will fail magnificently. Failure is ultimately a very hard metallic short circuit but along the way you can hope for bad smell, noise (shrieking, moaning etc), smoke, flame and explosion. In the "best" case I have had all of these in the order listed. For applications that DEMAND the Tantalum's capabilities consider using a SOLID aluminium capacitor instead. They have most of the capabilities of Tantalum and are of similar physical size for equivalent spec but do not self destruct in the same manner and may be cheaper. Note this is SOLID Aluminium - not wet film electrolytic. Philips, at least, make them. If you want to produce a product which MIGHT smell, shriek, explode and catch fire and/or short out various parts of its circuitry at indeterminate future dates then Tantalum caps are definitely the way to go ! Because this isn't what happens in all cases and because, if you design well and are very careful and lucky you may not have any problems, you are now liable to have people telling you what fine caps Tantalums are ;-) regards Russell McMahon -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email .....listservKILLspam@spam@mitvma.mit.edu with SET PICList DIGEST in the body

> For applications that DEMAND the Tantalum's capabilities consider > using a SOLID aluminium capacitor instead. Just to be picky. A Tantalum cap _is_ a Solid Aluminium electrolytic cap. Tantalum is one type of the family 'Solid Al'. Another type is Manganese Oxide (I think) which Philips market as Solid Al, preventing other solid Al cap makers from using the term. Ain't trademarks et al wonderful. Since both are are solid Al, the performance is about the same and they age about the same. The difference is that tantalum's breakdown characteristics get worse with heat. So both ambient heat and internal heat (especially due to ripple currents) effectively reduce the operating voltage. This characteristic is also a problem when they do fail. The Philips parts will short with a bit of resistance and get really hot and stay that way (pending any protection) and maybe blow the top off. When a tant gets really hot, there is more breakdown which generates more heat which causes more breakdown and so on, leading to flames. By the time you size them with some voltage overhead, temp overhead and made allowances for ripple, you are probably just as well off with a couple of wet electros and a decent sized ceramic. Steve ====================================================== Steve Baldwin Electronic Product Design TLA Microsystems Ltd Microcontroller Specialists PO Box 15-680, New Lynn http://www.tla.co.nz Auckland, New Zealand ph +64 9 820-2221 email: stevebKILLspamtla.co.nz fax +64 9 820-1929 ====================================================== -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email .....listservKILLspam.....mitvma.mit.edu with SET PICList DIGEST in the body

> > For applications that DEMAND the Tantalum's capabilities consider > > using a SOLID aluminium capacitor instead. > A Tantalum cap _is_ a Solid Aluminium electrolytic cap. Tantalum > is one type of the family 'Solid Al'. Another type is Manganese > Oxide (I think) which Philips market as Solid Al, preventing other > solid Al cap makers from using the term. Ain't trademarks et al > wonderful. This is partly right but not quite as I understand it overall. Here's a comment on how Tantalum caps are made http://www.ncc-matsuo.co.jp/product/tantalum/gijyutue.html There are indeed process & chemistry differences between brands but Tantalum caps use the metal Tantalum and Al caps use the metal Aluminum (or Aluminium if you are down here or in UK :-) ) .Al & Tantalum are close on the periodic table and have similar characteristics but for wehatever reason the Al metal caps do not exhibit the extreme over-voltage sensitivity that tantalums do.. Here is what a friend of mine had to say in reponse to my last post - He has had substantial experience with the two types. _________________________________ Russell, Agree whole-heartedly with your comments. In the old days I used to design TC's into products - every product turned out to be a ticking timebomb. A TC always failed eventually - usually in a catastrophic manner. I have also had a lot of TC's fail in products designed by other people (PC motherboards especially). Since switching to solid Al caps, I have never had one fail despite the fact that a good many of these products have a fairly hard life in terms of temperature. Several hundred thousand capacitors can't be wrong. Regards, Ken Mardle ______________________________________ regards Russell McMahon -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email EraseMElistservspam_OUTTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body

Russell McMahon wrote: > If you want to produce a product which MIGHT smell, shriek, explode and > catch fire and/or short out various parts of its circuitry at indeterminate > future dates then Tantalum caps are definitely the way to go ! Because this > isn't what happens in all cases and because, if you design well and are very > careful and lucky you may not have any problems, you are now liable to have > people telling you what fine caps Tantalums are ;-) I've seen resistors smoke and explode a lot too, maybe we should all stop using resistors. ;o) I don't know what bad experiences you had with the tantalums but I haven't seen any trouble with them. But then I never use a 16v cap on a 16v power supply rail. -Roman -- http://www.piclist.com hint: To leave the PICList piclist-unsubscribe-requestspam_OUTmitvma.mit.edu

