Gaming device potentially live case - help diagnosing needed

[B737]

ive got a gaming device that has lots of switches, servos and backlighting in it. It’s powered by 5v for main power, 5v for servos and 12v for backlighting.

It all sits in a metal case.

Just recently I’ve had a strange problem when using it that when I touch certain switches or even if I touch the metal case then my USB joystick on my pc disconnects and then reconnects. I narrowed this down to potential electrical interference (as I had disconnected everything from the gaming panel to my pc) and put 4 ferrite beads on the usb cable from joystick to pc which stopped the disconnects of the usb device when using the overhead panel but only when I used 4 x 5mm ferrite beads. The disconnects did reduce if I had 3 ferrite beads on but 4 seem to stop it all together.

This now leaves me with an overhead panel that may have a short circuit and a live case?.

My question is how do I best go about checking this with my multi meter.

I was advised to power the unit then plug one probe of the multi meter into a spare earth socket on a plug socket and then using the other probe touch the gaming metal outer panel (on a bare metal section - screw hole).

I tried this when it’s powered and got the following results when setting the multi meter to DC 2v. There was a lot of fluctuations and the range is listed below:-

All power leads connected = -1.8 to 1.8v
Just 5v main and 12v backlight = -1.5 to 0.5v
Just 12v backlight = -0.8 to 0.7v
No power = -0.3 to 0.2v

How can I have a reading when no power is connected?

Given the minus readings in the above tests I set it to AC 200v next. Here are the results:-

All power leads connected = 85
Just 5v main and 12v backlight = 83.4
Just 12v backlight = 63
No power = 20

I next tried a simple circuit test by unplugging the device and putting one probe on live plug pin and the other on neutral plug pin. My multi meter was set to circuit with audible sound. I expected to hear a sound as I complete the circuit but I didn’t? Any ideas?

The gaming device is actually fully working and functional with the exception of a couple of broken servos so in my mind the circuit tests should have worked?

Was my earth to case test even valid?

Any suggestions on how to diagnose this further?

Are the readings from above even a concern?

Thanks in advance

 

[B737]

It’s not home made.

Photos of the unit are attached.

It has 3 x jack plugs therefore I assume they are not earthed being jack plugs however the kettle leads that plug in to the transformer have and earth pin.

Should I run a wire from the case to earth?

I’m not even sure there is a fault with it. It was purely the fact I had to put ferrite beads on to a joystick usb to stop it interfereing with the joystick. And then when I did a voltage test on the case and a probe on earth that I got readings.

 

[Lucien Nunes]

The problem is very likely to be AC earth leakage from the power supply units / adaptors. The AC voltage you are seeing between the case and earth can result from normal leakage from the in-built interference filtering, or due to a fault. Only a proper safety test including insulation and earth leakage tests will be able to determine which of these two applies. The voltage itself is not very important - much electronic gear floats up to around 120V when the case is not otherwise earthed, it has no real significance because the source of that voltage is of such high impedance, and does not in itself indicate danger. Electricians often call it 'ghost voltage' which simply means a voltage detectable on something that is more or less floating from earth, where the leakage current is so small it's trivial and may be impossible to measure consistently.

If the DC outputs of the power supply units are NOT internally earthed (i.e. they are isolated, a.k.a. floating), and what you are seeing on the case of the device is nuisance leakage that cannot at the moment escape to earth, you could eliminate it by earthing the common DC negative terminals where they feed the electronics, and/or the case which is probably in contact with DC negatives already. If the power supplies ARE internally earth-referenced, then you have a problem with a broken earth connection somewhere in the mains power wiring because the earth connection is apparently not reaching them. You can confirm the internal earthing configuration by unplugging them from the mains, and using a multimeter on resistance to measure between the earth (centre) pin of the IEC inlet, and the negative of the DC connection to the electronics. An earth-referenced power supply will show a resistance anywhere between a fraction of an ohm to a few kilohms.

E2A the connectors you are referring to as jack plugs, which I understand are on the DC outputs of the power supplies, do not have a safety earth connection as they are for 5/12V DC. However, if the power supplies are internally earth-referenced, the earth connection is carried through the DC negative lead so although it is not present as a separate connection, anything in contact with signal & logic grounds (e.g.USB leads, the case, the internal PCB grounds) may be supposed to be at earth potential.

 

[Darkwood]

That trumps my PS4 controller

 

[B737]

So if I disconnect it from the mains and put one probe of my multi meter on Earth pin just inside the transformer and the other on the neutral pin I should get a circuit sound (bleep) and display some resistance? Will this also work if I do it on the live pin?

I guess then I repeat for the other 2 power supplies?

What will this prove? Should I see a particular reading to confirm it’s ok?

I’m struggling to know what to do next - maybe get an electrician to test it.

 

[B737]

Would you think I should Earth the case then?

If so do I simply run a single wire from the case to a plug that just has the Earth wire in it and plug it in?

 

[Lucien Nunes]

Earth pin just inside the transformer and the other on the neutral pin

No, not the neutral or live! The negative side of the 5V or 12V DC output. You would be checking to see whether the incoming earth connection in the IEC is carried through to the DC output. If so, then there is no reason why the case of the unit should float up to some arbitrary AC voltage, therefore you have a fault which you must find. If the earth is not carried through the power supply unit, then it is legitimately floating and you can solve the problem by earthing the case.

If any doubt remains, the power supply units should be subjected to a standard 'PAT test' to ensure their integrity.

 

[static zap]

Could start by testing individual supplies in isolation.
(If none have voltage on them - may be getting generated)
.. Do any of instruments illuminate with "Electro Luminescence" -Blue green Glow. Those are a bit like
laptop /LCD-TV Backlights (Hi freq-inverter)

 

[DPG]

You need to find someone who is competent to do these tests for you. No offence intended, but you are putting yourself at too much risk here based on your level of electrical knowledge.

 

[B737]

Could start by testing individual supplies in isolation.
(If none have voltage on them - may be getting generated)
.. Do any of instruments illuminate with "Electro Luminescence" -Blue green Glow. Those are a bit like
laptop /LCD-TV Backlights (Hi freq-inverter)

There are lots of annunciators that illuminate when required. These are fed by the 5v main supply feed.

 

[static zap]

Stay safe , and be wary if any inverters are present , they are a hazard to other circuits as any of their output getting into 5V CMOS logic ,is a story with a sad ending. ( if its more recent design , I would hope for LED) ... this is where , experience comes in ...
And I've never touched anything that expensive looking !

 

[B737]

No, not the neutral or live! The negative side of the 5V or 12V DC output. You would be checking to see whether the incoming earth connection in the IEC is carried through to the DC output. If so, then there is no reason why the case of the unit should float up to some arbitrary AC voltage, therefore you have a fault which you must find. If the earth is not carried through the power supply unit, then it is legitimately floating and you can solve the problem by earthing the case.

If any doubt remains, the power supply units should be subjected to a standard 'PAT test' to ensure their integrity.

Hi

I did the above test today. I had my multi meter on circuit with sound and also tried it on 200 and 2k ohm and can confirm I never got a sound or a reading. When on ohm stayed on 1 which was the same as when not testing anything.

So I theory I could earth the case?

 

[B737]

Update.

Had an electrician round who just looked at the device and said the problem was simply that the metal case is isolated and the voltage reading is just jumping across from from the circuits inside (ghost/phantom voltage) and it can’t get away. It needs earthing to allow the AC phantom voltage to get away so I’ve ran a earth cable from the case and now I get 0 volts ac on my multi meter when testing the case to earth.

Next test is to see if I still need the ferrite beads on the usb cable.