[lug] Aging Surge Protectors

[Nate Duehr]

Few days behind on this one, but yeah... the comments are all generally 
correct regarding surge supression.

I'll share some thoughts...

The two most common ways to protect against surges on AC lines are 
MOV's and gas-discharge tubes.

MOV's react sooner but can't withstand high current and will be 
destroyed or damaged by high surges of current.  Gas-discharge tubes 
are slow to fire but can then pass (relatively) large amounts of 
current.

Some very good designs (rare) use both... MOV's fire quickly at the 
beginning of the surge event and start shunting unwanted voltage to 
ground, then the gas-discharge tubes fire and soak up the rest of the 
current.  These designs last through just about anything other than 
direct or indirect lightning strikes if the right values of the 
components are used.

Many of the radio site systems I've worked with people on over the last 
10 years or so have had their power supplies internally modified to 
have both MOV's and gas-discharge tubes.  In those last ten years, I've 
seen about four complete MOV failures due to lightning either close-by 
or a direct hit to the radio tower itself, even after proper lightning 
protection (single point ground, halo systems, and Polyphaser panels at 
all entrances to the building) has been followed.  The gas-discharge 
tubes rarely fail, but some have had signs of arcing over the tube and 
into the power supply.

Since the one club's technical lead started adding these modifications 
to the power supplies, no power supplies have been completely lost due 
to lightning, however if the MOV's are completely destroyed by a hit,  
the AC feed to the supply is broken and the radio system goes dead 
until someone can arrive at the remote site and repair the MOV's and 
check the gas-discharge tubes for damage.

Another effect we have here in Colorado is that electrolytic filter 
capacitors in power supplies are prone to drying out -- and the caps 
are sometimes in the front sections of the power supply as either 
rectifiers or filters to help the supply provide "clean" power to the 
system.  Inside electrolytic capacitors is a liquid/jelly substance 
that is the electrolyte the makes them work -- at high altitude in dry 
climates like ours, 10 years of sitting in a non-environmentally 
controlled building at 10,000-13,000' MSL will cause the water content 
of the caps to become lower than their design tolerances allow and they 
then either degrade in performance or in the worst case, arc over 
internally and "pop", usually leaving a "strange" confetti-like 
substance inside the case of the power supply and two attachment leads 
soldered into the board with nothing attached to them anymore, or just 
small pieces left-over.

Anywhoo... that's just some of the fun things I've seen working on 
radios that live in harsh environments around here.

Cheap "power strip" surge protectors USUALLY only have MOV's in them... 
thus, if you think your store-bought surge protectors have survived a 
big surge... it's probably time to replace them.  As Dan mentions, many 
designs do lay in an LED across the MOV to show that it's still capable 
of passing current, but it does not prove that the MOV is still acting 
normally or "firing" at the same place in the curve after a big surge.  
Best to just chuck out the $10 surge suppressor and put in a new one at 
that price... or at least find a suitable replacement MOV and change 
them out if you suspect yours have been put through a large surge.

The number one rule in lightning "protection" that I've read and can 
back up with experience of having seen the damage after a number of 
direct strike incidents, is that you want your ground system to rise 
and fall for every device in the system at the same potential.  Older 
houses with multiple grounds (water pipe ground to a sub-panel in the 
basement, outside ground rod, telephone ground rod, ground rod at a 
detached garage - i.e. My 1960's house... oh no!) that are not directly 
bonded together suffer the worst from external surges.

Lightning strikes nearby and instead of the ground system becoming 
instantaneously charged to a specific potential and then tapering off 
as the system discharges through the single-point ground, the system 
"sees" each ground as a better or worse place to discharge -- and then 
high current flows between the grounding points... thus causing major 
damage in the house wiring (or other conductive paths like appliances 
and other household electronics) between the grounds.

Hope that's helpful info.  Boiled down it is:
- MOV's generally fail after surges - replace surge suppressors that 
have been "hit".
- Designs with gas-discharge tubes can offer greater protection but 
slower response.
- Think about your grounding system.  (Example: Putting a metallic 
power strip with a surge suppressor inside of it into your 7' data 
center rack and then expecting it to do any good is foolish... the 
MOV's inside drop the excess voltage and current to the CASE GROUND of 
the power strip... which is touching the metal rails of your cabinet... 
thus, energizing the cases of all the machines in the cabinet.  Oops.)

Oh... almost forgot...

Inexpensive consumer-grade UPS's are typically "off-line" type designs. 
  This means that they sense a loss of power and THEN kick over to the 
battery, faster than the downstream power supplies will notice.  This 
also means that in many cases these cheaper UPS's do little or no 
actual surge suppression, depending on their design.  Many only have 
MOV's similar to a "regular" surge suppressor.  Some people think that 
having a UPS is a good way to protect equipment because the batteries 
themselves act as slow capacitors and can help absorb transients -- but 
that's only true in the case of so-called "on-line" UPS's.

Okay... 'nuff said... probably too much.  ;-)  Heh.  Sorry.

--
Nate Duehr, nate at natetech.com