> Russell McMahon wrote: > > > If you want to produce a product which MIGHT smell, shriek, explode and > > catch fire and/or short out various parts of its circuitry at indeterminate > > future dates then Tantalum caps are definitely the way to go ! Because this > > isn't what happens in all cases and because, if you design well and are very > > careful and lucky you may not have any problems, you are now liable to have > > people telling you what fine caps Tantalums are ;-) > > I've seen resistors smoke and explode a lot > too, maybe we should all stop using resistors. ;o) > I don't know what bad experiences you had with > the tantalums but I haven't seen any trouble with > them. But then I never use a 16v cap on a 16v > power supply rail. Resistors fail when you exceed there specifications substantially or if they fail to meet specification. Tantalum caps fail when you exceed their specs by a very small amount. Also resistors usually miss out on the shrieking, smell much less evilly impressive and the explosion from a tantalum puts the average resistor to shame. Try it - you;; like it. Get a lab supply and a lowish voltage tantalum cap. Apply increasing voltage above rated value and see what happens. Great fun - as long as it isn't in a circuit that is meant to be reliable and useful :-). A PDF from Tantalum Capacitor maker NEC titled "Notes on the correct use of Tantalum Capacitors" makes very interesting reading. See - http://www.ic.nec.co.jp/compo/cap/english/products/c_tape/EC0332EJ1V0UM00.pdf I find the first comment in their preface interesting in view of the rest of the paper - something lost in the translation perhaps. It provides a formula and nomogram for failure rates which would make many designers blanch when they realised its significance. They provide interesting guidelines (following are as verbatim as possible but with some editing to reduce size) eg Operating voltage relative to rated voltage - General applications - 70% max - Power line or low impedance circuit 30% (50% max) - Derate voltage by 1% per degree C above 85C (many Al caps are only rated to 85 C and have very short lives at that temperature so this is not a major problem). Note that this means that, whereas an Al cap should be operated NEAR it's rated voltage, a Tantalum should be operated at about 1/3 of its rated voltage maximum or 50% if voltage transients are extremely well controlled. Also "Reliability is increased by inserting 3 ohms pere volt series resistance into switching circuits, charge-discharge circuits etc". "If the capacitor is in a low impedance circuit applied voltage should be less than 1/2 to 1/3 of rated voltage. "Do not connect two (or more) tantalum caps in series!!!" -- http://www.piclist.com hint: To leave the PICList @spam@piclist-unsubscribe-requestKILLspammitvma.mit.edu

> I've seen resistors smoke and explode a lot > too, maybe we should all stop using resistors. ;o) Late breaking thought :-) - Resistors require substantial external energy to destroy them and if run inside rated specs requires substantial energy either as a very large shortish spike or a sustained moderate overvoltage to die (provided voltage ratings are not grossly exceeded). Failure mode is most often open circuit. Whereas a Tantalum capacitor stores substantial energy within itself and when subject to a low energy transient that causes it to break down, uses its own stored energy to destroy itself. The transient can be VERY brief. Once the oxide layer is gone the cap is too. They usually produce a superb hard metallic short which makes for excellent crowbar protection. Few power supplies have enough energy to break the short of a short-circuit Tantalum. I have seen a 5v several hundred amp supply manage to do so (DEC PDP11 computer) - there was a nice Tantalum metal bead rattling around inside the case :-). Al wet electrolytics when subject to substantial overvoltage (or reverse voltage) will after a while blow out their end, eject a wad of innards and spray caustic electrolyte everywhere. But they lack the satisfying true explosion of a Tantalum and I have never seen an Al cap emit a jet of flame as Tantalums sometimes do. . Russell McMahon -- http://www.piclist.com hint: To leave the PICList KILLspampiclist-unsubscribe-requestKILLspammitvma.mit.edu

