60 V .3 MA from PSU on all rails is this normal?

[Codesmith]

I can sometime feel a current from the metal parts of my case. I meassued this with a multimeter and found it to be about 60 V .3mA AC.

The measurements are with me holding one prong and touching any wire from the PSU, or any conductive part of my case.

I checked and you need 16mA-110ma passing through your chest to pose any risk. http://www.cdc.gov/niosh/elecovrv.html

I get the same results from any of the rails when the PSU is on, but the system is off.

I get the similar results for multiple PSU plugged into multiple outlets while not connected to a motherboard and without shorting the green pin to start the PSU.

So either this is normal or my appartnemt has a wiring problem.

I would appreciate it if someone with a multimeter handy would tell me if thier PSU outputs any AC current when the system is on or off?

Or is someone who knows more about housewiring than me can tell me know how to test to see if my outlets are wired properly with a multimeter.

 

[blue68f100]

It soulds like you have a house wiring problem, Proably the earth ground. The earth ground is used to dispate stray current.

If you live in the US and have a 3 prong receptical. The 2 vertical slots are the main power, Line and netural, with the earth ground the round below them. You should measure 110-120vac across these (line netural). The Netural is the Longer Slot. If you measure from netural to the earth ground (round, below the slots) you should have 0 vac. Earth ground to Line should measure 110-120 vac. There are some cheap line tester that sell for $10-15 at your local hardware store. They have 3 lights which will give you the results and or problem.

Depending on the age of your house it was comon back in the 60's and 70's to try to use the metal box for ground, with out connecting it to the receptical. Then there were cases where painters would cut them off and not even connect them up. They also used and still do use the Quick connect connection. Never have liked these, over heat when exposed to high loads.

My dad's house (tree 10 ft from house) was struck with lightning bridging to his underground feeder to shop and to house plumbing. Blew out many breakers, connetions inside receptical boxes and Burned out the earth ground for his panel. With a scope connected across the earth and neutral it looked like a capacitor building up and dispating.

AC will take the path of least resistance, and want's to return to the source. Either your panel or line transformer.

 

[Kholonar]

There's two answers to this question. If you live in the UK then in theory you should never get built up charge on the metal casing as it should be grounded though the wall outlet. If you do live in the UK and are still getting this, try testing a new power wire on the computer, or, if you are an expert in electricity, check the wiring.

If you don't live in the UK then you won't have a grounding feature in your wiring and although build up of charge can mean faulty equipment, it's something you often have to live with.

0.3mA isn't enough to kill you. High voltages can cause burns and high currents can disrupt the heart but you aren't in any danger from either. I once experienced this build up of charge on objects when I went to france. The lamp in the hotel room became live with a small current, when I was turning on the lamp and put my feet onto the concrete I jumped about 5 feet in the air.

The PSU in your computer is encased in metal and touches the metal of your case. All it takes is for one piece of the PSU to touch the PSU protection and you get the build up.

**EDIT, just seen your post that the US uses grounding.

 

[blue68f100]

The higher voltages in the UK are actually saver than the lower ones in the US. Voltage 220 and > are powerful enough to knock you clear. 115 here in the US just makes you hold on, and will not knock you loose.

 

[maury73]

You didn't said how you take the measure. One probe to the case and the other? To the ground mains wire? To the PSU ground plane? To what?

 

[blue68f100]

Do ALL of the testing on the WALL receptical.

 

[Codesmith]

I held one probe while touching the other to my case to measuse the voltage and current I was recieving when touching the case.

Zero difference between that and simply touching the case, only this way I get a measurement.


I did the same with the power rails. Every line coming out of every PSU connected to any outlet has 60 v .3 mA AC.

I live in an apartment my outlets were all 2 prong but two were upgraded to 3 prong by someone who may not have know what they are doing.

Which maybe be why there is no difference between using a three prong outlet and 2 prong outlet via an adapter.

I just got a Black and Decker Home Wiring book from the library. By tomorrow I will know how to test and be able to correct any simple wiring errors.

 

[Pain]

You held one probe in your hand, and touched the other probe to the case?

That will measure nothing.

I did the same with the power rails. Every line coming out of every PSU connected to any outlet has 60 v .3 mA AC.

You're probably not measuring it correctly.

If the boxes behind the 3 prong outlets are metal boxes, then simply screwing them into the box will probably...probably... ground them sufficiently. It is not ideal and they should be physically tied to the metal box, but the screws will probably tie it to ground.

Now, lets talk about your house. For you to actually feel a voltage while touching the case, you yourself would need to be grounded [or touching some other potential]. It's sort of like a bird sitting on a power line. They are sitting on thousands of volts of electricity, but they aren't dead. Why? Because there is no current path. So, in order for you to feel the electricty, there has to be a path. Are you standing on a concrete floor in bare feet? On carpet? Wood? Are you sitting down? What kind of chair?

