|
Vintage Computers Any vintage computer systems, calculators, video games etc., but with an emphasis on 1980s and earlier equipment. |
|
Thread Tools |
2nd Mar 2021, 4:19 pm | #1 |
Diode
Join Date: Mar 2021
Location: Bangor, Gwynedd, Wales, UK.
Posts: 6
|
Tantalum identification..
This is my first post here so hello all! I spend my days (and most of my nights) repairing modern computers (most of which is basically like Lego these days) but I am newb to the older PCs that I now find myself fascinated by! Anyway, I am trying to bring life back into a Compaq Portable 386 and have found a couple of tantalums on the main board that have obviously gone pop. They appear to say '10-15' and below that '+M (2)' (with the 2 in a circle). Is there anyone here who would mind explaining how to work out what these markings mean and what modern component I can replace them with? Here's a piccy!
|
2nd Mar 2021, 6:52 pm | #2 |
Octode
Join Date: Jan 2014
Location: Oban, Scotland, UK.
Posts: 1,129
|
Re: Tantalum identification..
10uF 15V
Such tantalums are still available for a like-for-like replacement. |
2nd Mar 2021, 7:02 pm | #3 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Tantalum identification..
They are, but they are considered to be quite a nuisance for exactly the reason that you have found. Unless you want to keep it looking original (which would be a perfectly reasonable thing to want to do) it might be better to fit high quality conventional electrolytic capacitors in place of those old electronic firecrackers.
They often do 'just fail', but Tantalums really, really don't like being overvoltaged and that is one possible reason for them to fail in such spectacular fashion. Snip them out and measure the voltage across the pads to see what it is - if it is higher than it should be, find the reason for that first. |
2nd Mar 2021, 7:05 pm | #4 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,363
|
Re: Tantalum identification..
I have been informed that a common reason they fail is that people use them with too low a voltage value in terms of headroom - forgetting about peak values that occur during certain events - that is something that definitely takes them out.
|
2nd Mar 2021, 7:12 pm | #5 |
Nonode
Join Date: Jan 2009
Location: Bristol, UK.
Posts: 2,061
|
Re: Tantalum identification..
Judging by the burnt remnants, it is in a lowish source impedance circuit, like across the rails of a power supply! At the very least, replace with a higher voltage tant (25V) or you can, usually, switch to a low ESR electrolytic, again with a somewhat higher voltage.
Ideally, a tant bead capacitor will last almost FOREVER in a properly designed/specified circuit. The fact it didn't suggests its running too close to its operational limit, hence my suggestion to raise its voltage rating. I've seen quite a few do a passable impersonation of Mt Vesuvius but they have all been down to penny pinching on the voltage rating! dc |
2nd Mar 2021, 7:42 pm | #6 |
Pentode
Join Date: Jun 2012
Location: Bristol, UK.
Posts: 115
|
Re: Tantalum identification..
Les Lawry-Johns often said in his regular articles in 'Television' magazine: "Beware the blue tants!" There seems to be a trade-off in these capacitors. The higher the capacitance, the lower the working voltage.
|
2nd Mar 2021, 8:52 pm | #7 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,864
|
Re: Tantalum identification..
Any colour tant will get you!
Around 1990 there were recommendations put round that tants should not be used in circuits where there could be large inrush/outrush currents, and that they ought to be used with a small series resistor added to them. This completely wipes out their main advantages: Low ESR and low ESL. Rendering them most unsuitable for decoupling. Modern low ESL aluminium is best if it works. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
2nd Mar 2021, 9:33 pm | #8 |
Diode
Join Date: Mar 2021
Location: Bangor, Gwynedd, Wales, UK.
Posts: 6
|
Re: Tantalum identification..
Thank you all so much for your help, I think if it won't do any harm to use the same style of component then I should do that but I will go for the 10uF 25V version just to make sure. I have loads of other things to fix so I am sure I will be back and even though my knowledge is limited at the moment I will maybe even find someone here I can help! I just replaced the RIFA caps on an Osborne 1 and I have a PET (as mentioned in the Wanted section by Welsh Anorak who is a chum of mine and who led me here) with screen garbage, a Compaq Portable 1 which seems dead and two more PET's (or CPMs) on the way.
|
3rd Mar 2021, 7:33 am | #9 |
Pentode
Join Date: Jun 2012
Location: Bristol, UK.
Posts: 115
|
Re: Tantalum identification..
Nobody mentioned (especially me!) that the M indicates ±20% tolerance.
K is ±10. And the bar indicates the anode. If you have a real antique you will see 2 colour body with a spot, like a 1930's resistor. |
3rd Mar 2021, 9:01 am | #10 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,864
|
Re: Tantalum identification..
Yes, watch out for that polarity mark. For some screwy reason, the convention is to mark the positive leg of a tantalum electrolytic capacitor, while on aluminium electrolytic ones, it is normally the negative leg which gets marked with a bar.
