LongRunner's Mini-review Collection DISCUSSION

[Behemot]

Did I miss what is it for?

[LongRunner]

Did I miss what is it for?

To check if plastic parts of (mainly) electrical items contain enough fire retardants to limit flame spread and self-extinguish properly, of course.
(I already gave it a go on a broken piece from the FH-07A.)

Realistically, I can't afford to destroy one of every item I own (and I imagine few other people could either); so I end up limiting it to:

• Items too far gone (whether decrepit, abused, or just badly made), to sensibly re-use or repair
• Occasionally, if there's a sizeable extraneous bit of plastic (most often as molding flash), that I can remove without breaking the product.

Otherwise, I'm limited to tests which are non-destructive (or at least should be, in a competent product).

[Behemot]

Oh yeah, that one

[Behemot]

• If the neutral conductor breaks (or a plug just creeps loose) upstream of the power strip, then the load current will take a detour through the MOV to earth; if the circuit is on an RCD/GFCI that will trip immediately
• If you plug one into an outlet without an intact earth connection, whilst touching the casing of a Class I appliance (this definitely includes your desktop PC!) plugged into the computer murderer – and the active/hot pin makes contact before the neutral, and any appreciable load is switched on (one incandescent lamp or electric blanket is ample, never mind a space heater) – then you'll get electrocuted (or shocked at the very least).

MOV being a voltage-dependent semiconductor only opens after reaching the treshold for it to become conductive. When the neutral fails, nothing will happen as the MOV will still have resistance somewhere in the vicinity of megaohms. Some tiny currents may flow through it (as they always do), as well as through leaking Y capacitors. If that sums to high enough current, RCD will kick in and shut the circuit down.

You get electrocuted because of leaking Y capacitors, induction (SMPS does induce some currents in the metal casing/shielding) and bad transformer insulation, not because of MOVs. Those usually have much lower leakages than Y capacitors, as known from class II/defect class I bricks powering laptops and similar equipment which sometimes give some ppl shocks (especially the metal ones). The voltages usually reach about 50-100 V so you can feel it but the currents are very low and (usually) harmless.

In general, I do not really see any problem with MOVs on ground, good surge protectors and UPS usually still have those, just the cheapos don't (to save costs). While line-neutral MOV is usually enough, not always, especially with distrasous surges when the normal-mode MOV immediatelly vaporizes it could help to conduct another small part of the impuls to the ground.

Protection against disasters like lightning is a science of it's own and every protection mechanism you bring in could actually harm you as it brings other problems. You would think that commong quality grounding even between buildings makes sense, right? Well the problem is when you have that and lightning of 1 MV strikes one building: even if you have only 1 ohm resistance on the ground, that makes 100 000 volts induced potential which affect the nearby buildings too. You don't want to touch anything earth grounded when that happens, lol. But equipment-wise, that makes the potential to be 100 kV - 230 V (for us)=pretty much still 100 kV When that happens, no ordinary component will survive such potential difference. But when you have common mode MOVs in there, they will open, allow the current to flow FROM ground to both line/neutral and lower that potential difference. Off course they will blow in an instance, but so will the lightning energy pulse get conducted to the ground too. This is very complicated problem of many transient phenomenons but in the end the MOVs could give some equipment just the few extra microseconds it needs to not break insulation of its components (especially semiconductors).

[LongRunner]

Clarifying it piece-by-piece:

• If the neutral conductor breaks (or a plug just creeps loose) upstream of the power strip, then the load current will take a detour through the MOV to earth; if the circuit is on an RCD/GFCI that will trip immediately
• If you plug one into an outlet without an intact earth connection, whilst touching the casing of a Class I appliance (this definitely includes your desktop PC!) plugged into the computer murderer – and the active/hot pin makes contact before the neutral, and any appreciable load is switched on (one incandescent lamp or electric blanket is ample, never mind a space heater) – then you'll get electrocuted (or shocked at the very least).

MOV being a voltage-dependent semiconductor only opens after reaching the treshold for it to become conductive. When the neutral fails, nothing will happen as the MOV will still have resistance somewhere in the vicinity of megaohms. Some tiny currents may flow through it (as they always do), as well as through leaking Y capacitors. It that sums to high enough current, RCD will kick in and shut the circuit down.

