New sewing machine, dangerous fault!

[TriodeTiger]

So, I got a new Sears Kenmore 385, naybe $100-150ish range for my 21st (random, aren't I?) to do some random clothing repairs and play around with, and I jammed it with a crazy monogram I was playing with.

I pop the clutch while it is on so it has light, but doesn't let me hit the pedal and do something horrid to myself, and I fish the thread carefully out (of nowhere special) with a metal rod. Guess what? Speakers crackle.

Two things I quickly think of after hastly turning it off - it was a single point I touched, and, it's a polarised plug. I must've provided a path through to earth to upset the speakers. Scary.

I pull out my little multimeter and find this - on ALL metal surfaces:



(56.6VAC, if that is unclear)

At 4am I was fancying ideas of murder attempts, or conspiracies against hobbyists, but I decided to play it safe before opening it beyond my ability to reassemble it and post on some repair forums. I thought soon after "Ah-huh! It might be some sort of motor starting capacitor" - as I've heard someone on here experiene a shock from them on his father's washing machine that had a (I assume) broken ground lead. A moderator with 14k+ posts said it probably wasn't, and that I should return it. Eek.

Made in Thailand it seems, although I guess not China, I will certainly test any mains run equipment I get next time more thoroughly ... I treated this thing like a child in the short time I had it.

Tiger.

[jh15]

Just for grins put a 1500 ohm resistor across the meter input, preferably also with a .15 microfarad in parallel with the resistor.


You might be reading electrostatically coupled voltage.

Before getting my Sencore powrrite with integral leakage tester, I used that method described in most consumer electronics service manuals.


Tell us other measurements such as each pin of plug to all exposed metal parts of the machine.

I would then report it to the Consumer products safety commision or equivalent in your country.

[TriodeTiger]

Okay: A spur of tests.

(just plain probing)
while on, plates to neutral: 55 VAC, plates to hot: 10.4VAC
while off, plates to neutral: 0 VAC, plates to hot: 65 VAC

Uhh, resistor between live and plate scares me, so I'll just do to neutral:
while on, plate to neutral with 2.2k (nearest I had) in parallel: 0.1V (it seems to drop it to nothing.)

I stuck a resistor carefully in neutral beside the probe, and ever so carefully touched it against the plates for my tests...

What is it coupled to or from? I might be able to make sense of it soon ... could the transient clicks on my (old, possibly capacitor-leaking speakers) be just the charge building up numerous times and discharging? I wonder if a low-Z multimeter would be used to this, but maybe that is what your test was for

Tiger.

[c4757p]

Just ground the case.

What is it coupled to or from?

My 60s tube voltmeter (before I added a grounding strap to the case) had 40-60VAC to ground on the metal case when probed at high impedance. I measured the impedance of the voltage by probing at two known input impedances - upper megohms. There isn't anything touching the case at all (it originally had the common pin connected to the case, but this really bothered me so I changed it) - it's all just EMI pickup from (I'm assuming) the transformer.

[TriodeTiger]

Ah! Touching the plate with my finger and the multimeter attached seems to serve the same purpose, the voltage nicely goes down to 0-20VAC depending on how much I allow it to discharge through me I guess.

Okay - so, should any product really do this now-a-days? I guess it is polarised and isolated with the transformer and all, so no need for ground lead ... (for safety purposes at least), but should it be added to protect from noise and possibly a fault in the future? Mains does go straight in the the machine, so it could possibly make the case live if something cracked near there...

I'll look in to grounding it, bet it'd look ghetto, and it's a strange plug so I'd have to get it from a separate socket or something and that'd look silly when moving it around to other places. Maybe I'll ignore it in this exception now that I am a bit more confident it is okay.

Tiger.

[c4757p]

Okay - so, should any product really do this now-a-days?

No, in my opinion - I'm sure you can get more qualified answers than "IMHO" though. Still, it's not a safety issue, it just looks like one at first glance.

[c4757p]

I'll look in to grounding it, bet it'd look ghetto, and it's a strange plug so I'd have to get it from a separate socket or something and that'd look silly when moving it around to other places. Maybe I'll ignore it in this exception now that I am a bit more confident it is okay.

Can't you change the plug?

[David_AVD]

Sounds like a typical "double insulated" appliance to me.

[c4757p]

Sounds like a typical "double insulated" appliance to me.

Good point. I usually think of plastic enclosures when I think of double insulation.

[TriodeTiger]

Sounds like a typical "double insulated" appliance to me.

I'd think that too, and it is why I thought at first "There's no way I should get anything relative to earth on a metal surface with polarised plug!"

I see no rating for it however, just SA on the metal label, and I am not sure I've seen other similar mechanical appliances be that isolating, having metal everywhere.

Can't you change the plug?

I'm pretty unsure, it's two prong in to the pedal internally, then three prong in to the machine! Funny, depressing the pedal doesn't seem to change the voltages between any of the prongs or maybe I am doing it wrong - but it doesn't seem like I can do much there.