 

[blue68f100]

You didn't said how you take the measure. One probe to the case and the other? To the ground mains wire? To the PSU ground plane? To what?

I will make it simpler for you.

Set you Volt Meter to Volts AC range 0 - 200v

1. Unplug your computer from the wall.

2. Stick you probes in to the 2 Slots on the wall plug.
(slot to slot) Write your reading down.

3. Now move the Short slot Probe to the round hole.
(long to round) Write your reading down.

4. Now move the Long slot probe to the narrow slot.
(short to round) Write your reading down.

Post Results.

 

[chaynz]

Basically, here's what I got and what it should be. I used to have this problem all the time at my other house, come to find out it was because the outlets weren't grounded properly. At that time, I was seeing anywhere from 60vac to a full 115vac from the computer case to the common. Solution? The house ground was bad. The way the electrician fixed it was by joining the common (aka Neutral) and the household grounds to a water pipe.

Visit this website

http://hyperphysics.phy-astr.gsu.edu/hbase/electric/hsehld.html

 

[blue68f100]

Grounding to pluming is no loger recomended. Mainly due to the use of pvc piping. My dad was grounded that way and a by ground rod, when lighting struck a tree near this house. Bridge over and burned holes in the copper pipe under the slab. Expensive repairs. Use ground rod only from your distribution box.

 

[Pain]

It is code in my area that you must be connected to the cold water pipe and 2 ground rods. Most areas bring water into the house via copper, and the code specs to connect to that copper pipe within so many inches or feet [I forget the code] from where it enters the house. This is obviously because, as you said, you don’t know that all the copper in the house isn’t broken by pvc, so you have to go back to where it enters the house and attach the ground there.

As for testing at the outlet. You need to measure if the ground and neutral are tied together, i.e. you need to measure the resistance between ground and neutral. Measuring voltage between neutral and ground will not tell you if the ground lug is actually tied to ground....unless of course there is some voltage on it. If it's floating, you won't know it by measuring only voltage.

It is still possible that the earth ground at the panel is broken, even if the neutral and ground appear tied together at the outlet.

 

[blue68f100]

Pain, I have given him detail instruction on what to measure. I do not know or think he know what all of the terminology is. The first time He was tring to measuring the computer.

I worked on power distrubition systems from 110 to 13.4 KV and helped another person that did 134 KV. 134KV will make your hair standup if you get near it on humid days. DID NOT LIKE IT AT ALL. Hated going in to the sub-station to roll feeders.

PS require a reference to ground, or they float. It's posiable that he has a PS in the process of failing. But Can not know that till the source is checked.

It is highly possiable that his line and neutral are swaped.

If he had a Newer APC UPS he would know this, because they check the source. A light will come on saying Check Wiring.

 

[Pain]

I never said you didn't give detailed instructions.

I did say that code in my area requires ground tied to the cold water copper from the outside of the house. Basically for conversation, but whatever.

 

[blue68f100]

I think the codes changed with in the last 2 code releases (6yrs). I took a course for the Journmans Liencen 5-6 years ago, passed 1st try. They were talking about the change then. They covered your area and all older house requiring 2 grounding rods. The code books are only update every 3 yrs. Most cities run 1 - 2 books behind. And add other requirment to the code. My house uses the under ground conduct that the main feeder comes in with. Have not found where or if it uses the plumbing. A lot of panels have a bridge tying the Netural and earth together. Based on weather it is a main or sub.

I'm not sure of Codesmiths intention. He did ask for help. Then proceded on his own direction. Must be a geologist, there trouble shooting skills are backwards.

 

[Pain]

heh.

Yeah, I just replaced my panel and had to use 2 rods and a #4 copper to the cold water line, clamped above and below the pressure reducer. I have above ground service.

I have to add a subpanel in my house and replace the one in my garage too, and yeah, the main panel has the neutral and ground tied together, the subs have them seperated. That's so if you lose the neutral in the sub you don't have the possibility of current flow through the ground, but you know that.

 

[blue68f100]

I have a friend that has a house that is so old it was wired with just 2 wire, no ground.

He had to replace all the breakers with GFCI, you known that was expensive.

 

[rantonrave]

The 60VAC is caused by an ungrounded wall outlet connection and a pair of 5nF capacitors connected between each line and the case ground. Both capacitors conduct equally and therefore divide the 120V in half where they meet.

 

[Codesmith]

----
Stop and think for a moment. Anytime you can feel electricity, you are completing a path to ground and therefore you can use a multimeter to measure it with a multimeter by holding one prong and touching the other to the source.