You get electrocuted because of leaking Y capacitors, induction (SMPS does induce some currents in the metal casing/shielding) and bad transformer insulation, not because of MOVs. Those usually have much lower leakages than Y capacitors, as known from class II/defect class I bricks powering laptops and similar equipment which sometimes give some ppl shocks (especially the metal ones). The voltages usually reach about 50-100 V so you can feel it but the currents are very low and (usually) harmless.

That section is specifically about the frightening U.S. versions, as I wrote immediately above:

At least in these Australian versions, the neutral—earth MOV is rated to withstand the full mains voltage. In the USA (which probably has, in general, the least safe electrical system of any "first world" country), even that doesn't necessarily apply. This introduces two major further dangers:

I was not claiming that particular danger applied to the Australian versions (as shown).
(Indeed, the use of 680V MOVs in the L—E and N—E positions is specifically to pass a 500VDC insulation test.)

In general, I do not really see any problem with MOVs on ground, good surge protectors and UPS usually still have those, just the cheapos don't (to save costs). While line-neutral MOV is usually enough, not always, especially with distrasous surges when the normal-mode MOV immediatelly vaporizes it could help to conduct another small part of the impuls to the ground.

Zero Surge explains the problem very clearly indeed. In a nutshell:

By sending surges to the ground line, the voltage rise on the ground can disrupt audio, video, data and communications signals and also damage interconnected equipment.

And in more detail:

Any surge suppressor which diverts surges to the ground wire is a Mode 2 suppressor.
<snip>
Unfortunately, this “three mode protection” process diverts high energy powerline surges directly into delicate low voltage audio, video and computer datalines, because these lines use the powerline ground wire circuit for their reference voltage.

Computers with modems or datalines to other equipment, such as LANs and shared printers, should never use Mode 2 surge protectors which divert surges to the powerline ground, because this will increase the likelihood of damage.

A surge which is not diverted by a surge “protector” will hit the computer’s power supply, which is considerably more surge tolerant than the delicate dataline circuitry that Mode 2 suppressors endanger.

It's not just a case of “more is better”, either. (And supposing that you need stronger protection, is it really that hard to just use a bigger MOV?)

In the units I smashed apart, the L—N MOVs showed no visible damage and their thermal fuses remained intact; so clearly they weren't pushed too hard.
(This is only a small office, occupied by just my Mum + her business partner + one receptionist + sometimes my sibling; the only appliance in there that could conceivably generate any meaningful surge is a fridge, but even that's just a small “bar” unit. The air conditioners are all inverter types.)

And even if they did no harm – what could you possibly gain, given that:

Modern equipment is inherently immune to Common Mode surges.

(Within the dielectric strength of the insulation, of course.) Granted, non-safety-rated 1kV or 2kV ceramic caps (in non-compliant SMPS) could break down; but I'd never pass a unit containing those, and I don't suppose you would either.

Protection against disasters like lightning is a science of it's own and every protection mechanism you bring in could actually harm you as it brings other problems. You would think that commong quality grounding even between buildings makes sense, right? Well the problem is when you have that and lightning of 1 MV strikes one building: even if you have only 1 ohm resistance on the ground, that makes 100 000 volts induced potential which affect the nearby buildings too. You don't want to touch anything earth grounded when that happens, lol. But equipment-wise, that makes the potential to be 100 kV - 230 V (for us)=pretty much still 100 kV When that happens, no ordinary component will survive such potential difference. But when you have common mode MOVs in there, they will open, allow the current to flow FROM ground to both line/neutral and lower that potential difference. Off course they will blow in an instance, but so will the lightning energy pulse get conducted to the ground too. This is very complicated problem of many transient phenomenons but in the end the MOVs could give some equipment just the few extra microseconds it needs to not break insulation of its components (especially semiconductors).

Well, we have yet to deal too closely with lightning. (No PSUs have yet catastrophically failed here; the only dying supplies have been from bad capacitors.)
And I severely doubt that the mere 14mm L—E and N—E MOVs in the unit shown could possibly save anything whatsoever, if we were directly struck.

