Safely working on older equipment?

for all subjects/topics not covered by the other forum categories
Post Reply
User avatar
SarahWalker
Posts: 1129
Joined: Fri Jan 14, 2005 3:56 pm
Contact:

Safely working on older equipment?

Post by SarahWalker » Sat Apr 21, 2018 4:16 pm

I had some time off recently, and decided to try recapping an old PC motherboard that had stopped working. Having replaced all the capacitors, I powered on the board and one of the caps quickly went bang. Quickly powered off, replaced the blown cap on the thinking that maybe it was just bad, powered back on, it goes bang again. At this point I get nervous about a 25 year old motherboard connected to a 25 year old PSU connected to the mains with components exploding, and stop working on it.

So as there are numerous people here who work on equipment older than this (like Beebs!), are there any actions you take to reduce the risk of anything going seriously wrong? I'd prefer to avoid electrocuting myself or burning the house down if at all possible!

(for the record, I'm pretty certain the exploding cap was due to inaccurate markings on the motherboard. The cap is between -12V and 0V, with the + marking on the -12V line, which I'm reasonably certain is wrong...)

User avatar
1024MAK
Posts: 8111
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Safely working on older equipment?

Post by 1024MAK » Sat Apr 21, 2018 5:58 pm

For future reference, it's wise to either mark the board with a marker pen (showing the + and - leads), and / or take photos showing the polarity of electrolytic capacitors before replacing them.

If an electrolytic capacitor is fitted with the incorrect polarity, it is very likely to go bang with the end cap exploding off. Sometimes with enough force to embed itself in a ceiling.

So never stand or lean over any electrolytic capacitors when first switching on.

The other cause of electrolytic capacitors going bang, is a severe over voltage.

If you have a multimeter, attach the negative/ black lead to a 0V / GND / ground point on the circuit board. Switch to a low resistance range, such as 200 ohms (or equivalent). Then test each connection point on each capacitor. For any capacitor on a positive supply rail (like +3.3V, +5V, +12V), the meter will show a low reading (0 ohms to 5 ohms) when you touch the probe on the connection pad for the negative lead.

For a -12V supply, if there is a RS232 serial port on the board, or the board supports ISA or PCI serial boards, the meter will show a low reading (0 ohms to 5 ohms) when you touch the probe on the connection pad for the positive solder pad. But there will only likely be one or two low value capacitors for the -12V supply.

To prove if a capacitor is on the -12V rail, test (again on a low resistance range) between the motherboard power supply connector, the pin for the -12V rail and what you think is the negative lead or pad for the capacitor. You should get a result of between 0 ohms and 5 ohms.

Mark

Coeus
Posts: 1118
Joined: Mon Jul 25, 2016 11:05 am
Contact:

Re: Safely working on older equipment?

Post by Coeus » Sun May 13, 2018 7:58 pm

My daughter just turned on an old tower PC in her room and complained that it made a weird noise and then gave out a red smoke which smelt very bad indeed. It also tripped the MCB on the ring main it was connected to, so overload rather than earth leakage. I haven't taken a look yet. Switch mode PSUs are one of the things I am a bit wary of given that they often rectify the mains and out it across one or more capacitors. I did change the caps on the BBC micro PSU but I figured in that case there being about 30 years since it was last powered on there probably wouldn't be any residual change in it.

User avatar
1024MAK
Posts: 8111
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Safely working on older equipment?

Post by 1024MAK » Sun May 13, 2018 9:27 pm

Assuming you are in the U.K., you will probably find that the mains plug fuse has blown as well. Whatever died, resulted in a low impedance short circuit between line (live) and neutral. MCBs can trip, as they can have a faster response time to such faults compared to a fuse.

There are various different switch mode power supply designs. The most common mains powered type does indeed rectify the mains voltage, which is then smoothed by one or two high voltage electrolytic capacitor(s). The total DC voltage across these will be about 339V (which is referenced to the mains).

Assuming the PSU has/had a load connected to the output, depending on what component(s) failed, there may, or may not be a high voltage charge left on the/these voltage electrolytic capacitor(s).

Any high voltage capacitors will have a voltage rating of 200V to 450V. Using a decent multimeter, you can test to see if there is a significant voltage on them. You can discharge them by using a 1k ohm resistor (rated at 1 Watt or higher) covered in insulating heatshrink sleeving, and connected to suitable wires with croc clips or test probes.

It's wise to always double check that the power supply is properly disconnected from the mains supply. And to discharge any high voltage capacitors even if you think they are not charged.

Mark

Prime
Posts: 2601
Joined: Sun May 31, 2009 11:52 pm
Contact:

Re: Safely working on older equipment?

Post by Prime » Mon May 14, 2018 4:13 am

Though if it's a standard PC replacing the PSU may be safer and easier. Chances are if it's gone bang in such a spectacular way that the fault is in the PSU as generally faults in the low voltage components, motherboard, drives etc just tend to shut the psu down without blowing mains fuses / trips.

Cheers,

Phill.

Coeus
Posts: 1118
Joined: Mon Jul 25, 2016 11:05 am
Contact:

Re: Safely working on older equipment?

Post by Coeus » Mon May 14, 2018 4:28 pm

Thanks, guys. I finally got round to taking the PSU out of the PC with the intention of checking for obvious blown caps etc. While looking round the case to find the screws to remove I noticed that someone has moved the input voltage selector from 230V to 115V! No doubt that is the cause of the failure!

Looking round there is nothing that has obviously exploded. One of the large electrolytic looks a little brown on top but that's all. It does seem there is something not quite right as measuring the resistance between the mains input pins I get 220 ohms with the selector at 115V and 450K ohms if I move it to 230V.
20180514_174533.jpg

User avatar
1024MAK
Posts: 8111
Joined: Mon Apr 18, 2011 4:46 pm
Location: Looking forward to summer in Somerset, UK...
Contact:

Re: Safely working on older equipment?

Post by 1024MAK » Mon May 14, 2018 5:45 pm

Coeus wrote:I noticed that someone has moved the input voltage selector from 230V to 115V! No doubt that is the cause of the failure!
The effect of the input voltage selector being in the wrong position, is effectively the same as feeding it 460V AC.
When the input voltage selector is in the 115V position, the mains is alternately applied to each of the two large electrolytic capacitors. These will be rated at either 200V or 250V DC. But if the supply voltage is 240V (the typical actual mains voltage in the U.K. mainland) the rectified DC voltage from the rectifier will charge these to about 339V each! These are connected in series, giving a total DC voltage of about 678V! The electronic switch transistor and control circuit are fed from this DC voltage, but are only expecting a nominal voltage of about 339V (which is what you get if the input voltage is 115/120V in this configuration, or with a 230/240V mains input with the input selector set in the 230V position).

My suggestion is to buy a new PSU...

Mark

Post Reply