From the late eighties onwards, the market price of tantalum dust rose exponentially. Capacitor makers minimised the amount per component and the capacitors not only fell to the lower end of their capacitance tolerance, the working voltage also fell. Industrial users noticed a fall in reliability, and designers were advised to de-rate the working voltage of all tantalum capacitors they used, and to add series resistance. Some capacitor manufacturers noticed this and down went the voltage that the capacitors would really stand. They still had the same voltage ratings on their data sheets and markings, but the mean time to failure kept falling, and the de-rating got more, and so on. It was a race to the bottom. As a result of all the ensuing failures in equipment, tantalum capacitors got a stinking reputation. The bead ones were worse than their metal-canned 'wet-slug' variants. So there you have it... avoid the things like the plague. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
4th Mar 2021, 12:11 am | #11 |
Octode
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,437
|
Re: Tantalum identification..
Yes, with Surface-Mount ones I've found quite a few Engineers hadn't realised that the marked end was positive.
And equipment manufacturers, including Sony have been known to fit the odd polarised capacitor the wrong way round (sometimes with the PCB incorrectly marked - maybe due to a schematic error), so leading to early failures / corrections required. I recall Watford Electronics making memory upgrades for the Acorn Archimedes computers, with all the Tantalum capacitors the wrong way round! So it was surprisingly they didn't go bang immediately, as tantalums tend to conduct as soon as you reverse-bias them (and it's been known fort people to put 2 back-to-back in series, to make an un-polarised capacitor). But I also came across someone's RS232 DC-DC converter that had some backwards, and had worked for a few months before slowly dying, so we had to get them all back and fix these properly. Tantalum capacitors are also not usually rated for much ripple / surge current - probably due to quite-high ESR on lower values (as standard Aluminium Electrolytics can be). So this may also be a factor in their early demise. I have also read that there was a Post Office guidelines to derate voltage to 50% and have 10R resistance in series with Tantalums, to limit surge currents (which as stated, does defeat their effectiveness somewhat). You could also get 'mil spec.' high reliability ones from Sprague, with built-in fuses! Over the years I've found many tantalums going short - even catching fire right in front of me, on a powered circuit, as it seems that one was starting to. But have also had them fail in a watch-batteries powered microphone. When you read how Tantalum beads are actually made, with so little effective insulation thickness to get the high capacitance, it's a wonder any last that long at all! Racal HF receivers also have the dreaded Thompson make (Blue?) ones, that so many that fail in these (I think upto 4 went in mine) with age, that you are best to just replace the whole lot before more go. The Yellow AVX etc. ones do seem rather better long-term, if not subjected beyond ratings. Quite a few years ago, there was a bit of a World-wide Tantalum 'shortage' / cost increase, that lead to superior Niobium and Organic-Semiconductor electrolytics plus high-density ceramic being used. Although not that common, you can now get small-size 10uF ceramic leaded capacitors (basically packaged versions of the surface mount ones) - but voltage rating may not be that high - that should be rather better reliability (although Tantalums don't suffer from DC-bias capacitance variation issue) For 0.1uF Tantalums etc., I'd always just put a standard ceramic in place, which are also cheaper. |
4th Mar 2021, 7:25 am | #12 |
Hexode
Join Date: Apr 2017
Location: Buderim, Queensland, Australia.
Posts: 428
|
Re: Tantalum identification..
Like others have said, I suggest -
Replace with 10uF 25V bead tantalums for low ESR, safer higher voltage, and replacement ease on high density PCBs. Replace with tantalums particularly in pulse circuit bypass situations, instead of ordinary electrolytics, which can be used as satisfactory replacements in less demanding situations. Blue 16V tantalum caps sometimes failed because they were used in 12 Volt PS rail situations. The industry only realised late (about 1980?) that this type of component should be used at 50% or less of rated voltage. Whenever I see a blue cap in a failed unit with 12V PS rail, it gets replaced. They can even explode, fail OC gracefully, and the PCB functions quite OK due to other nearby bypassing. . |
4th Mar 2021, 10:17 am | #13 | |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,864
|
Re: Tantalum identification..
Quote:
This can lead to trouble in decouplers near regulators, because the designers of the regulator chips often relied on the ESR to keep the things stable! Check datasheets to look for anything about capacitor types. Often it's not mentioned, so take a shufti at the applications note circuits and see what they used themselves. Sometimes you have to pay a lot of attention to what people don't say... David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
|
4th Mar 2021, 11:23 am | #14 | ||
Octode
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,437
|
Re: Tantalum identification..
Quote:
Yes, the much-lower ESR is usually an advantage as stops self-heating, and increases ripple current handling. And I recall reading a few decades ago in Television magazine, that some Service Engineers would often add low-ESR capacitors in parallel with existing electrolytics to improve HF ripple current handling and improve reliability. Fortunately, most of the older (Non Low Dropout) regulator IC's, used in more vintage equipment were generally unconditionally stable, so low-ESR capacitors were not an issue and also didn't require a minimum bulk output capacitance for stability. But many early Low Dropout Regulators regulators had a much narrower stability region, and might need to insert a small resistor in series with low-ESR capacitors to ensure stability. But could sometimes use a much larger capacitor, with a bit of ESR, to damp down potential oscillation. The latest regulator IC's tend to be designed to be fairly non-critical on output capacitor ESR, as the trend has been to use ceramic ones as much as possible and avoid Tantalums due to expense / reliability. I also like to avoid using Electrolytics wherever possible / only use the highest temperature rated ones and also much longer >> usual 1000h at max rated temperature ones, as have seen too many fail after a short time in consumer equipment |
||