I'm pretty sure bonding the metal stuff to neutral which leads to ground probably makes it less noisy but more dangerous, so maybe not

I'll post back on those repair forums, maybe the repair guy would be interested in hearing what seems to be the case.

Tiger.

[David_AVD]

Modifying a double insulated device to have a 3 pin plug is not legal over here as far as I know.  No idea about USA or Canada.

[c4757p]

I've got no idea about Canada either, or that specific case here (though I am a bit more familiar with our electrical laws generally). Would there be any specific reason for this, beyond the usual rules about modifying stuff in general? I mean, if it's double-insulated, why would it matter if the exterior is grounded?

[David_AVD]

I don't know the reasoning.  I could even be wrong about it of course.  It's just something that stuck in my head for some reason.   

[TriodeTiger]

We follow the same rules as the US generally, but... I'm not sure about modifying it. I could do something weird like open the pedal, strip the live/neutral cable out and replace it with a grounded one. I could then bond earth to neutral at the point of the pedal, or something, although that's kinda the same thing as mentioned about about bonding neutral to the case. Would that be bad at all seeing as I can't reverse the plug? *shrug* I won't, but who knows.

The earth on the socket is quite accessable still, wonder if there's some crazy ground-prong -to-wire thing I can couple to the plug and screw it near the socket on the machine.

[c4757p]

I could do something weird like open the pedal, strip the live/neutral cable out and replace it with a grounded one. I could then bond earth to neutral at the point of the pedal, or something, although that's kinda the same thing as mentioned about about bonding neutral to the case. Would that be bad at all seeing as I can't reverse the plug? *shrug* I won't, but who knows.

No! Nonononononono. Neutral is grounded back at your breaker panel and nowhere else. Period. Even neutral by itself can be dangerous.

[David_AVD]

We follow the same rules as the US generally, but... I'm not sure about modifying it. I could do something weird like open the pedal, strip the live/neutral cable out and replace it with a grounded one. I could then bond earth to neutral at the point of the pedal, or something, although that's kinda the same thing as mentioned about about bonding neutral to the case. Would that be bad at all seeing as I can't reverse the plug? *shrug* I won't, but who knows.

I would never dream about connecting ground to neutral in an appliance!   

[c4757p]

Do you know how an inductor works? If not, that's OK - think of it like a resistance that gets higher the faster you try to change the current. Also remember that your house's wiring has inductance.

If (not if - when) you get a current spike (motors turning on/off, especially), the neutral line briefly becomes a resistor, and can stand off that voltage. You will briefly have a voltage on neutral, even if it's "grounded" back at the breaker panel. Grounding it at the appliance will at best form an inductive voltage divider, cutting the spike in half - and half of dangerous is still dangerous.

[TriodeTiger]

Do you know how an inductor works? If not, that's OK - think of it like a resistance that gets higher the faster you try to change the current. Also remember that your house's wiring has inductance.

If (not if - when) you get a current spike (motors turning on/off, especially), the neutral line briefly becomes a resistor, and can stand off that voltage. You will briefly have a voltage on neutral, even if it's "grounded" back at the breaker panel. Grounding it at the appliance will at best form an inductive voltage divider, cutting the spike in half - and half of dangerous is still dangerous.

Umm, I dunno. It seems it doesn't matter where I reference the plate (my body, ignoring it loading the circuit), neutral, or the earth in the socket it still reads this higher voltage. I think it's just that coupling or leakage from the transformer. Neutral seems to be a few volts above earth, but, that's not uncommon from what I can gather, and some things use neutral to communicate.

Also, it's all happening when I don't have the motor actually doing anything, just flipping the switch and having power run (through the light, too) makes the voltage appear. Any voltage on neutral I think would be long gone since I haven't done anything in the past while.

edit: In terms of putting neutral on to the case - it's a return path, the resistor I put between neutral and the plate removed the voltage right? so it's not technically "earthed" .. but it certainly dissipates the charge permanently, right?

Tiger.

[c4757p]

Wherever the voltage is coming from, don't even think about joining neutral to ground.

Neutral seems to be a few volts above earth, but, that's not uncommon from what I can gather, and some things use neutral to communicate.

Yes, it's a few volts above earth, because the current is changing at 60 Hz. It will be much higher when the current change is faster. Anything that uses neutral to communicate will protect itself against those spikes - it's not hard to do.

[amyk]

Doesn't sound like a safety issue as it is; use AC current mode on the multimeter to see how much current you can draw through; a few tens of uA is perfectly fine. IEC60950 says up to 250uA is acceptable!

SMPS are well known for doing this: http://www.aplomb.nl/SMPS_leakage/Doc_ie.html

[Rufus]

You will briefly have a voltage on neutral, even if it's "grounded" back at the breaker panel. Grounding it at the appliance will at best form an inductive voltage divider, cutting the spike in half - and half of dangerous is still dangerous.

Bonding neutral to chassis is not dangerous because of spikes. It is dangerous because an open circuit anywhere in the neutral feed makes the chassis live through the load in the equipment. If you hold the chassis while plugging in you will be electrocuted if the live makes contact first for example.