I've never had any Ta cap fail except through reverse voltage (because they were inserted the wrong way round). Then they pop and a bead of molten tantalum pokes out of a pinhole in the case. Aluminum ones take a little longer to manifest their discomfort and can sometimes still be used once they are reoriented. /\/\/\/*=Martin ----- Original Message ----- From: "Douglas Butler" <RemoveMEdbutlerTakeThisOuTIMETRIX.COM> To: <spamBeGonePICLISTspamBeGoneMITVMA.MIT.EDU> Sent: Monday, December 17, 2001 12:12 PM Subject: Re: [EE]: Aluminum caps vs. Tantalum > I remember something about tantalum caps needing to be used at a goodly > fraction of their rated voltage. I.E. on a 5V bus a 25V cap would fail > sooner than a 10V cap. > > Does anyone else remember that? > > Sherpa Doug > > > {Original Message removed}

All I'll say on the subject at this point is that tantalum caps are great when you are dealing with lots of noise in a digital circuit, when you just can't seem to decouple well enough with cheap ceramic discs. I don't care why, nor do I care how many more pennies they cost. They just make great decoupling caps. CL {Quote hidden}> > > Russell McMahon wrote: > > > > > If you want to produce a product which MIGHT smell, > shriek, explode and > > > catch fire and/or short out various parts of its circuitry at > indeterminate > > > future dates then Tantalum caps are definitely the way to > go ! Because > this > > > isn't what happens in all cases and because, if you > design well and are > very > > > careful and lucky you may not have any problems, you are > now liable to > have > > > people telling you what fine caps Tantalums are ;-) > > > > I've seen resistors smoke and explode a lot > > too, maybe we should all stop using resistors. ;o) > > I don't know what bad experiences you had with > > the tantalums but I haven't seen any trouble with > > them. But then I never use a 16v cap on a 16v > > power supply rail. > -- http://www.piclist.com hint: To leave the PICList TakeThisOuTpiclist-unsubscribe-requestEraseMEspam_OUTmitvma.mit.edu

At 01:01 PM 12/17/01 -0500, Chris Loiacono wrote: >All I'll say on the subject at this point is that tantalum caps are great >when you are dealing with lots of noise in a digital circuit, when you just >can't seem to decouple well enough with cheap ceramic discs. I don't care >why, nor do I care how many more pennies they cost. They just make great >decoupling caps. If you look at what you're trying to decouple, you can do a better job, rather than just shotgunning it. Shotgunning can also induce problems. (you have a 15 ohm resistor in series with the VCC supply, right) Generally, smaller caps for higher frequencies. It is also important that the PCB layout supports bypassing. Layouts that have the cap from VCC on one chip to GND on another chip are fatally broken. I generally use a few zero ohm resistors to steer grounds, a few 1-10 ohm resistors, a few 0.1 and/or 0.01uF caps, and that's about it. -- Dave's Engineering Page: http://www.dvanhorn.org Got a need to read Bar codes? http://www.barcodechip.com Bi-directional read of UPC-A, UPC-E, EAN-8, EAN-13, JAN, and Bookland, with two or five digit supplemental codes, in an 8 pin chip, with NO external parts. -- http://www.piclist.com hint: To leave the PICList RemoveMEpiclist-unsubscribe-requestTakeThisOuTmitvma.mit.edu

----- Original Message ----- From: "David VanHorn" <dvanhornEraseME.....CEDAR.NET> To: <EraseMEPICLISTMITVMA.MIT.EDU> Sent: Tuesday, December 18, 2001 4:58 AM Subject: Re: [EE]: Aluminum caps vs. Tantalum snip > It is also important that the PCB layout supports bypassing. > Layouts that have the cap from VCC on one chip to GND on another chip are > fatally broken. > > I generally use a few zero ohm resistors to steer grounds, a few 1-10 ohm > resistors, a few 0.1 and/or 0.01uF caps, and that's about it. > snip On this subject, why, on a four layer board (or 6 or 8 or 16), do the power planes usually go in the centre and the signal layers on the outside. Isn't it more logical to put plane on the outside to catch EMI/RFI? Why have power planes at all. If you used them as signal layers you would have almost an extra 2 layers. -- http://www.piclist.com hint: To leave the PICList RemoveMEpiclist-unsubscribe-requestEraseMEEraseMEmitvma.mit.edu

> On this subject, why, on a four layer board (or 6 or 8 or > 16), do the power > planes usually go in the centre and the signal layers on the > outside. Isn't > it more logical to put plane on the outside to catch EMI/RFI? > Planes on the outside would be better from an EMI point of view, and this is occasionally done in bleeding edge technology. But normally the planes are on inner layers so that they have fewer holes in them. If you put your ground plane on the top and you put a big BGA on your board you would have to make a big hole in your ground plane for all the signal pads. Sherpa Doug -- http://www.piclist.com hint: To leave the PICList RemoveMEpiclist-unsubscribe-requestTakeThisOuTspammitvma.mit.edu