As strange as the examined unit's design may seem, it actually makes a surprising amount of sense when viewed in the correct context:
The L—N MOV is 25mm diameter because it's the only one that does any legitimate ‘protection’; and because it's the only one that does real work (and the other two are set for a higher voltage), it's also the only one they bother putting a thermal fuse on. (Not that it's especially safe, to not have fusing on the L—E and N—E MOVs.) The ‘token’ X and Y capacitors are just so the manufacturer can factually claim to provide filtering, without spending much actual money.

I can be confident enough that whatever price the units sold for back in 2002, it was horribly excessive for the little stuff in them.
(Perhaps they used the Tamperproof Torx Plus screws, to hide that reality from the casual observer .)
I mean, if you're going to charge a lot of money for a power “filter”, the least you could do to justify it would be to add some beefy inductors.

I'll bet that had we known better at the time, we could have instead put the money towards:

• 2 × ST340016A (rather than the unlovely ST340810As), for the main PCs
  (Even the cute little ST320011A was surely a much nicer drive than any capacity of U6. I do have one built 2001-11-15, and it still sounds OK.)
• 2 × ST360021A or even ST380021A (rather than the Maxtor 6L060J3s, whose ball bearings eventually seized), for the office server
• Better PSUs, perhaps? (Just hopefully not the Fuhjyyu-ridden CWT-built Antecs )

Incidentally, the fried HDD (a Samsung HD082GJ) did have TVS diodes on both the +12V and +5V inputs (examining the other, surviving unit). Presumably, the data lines were where the surge got in (TVS diodes provide reasonably robust protection for the power lines, but the data lines are the most vulnerable).
Its replacement was actually the ST380817AS (with an Agere MCU and Texas Instruments SH6950 motor driver; I'm not sure if their weaker STMicroelectronics counterparts would have survived) mentioned in my HDD mini-review thread (these older Barracudas are so awesome ).

The WD800JD-00LSA0s have no TVS (of any description), their only discrete diodes being 3 × STPS2L30A (2A 30V Schottky, for the DC—DC converters); yet they still served reliably in my previous home PC (I was never daft enough to use those surge-to-ground devices at home; it's just whatever MOVs are built into the PSUs, and that's about it). I did add one to our previous reception PC (originally built in 2008, my WD800JD-00LSA0s dated from March 2006), which later died.
(If it's indeed the WD800JD-00LSA0 that stopped working, then I'll bet on it having been electrically murdered, rather than failing of its own accord. R.I.P. )

Anyway, well-engineered electronics are much less vulnerable under normal circumstances, than the marketers of those “protectors” want us to believe…
(Basically, they're marketed not unlike the endless health frauds – indeed, these abominations are essentially quackery for electronics. )

[LongRunner]

For the convenience of readers, I've added a list of the reviewed items in the initial post (including brief overviews of what they're like).

[Behemot]

The adapters are not as bad as the cheapest chinese ones and it's good that they have normal screws, when the contacts loose tension, you can disassemble it for repair

I do not really get the need for fuse on cable in AU and around though. Don't you have circuit breakers? And any decent appliance has fuse or breaker built in as well.

[LongRunner]

Only the adapter from US/UK plugs to an Australian outlet has Phillips screws, the others have tri-wing.

BS 1363 was designed for use on ring circuits (originally protected by a 30A rewireable fuse, nowadays a 32A MCB), so BS 1363 plugs need a fuse to protect the flex under short-circuit conditions (the same goes for adapters that plug into a BS 1363 outlet, and accept unfused plug types). (If ring circuits were really a good idea is disputed, but that doesn't change the requirements of BS 1363.)

Other countries use radial outlet circuits (usually protected at no more than 20A), which is why they can get away with using modest cords without plug fuses.
(And no, Australian plugs aren't fused.)

[Behemot]

I see. I don't really see any benefit of that these days when the actual wiring work is several times more expensive than the cable used, and thinner cable is about the only benefit of this.

For once britons could take their heads from their bottoms and use something normal. Just look at how bulky their sockets are because of this.

[LongRunner]

Added a shorter post on a few internal differences (one somewhat unfortunate) between the KFH600 and KFH660.