Troubleshooting a DVD player with case at ~170VAC

[Whales]

Hey guys,

My family DVD player (we still have one) has been sitting disconnected for several months after I discovered its case was being held at a non-ground potential.  I accidentally noticed that dragging a HDMI connector across the back of the case yielded small sparks and from there I investigated.  Whilst my TV is earthed the DVD player is a two-prong AC plug affair, and both function perfectly fine.

Measuring the case with my non-true RMS meter reveals a case pot of about 177V AC compared to house ground.  The ground in the house is functional.   Grounding the case over a 10K resistor yields a 400mVAC drop (less than a mA average) and over a few ohms it is not really measurable with  my DMM. ( ~0.003 VAC)

Running the DVD player's power supply after removing it yields the same results, so I know the problem is isolated to this one board.

This is the left (primary) side  of the PSU board.  What look like fine separations in the ground plane around the edge screws are not: they are just black silkscreen I didn't notice to clean up 



I can only see a few ways my problems could occur:

C110 (MOV?) lets too much current through
The small amount of capacitance this has would let some  mains through, but if it were big enough to cause this problem then why would they put this on a non-grounded device?  DC resistance checks out to be above my meter's max of 20 meg, so it's not a short.

EDIT:  C111 and the optocoupler both measure good for DC resistance too

Insulation on the large transformer has gone
... but surely then the device would have its voltage taps skewed unevenly -> things would be out of spec -> the device would not work.  When I discovered the problem it was still working.

Fault on the secondary side
Unlikely given the voltage levels & the fact it's AC.


Should I try removing C110?  As far as I can see it's for noise suppression, not safety, but I could be wrong.  Is this behaviour normal for non-grounded devices that use their case as ground return path?

[Monkeh]

C110 should be a Y-class capacitor, and is where the leakage is coming from. It should be a very small current, <3mA certainly, typically below 1mA.

Removing it should remove the AC bias on the chassis. It's quite normal, though.

[Whales]

So it is safe to assume that although the case is held at a potentially lethal voltage, it cannot deliver enough power for long enough to cause harm?  Ie it drains a small capacitance when you touch it and cannot maintain anything near a dangerous voltage if you continue your contact? I don't like the sound of this 

If the device is only able to leak a few mA though the case then I will just keep it earthed through the cables to the TV rather than remove these caps.    As I understand the caps provide some noise suppression, so it would be better to leave them if it's safe.  Ideally I would add a three prong plug and properly earth it, but mounting that would be difficult.

It's quite normal, though.

The case of this tingled my fingers when I played with the HDMI cable.  I've now just lost a lot of faith in non-grounded appliances 

[EMUD]

Hi,
this is a capacitor to keep the CE conformance (radiated emissions). This caps (usually some 100 to some 1000p ) are needed to keep the switch mode power supplies in conformance with UL and CE. To avoid such voltage there are some methods:
- buy only things wich have are earthed by mains earth (3 pin plug)
- do not use things with a primary switched mode supply


EMUD

[amyk]

So it is safe to assume that although the case is held at a potentially lethal voltage, it cannot deliver enough power for long enough to cause harm?  Ie it drains a small capacitance when you touch it and cannot maintain anything near a dangerous voltage if you continue your contact? I don't like the sound of this 

If the device is only able to leak a few mA though the case then I will just keep it earthed through the cables to the TV rather than remove these caps.    As I understand the caps provide some noise suppression, so it would be better to leave them if it's safe.  Ideally I would add a three prong plug and properly earth it, but mounting that would be difficult.

It's quite normal, though.

The case of this tingled my fingers when I played with the HDMI cable.  I've now just lost a lot of faith in non-grounded appliances 

400mV over 10K is 40uA - well below the applicable safety standards (I believe it's something like 250uA max) and so perfectly safe.

[Whales]

C110 should be a Y-class capacitor, and is where the leakage is coming from. It should be a very small current, <3mA certainly, typically below 1mA.

Removing it should remove the AC bias on the chassis. It's quite normal, though.

 
I might remove it then.  If it causes reception issues for my TV I will consider placing it back in.

400mV over 10K is 40uA - well below the applicable safety standards (I believe it's something like 250uA max) and so perfectly safe.

In this continuous situation a definite yes, but upon initial contact it's enough current to make small arcs. 



From a logical point of view this appears to not be a safety issue and I'm letting my fear of superheated iron molecules gunning themselves off the back of my DVD player get the best of me.  I'll  worry some more and then hopefully get to a resolution I'm happy with 

Thankyou to everyone who has replied.

[EMUD]

Hello Whales,
the best solution will be to use an isolation transformer!

EMUD

[Whales]

Hello Whales,
the best solution will be to use an isolation transformer!

   Perhaps if I switched two leads on L102....

When I first looked at the board I was confused as to why there were two transformers.  "Why would they go to the effort of isolating this?  They didn't even want to pay for a discharge resistor across the main filter cap".  Tis a shame L102 is too small to actually make into a proper 1:1 transformer 

Always interesting: compare the size of the component on the silkscreen to what was used in the shipped product.