> -----Original Message----- > From: David Lions [EraseMEyizgarnoffspamspamBeGoneOPTUSHOME.COM.AU] > Sent: Monday, December 17, 2001 5:11 PM > To: RemoveMEPICLISTKILLspamMITVMA.MIT.EDU > Subject: Re: [EE]: Aluminum caps vs. Tantalum > > > Why have power planes at all. If you used them as signal > layers you would > have almost an extra 2 layers. > stated rather simplistically; in general, the more you break up your ground plane, the less effective it becomes... -- http://www.piclist.com hint: To leave the PICList piclist-unsubscribe-requestSTOPspamspam_OUTmitvma.mit.edu

> >On this subject, why, on a four layer board (or 6 or 8 or 16), do the power >planes usually go in the centre and the signal layers on the outside. Isn't >it more logical to put plane on the outside to catch EMI/RFI? You'd think so, but it makes modifying the PCB practically impossible, and copper sheets are reflectors, not magic black holes. Proper routing means you don't radiate nearly as much to start with. I produce multi-processor designs at >10MHz with switcher power supplies, that pass part 15 with only a plastic enclosure and two layer boards. No shields, coatings etc. >Why have power planes at all. If you used them as signal layers you would >have almost an extra 2 layers. In my experience, layer three is worth about .5 layer for routing, and it goes down from there. I don't use blind or buried vias though. -- Dave's Engineering Page: http://www.dvanhorn.org Got a need to read Bar codes? http://www.barcodechip.com Bi-directional read of UPC-A, UPC-E, EAN-8, EAN-13, JAN, and Bookland, with two or five digit supplemental codes, in an 8 pin chip, with NO external parts. -- http://www.piclist.com hint: To leave the PICList spamBeGonepiclist-unsubscribe-requestSTOPspamEraseMEmitvma.mit.edu

At 05:49 PM 12/17/01 -0500, Chris Loiacono wrote: >I remember working with 'multi-wire' boards, which had layers of wiring >buried in the resin similarly. Heavier wires needed to be buried deeper. >Troubleshooting power 'layers' was a real bear, but it sure was a good thing >the signal wire layers were nearer the surface. ECN's seemed to focus on >them regularly. That's a good "whatever happened to".. Like Space Food sticks, and fizzies! I remember, the idea was that they just needed a netlist, no routing information, and it was supposed to be faster and cheaper than prototype PCBs. I suspect SMD is what happened to it. -- Dave's Engineering Page: http://www.dvanhorn.org Got a need to read Bar codes? http://www.barcodechip.com Bi-directional read of UPC-A, UPC-E, EAN-8, EAN-13, JAN, and Bookland, with two or five digit supplemental codes, in an 8 pin chip, with NO external parts. -- http://www.piclist.com hint: To leave the PICList EraseMEpiclist-unsubscribe-requestEraseMEmitvma.mit.edu

> >I remember working with 'multi-wire' boards, which had layers of > >wiring buried in the resin similarly. > That's a good "whatever happened to".. Like Space Food sticks, and > fizzies! > I remember, the idea was that they just needed a netlist, no routing > information, and it was supposed to be faster and cheaper than > prototype PCBs. They were still around a couple of years ago, being marketed towards high speed, dense backplanes. They were cost effective against the high layer count PCBs you might otherwise use. ====================================================== Steve Baldwin Electronic Product Design TLA Microsystems Ltd Microcontroller Specialists PO Box 15-680, New Lynn http://www.tla.co.nz Auckland, New Zealand ph +64 9 820-2221 email: @spam@steveb@spam@spam_OUTtla.co.nz fax +64 9 820-1929 ====================================================== -- http://www.piclist.com hint: To leave the PICList spamBeGonepiclist-unsubscribe-requestKILLspammitvma.mit.edu

> I remember something about tantalum caps needing to be used at a goodly > fraction of their rated voltage. I.E. on a 5V bus a 25V cap would fail > sooner than a 10V cap. > > Does anyone else remember that? This is generally true for wet electrolytic aluminium caps which like to be run near (but below) their rated voltage. Tantalums (see NEC pdf I mentioned) are much better run at 1/3 to 1/2 their rated voltage. Russell McMahon -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads