I personally know of two people who had a tantalum explode in their face and almost got their eyes hit.
- VCF SoCal - February 15 - 16 2025, Hotel Fera, Aneheim CA
- VCF East - April 2025, Infoage Museum, Wall NJ
- VCF Southeast - July 2025, Atlanta, GA
- VCF West - Aug 2025, Computer History Museum, Mountain View, CA
- VCF Midwest - Sept 2025, Schaumburg, IL
-
Please review our updated Terms and Rules here
Fixing my 5160
- Thread starter Roland Huisman
- Start date
modem7
10k Member
I am going to add my view (repeat: view) on this subject:
I recognise that different people have differing opinions on the subject. In my life experience, EVERYTHING has both pros and cons. Looking at the pros and cons, I do not see either action (replace or don't replace) as 'best practice'.
My experience is like that of maxtherabbit. To quote myself from [here], "My own experience is that the failure rate is relatively high if the tantalum capacitor has been unpowered for many years (e.g. I buy an 'unknown working status' motherboard from eBay), but the failure rate is very low for a tantalum capacitor that is periodically powered (e.g. I use my IBM 51xx computers at least every few months.)" I cannot remember the last time I replaced a failed tantalum (but I seem to remember it was on a Tandon floppy drive). The failure rate that I observe just does not justify my time in replacing all tantalum capcitors. That's just me. I do know that if I was to buy a 'I found this IBM PC in the attic' machine, I expect one or more tantalums to fail, over the course of say, the first 10 hours of operation. Then stability. It doesn't mean that another tantalum won't fail down the track, but it is very rare, and I attribute that to the fact that I run my 51xx computers (now down to 7 of them) at least every few months (for a minimum of an hour).
But that's my attribution, and if it's correct, I recognise that not everyone runs their computers as often as I do.
As for exploding components, as VERAULT brought to our attention, other component types can explode as well. And is there a reason why a new tantalum won't explode? My suggestion: Wear safety glasses when your computer is open.
I have military soldering skills. On the opposite end of the spectrum, there are people here with nil or poor or extremely poor soldering skills. It scares me to think of the damage that they could do to a printed circuit board. Those people should keep their soldering to a minimum, until their skill becomes at least adequate for the task (or find someone else to do the soldering).
I recognise that different people have differing opinions on the subject. In my life experience, EVERYTHING has both pros and cons. Looking at the pros and cons, I do not see either action (replace or don't replace) as 'best practice'.
My experience is like that of maxtherabbit. To quote myself from [here], "My own experience is that the failure rate is relatively high if the tantalum capacitor has been unpowered for many years (e.g. I buy an 'unknown working status' motherboard from eBay), but the failure rate is very low for a tantalum capacitor that is periodically powered (e.g. I use my IBM 51xx computers at least every few months.)" I cannot remember the last time I replaced a failed tantalum (but I seem to remember it was on a Tandon floppy drive). The failure rate that I observe just does not justify my time in replacing all tantalum capcitors. That's just me. I do know that if I was to buy a 'I found this IBM PC in the attic' machine, I expect one or more tantalums to fail, over the course of say, the first 10 hours of operation. Then stability. It doesn't mean that another tantalum won't fail down the track, but it is very rare, and I attribute that to the fact that I run my 51xx computers (now down to 7 of them) at least every few months (for a minimum of an hour).
But that's my attribution, and if it's correct, I recognise that not everyone runs their computers as often as I do.
As for exploding components, as VERAULT brought to our attention, other component types can explode as well. And is there a reason why a new tantalum won't explode? My suggestion: Wear safety glasses when your computer is open.
I have military soldering skills. On the opposite end of the spectrum, there are people here with nil or poor or extremely poor soldering skills. It scares me to think of the damage that they could do to a printed circuit board. Those people should keep their soldering to a minimum, until their skill becomes at least adequate for the task (or find someone else to do the soldering).
VERAULT
Veteran Member
ITs true. Faulty new components can explode as well.. Absolutely.
And I get what you are saying about the people with zero soldering skills. I have been soldering since I was 8 years old. These people need practice.. I know they dont need to practice on "important" stuff. But Im not going to tell someone they cant solder something they own. ITs theirs.
And I get what you are saying about the people with zero soldering skills. I have been soldering since I was 8 years old. These people need practice.. I know they dont need to practice on "important" stuff. But Im not going to tell someone they cant solder something they own. ITs theirs.
channelmaniac
Veteran Member
I'm on that list as well... It was off to the side and hit me 1/4" to the side of my right eye.
Had to burn those underwear. Now I just put a bit of pink bubble or foam sheet over the PCB when first powering it up to catch the shrapnel.
Good idea.
I have only had 2 boards with prefailed tants, and I replaced those and both were on your fail list. I haven't replaced all of them bit but understand the caution and won't fault anyone who wants to replace them all. I've been soldering since I was a kid so over 30 years and am one of those oddball that does smd/smt work by hands (no hot air).. you haven't lived till you solder a 100pin Intel ethernet chip with a non temp controlled RadioShack iron with no microscope (just your eyes). it worked first try.
channelmaniac
Veteran Member
I do hot air removal and hot iron placement. Pace MBT250. It's fun flowing 144 pin QFPs that are barely over 1" square.
The only tantrum I had explode was on a 486 motherboard, and it was quite violent. Luckily I wasn't in its path. Had another on my 5160 that when turned on just lit a flame on the top of it. This is in some ways more worrying as a loud bang alerts you to a failure. Had the case been on, I very well may have not noticed and have a fire on my hands. In both cases the systems had not been powered on for over a year.
Roland Huisman
Veteran Member
Yeah, I'm fixing and working with electronics for about 35 years on a daily basis. And still after all these years electronics can surprise me at any time... Especially with unknown state electronics you have to be a bit careful. Once it runs again the surprises will be less big... But not using a machine for 10 years might be a bit tricky... Just like this 5160... Repairing stuff now will not mean that it still works within 10 years. And replacing good old capacitors with modern Chinese crap will also mean you can expect problems within a few years... I've seen modern ATX power supplies turning into a smoke bomb within seconds... And I've seen 50 year old power supplies still going strong without any modifications...
We all have our experiences and our ways to fix stuff... To me nothing is right or wrong, as long as you are comfortable with it and taking care of these old machines... It is just to have a bit of fun with it...
Regards, Roland
We all have our experiences and our ways to fix stuff... To me nothing is right or wrong, as long as you are comfortable with it and taking care of these old machines... It is just to have a bit of fun with it...
Regards, Roland