Surge Protection

A lot of people still do it though.......................................................

It was a big time trick of the service trade to bridge a blown fuse with a

60W bulb and "see what smoked".

In reality; the PTC bulb surge current can damage a few semiconductors downstream.

Not that long ago; my monitor produced a nasty niff instead of anything on the screen when I plugged it in.

More or less resigned to scrapping it, I opened it up to see if there was any worthwhile salvage. The fuse was blown, so I fired up the iron and did a few checks - all the usual suspects were innocent, so I resorted to the bulb trick. A whisp of smoke started rising from the mains in common mode choke.

I'm sure you do. But Phil doesn't take criticism *at all* well (even if it's sprinkled with fairy dust and castor sugar;-)).

We had one product that had a giant (850 watt) 50/60 Hz power transformer. The combination of capacitor charging and magnetic saturation from random turnoff magnetization tended to weld the power switch. We put a biggish disk NTC in each of the transformer (120/240) primaries, and that fixed it. I tried teasing the switch on/off with various timing, and the NTCs still helped. The switch failures sure stopped.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

When my welding set welded its on/off switch, I couldn't get hold of a big enough NTC.

A cooker oven switch did the trick, but I had to remember to only turn the knob far enough to click - the rest of rotation is the simmerstat bit.

The place I worked at had almost identical stackable switch units to the original equipment - but they kept them too well guarded.

** You ought to patent an idea like that.

** Fraid I am in the electronics business, while audio is something you are pig ignorant of. ** No fooling? Who would have thunk.

Your lack of experience with equipment you did not design is showing - same problem for most designers.

And your asinine attitude.

.... Phil

** Any reactance is negligible at the frequencies involved, so the resistan ce alone is doing the job.

You say you have seen "surge damage" to SMPSs in scopes etc - but is that l inked to *inrush surges* or something else ?

The only components likely to suffer from inrush surges are the rectifiers acting on the incoming AC supply - something designers have know about sinc e Noah was a boy. Very likely, the SMPSs you are dealing with already have inrush surge suppression in the form of WW resistors, NTC thermistors or si mply the resistance of common mode chokes used in line filters.

It is only necessary to limit the worst case surges to about 30amps peak to protect the sort of diodes and bridges that might be involved and this mea ns the total resistance in series with the AC supply needs to be 10ohms.

With no resistance included at all, peak surge currents would be well over

100amps and destroy diodes immediately and/or trip supply circuit breakers at switch on.

.... Phil

Metal oxide varistors(MOV) are very good surge suppressors - we use it for power line surge suppressors, satellite TV surge suppressors, wall plug mount LED lamps etc., Very easy to incorporate in any circuit.

.

ance alone is doing the job.

linked to *inrush surges* or something else ?

s acting on the incoming AC supply - something designers have know about si nce Noah was a boy. Very likely, the SMPSs you are dealing with already hav e inrush surge suppression in the form of WW resistors, NTC thermistors or simply the resistance of common mode chokes used in line filters.

to protect the sort of diodes and bridges that might be involved and this m eans the total resistance in series with the AC supply needs to be 10ohms.

r 100amps and destroy diodes immediately and/or trip supply circuit breaker s at switch on.

The problem is gradual degradation of the NTC Rs. I suppose one could alway s replace one with 4.

NT

MOVs are subject to cumulative damage, though. They should really be replaced periodically. I would say NTC thermistors, suitably rated, would be a better bet. Or Phil's suggestion of a current limiting resistor.

** MOVs and NTCs do different jobs, one is not a better bet than the other.

** A resistor does not change its value, so will guarantee a known and constant surge current limit every time the unit is switched on.

An NTC is a poor solution, as it only works as intended the first time unless given many minutes to cool down OR works in conjunction with a relay bypass system that must be carefully designed to drop out soon as the AC supply is removed.

.... Phil

tastes great vs less filling

I would put in enough fixed resistance so the peak current is within spec, then also add an NTC to further reduce the peak during a normal start.

This way, 99% of the time when there is a cold start, the surge will be very low...

and those 1% re-rush corner cases, the fixed resistance will keep the peak safe.

Many times, the hot resistance of the NTC is sufficient and a separate fixed resistor is not needed.

mark

Using both is pointless - the downside of a fixed resistor is increased volt drop if current increases for any reason.

The NTC is better, but as Phil says; an outage re applies current with a hot NTC, so no inrush protection for that event.

I disagree that timeout relay design is difficult - a relay means most of the circuitry is on the secondary side. A basic monostable will do the job, you just use a steering diode to clamp the timing capacitor to a secondary side Vcc rail - when that goes down; it resets the timing circuit.

e

s

ec,

olt

hot

b,

y ** There are too many different scenarios for one scheme to be the best so lution. Some PSUs must start under full load while others can have that loa d delayed until things settle. Audio amplifiers are an example of the latte r and it is normal to mute the inputs until a few seconds after switch on.

If the inrush surge is so large it must never happen, as is the case with l arge toroidal transformers, then it is worth using a separate transformer j ust for the relay. It is then easy to have the relay de-energised immediate ly the AC supply is disconnected.

... Phil

The OP wants it for the SMPSU section of things like oscilloscopes - the scheme I suggested would fit that bill.

** But is more complicated to design and install than needed.

The OP to wants to upgrade existing designs that work well enough - cos he is paranoid about imagined large inrush surges.

If he merely replaced the old supply side PSU diodes or bridges with larger ones and maybe the on-off switches too, the upgrade is done.

..... Phil

Probably nearly half of commercial electronics come out the factory with just the basic NTC, probably more in the case of professional gear.

Usually delay relays turned up on CRT VGA monitors for the degauss posistor. Controlled by the front panel micro - sometimes also a manual degauss button.

You only need a similar delay for the inrush NTC if petty little weenies spend all their time picking fault with every suggestion put forward.