[EMUD]

Hi,
I think an external xformer one! Look, not only C110 but C111 will make a current flow to your case!

EMUD

[wraper]

Hurry to throw away most of mains powered devices around you which do not have earthed mains plug  . Did you check your TV as well, likely it is the same as your DVD player. Most 10+W and almost every 30+W double insulated SMPS will do the same.

[EMUD]

Hi Wraper,

absolutely right. I know some friends who have toasted their pc interfaces (it was RS232) of their notebooks by plugging them when they are connected to mains by such a power supply! The current is not so dangerous for us, but interfaces dont like this.

So, before making connections to such devices pull the mains plug ! This may be more safe for you, too.

EMUD

[macboy]

Hello Whales,
the best solution will be to use an isolation transformer!

   Perhaps if I switched two leads on L102....

When I first looked at the board I was confused as to why there were two transformers.  "Why would they go to the effort of isolating this?  They didn't even want to pay for a discharge resistor across the main filter cap".  Tis a shame L102 is too small to actually make into a proper 1:1 transformer 

Always interesting: compare the size of the component on the silkscreen to what was used in the shipped product.

L102 is not a transformer but a common-mode choke, part of the noise filtering to keep this SMPS from polluting (too badly) the mains with high frequency garbage.

There should be essentially no leakage through C110, because it is on the neutral side, which should be equal to ground... or very close, only difference due to V=IR drop in the house wiring from the breaker panel to outlet. Measure the voltage across C110, or measure from R100 (the side nearest the label, connected to C110) to ground. If there is more than a couple volts there, then you may have your live/neutral swapped, a potentially dangerous problem.

[EMUD]

Hi Macboy,

I have never seen a marked mains plug in Germany. So you will swap neutral and live. English plugs have a live and a neutral wire,
using symmetrical plugs you will not know where the live wire is.

EMUD

[Monkeh]

Hi Macboy,

I have never seen a marked mains plug in Germany. So you will swap neutral and live. English plugs have a live and a neutral wire,
using symmetrical plugs you will not know where the live wire is.

EMUD

The OP is not in Germany.

[David_AVD]

If you're concerned about the bleed voltage, connect the DVD player chassis to mains ground via a separate mains plug.  Use a series resistor if you like.

We've done the same thing in our workshop on TVs that we're plugging sources into for testing.  Getting the tingle almost every time you grab the video lead gets old after a while!

[Whales]

L102 is not a transformer but a common-mode choke, part of the noise filtering to keep this SMPS from polluting (too badly) the mains with high frequency garbage.

I know.  I was making a joke about rewiring it into an isolation transformer

There should be essentially no leakage through C110, because it is on the neutral side, which should be equal to ground... or very close, only difference due to V=IR drop in the house wiring from the breaker panel to outlet. Measure the voltage across C110, or measure from R100 (the side nearest the label, connected to C110) to ground. If there is more than a couple volts there, then you may have your live/neutral swapped, a potentially dangerous problem.

This player is powered by a figure eight cable (C7/C8) which can be plugged in either way.


Thanks for pointing this out however -- I will try reversing the polarity (as pointed out by Macboy) and see if the issue is diminished.

Hurry to throw away most of mains powered devices around you which do not have earthed mains plug  . Did you check your TV as well, likely it is the same as your DVD player. Most 10+W and almost every 30+W double insulated SMPS will do the same.

I'm trying to fix them rather than throw them out wraper    I've checked my TV -- its case is well earthed and the only other non-earthed devices in my house have plastic housings.  This DVD player is the only thing giving me issues.

If you're concerned about the bleed voltage, connect the DVD player chassis to mains ground via a separate mains plug.  Use a series resistor if you like.

We've done the same thing in our workshop on TVs that we're plugging sources into for testing.  Getting the tingle almost every time you grab the video lead gets old after a while!

I'm glad I'm not the only person who finds this uncomfortable.  Grounding the device with a resistor might not be the best: if no other grounded cables are plugged in and this resistor fuses then the case would be held at a high pot without tripping off my RCDs or blowing a fuse.  Granted this is a risk the current design already has.

At this point I'm going to attach an earthing cable to a screw on the back of the case.  It's non-invasive, simple, and probable safest in worst-case situations (ie y-caps going short).

Overall it's a device I'm happy to use as an individual knowing what to expect, but it's definitely not something I'd feel good about selling to other people en-masse.  I'm surprised non-earthed metal-cased devices are permissible, but hey, the regulation agencies draw a line between various factors somewhere and on the most part they do a good job.

[David_AVD]

The good ol' double insulated metal case sitting at half mains voltage issue has been around for a long time.

It used to be mainly TVs as they were probably the first A/V gear to use switching power supplies in a double insulated casing.

Early hi-fi gear used linear power supplies and didn't have those capacitors fitted to the double insulated models.

The explosion of switching power supplies in CD and DVD players is where the issue really started to be noticeable in my opinion.

[ornea]

I have experienced tingling when dragging my finger across a none earthed case and measured approx 80v A/C. Not pleasant.