(Assuming the current/voltage is withing the range of your multimeter and it lasts long enough)

60 volts of electricity is more than capable of flowing through a mans arm, down his body, through his shoes and carpet, through the carpet mat, wood flooring .....

---
Anway I know how to use a multimeter, and I know the basics of electricty from my physics classes,

So I understand volts, ohms, wattage, ohms law, flux, electromagentic fields ...

We never covered capacitance, impedence and AC current, nor did I know anything about how houses were wired until I got a book from the library.

Which is why five years ago I had my friend who has a 2 year Electronics degree put in some three prong outlets near my computer.

He knows that the bare wire is usually the grounding wire, but neither of us can remeber if he tested the outlet to see if it was properly grounded.
----
I read half the Black and Decker book and have tested all my outlets with a multimeter.

The three prong outlets that a friend put in near my PC have hot to ground 75v, neutral to ground is 44v.

It may not be connected to anything, or maybe they used repurposed the grounding wire for the doorbell or thermostat?

The only outlets in the appartment that are grounded properly are the GFCI ones in the kitchen that were recently upgraded due to a change in building codes.

With the other ungrounded outlets I get 40-60v hot to the center screw and 0-5v neutral to center screw.
---
I used an 3 to 2 prong adapter so nothing connects to the grounding wire as a temporary fix.

When I finish reading the home wiring book I will cut the power and take a look.

My guess is what I will find is what should be the grounding wire is connected properly, but its not actually grounded to anything.
---

I also know nothing about how powersupplies actually work other than what they due, convert 120 V AC to various DC voltages and keep those voltages steady depsite a constantly varing load.

So I am not sure what is ment my a Power Supply "floating".

Does that mean I should expect to see a small, but measurable current leaking to ground when connected to an ungrounded recepticle, or does that indicate that my PSU is somehow defective.

Sometime this week I am going to take a bunch of PSUs into the kitches, connect them to the properly grounded outlest via a 2 prong adapter then measure the voltage between the casing of the PSU and the grounding pin of the outlet.

If they all leak a small amount of current to ground then I will assume that is its normal for a PSU to do so if ungrounded.

---

 

[pcsteve]

Hey Pain, I totally agree mate. There is no reference for a true voltage to be measured..
Current is measured in series with the device and voltage is measured across (parallel) the device.
Holding one probe gives no true reference point.

 

[Pain]

----
Stop and think for a moment. Anytime you can feel electricity, you are completing a path to ground and therefore you can use a multimeter to measure it with a multimeter by holding one prong and touching the other to the source.

(Assuming the current/voltage is withing the range of your multimeter and it lasts long enough)

60 volts of electricity is more than capable of flowing through a mans arm, down his body, through his shoes and carpet, through the carpet mat, wood flooring .....

That's what I said about a current path, but I don't think you're getting any usable reading by holding a probe in one hand and touching the other to the case.

If you want to know how much voltage is on the case [or anything else for that matter] then measure directly from a known ground connection, to the case. I've done enough electrical/electronic work in my life to know that 60V isn't enough to pass through all that stuff and give a valid reading, or give a shock. If you were standing on a concrete floor in bare feet, yes. On carpet over wood with shoes on, no...unless everything is soaking wet.

With that said, I'd check it out immediately or get an electrician to do it. Often, if there is a float [or open circuit] you'll get various readings with a digitial multimeter. It doesn't necessarily mean you are reading all those various readings you say, but in any case, you need to have it checked ASAP just to be exactly sure what is happening.

 

[pcsteve]

Damn good advice. ALWAYS better to be safe than ... 8O

 

[Codesmith]

Sorry I didn't mean to imply I was getting an accurate measurement of what the PSU was actually leaking to ground.

I ment that my method is giving me an accurate measurement of the voltage and current actually passing through my body I when touched grounded metal on my case on my case. (grounded in the sense that it is connected to the ungrounded grounding wire)

It was constant, but very weak and didn't feel strong enough to be dangersous, but I wanted to measure it, not guess.

---
I have been wondering why I get shocked sometimes and not others. Now I think I found the answer. For example I don't think I was shocked once all winter.

I seem to only get shocked or have measurable current on days when it is extreemly humid.

Thursday and Friday it was extremely humid and I could feel and measure current from four different PSU's that were plugged in to a 2 prong outlets, but otherwise "off". (I wasn't shorting the green grounding wire). I also could get a shock from one monitor's signal cable, but no the other.

The results varied from .2 mA to .4 mA but managed to convince me that there was nothing wrong with my PSU other than it being connected to an ungrounded outlet.

Today I can neither feel nor measure anything. I even took the PSU's into the kitchen and took measurements using the ground from a properly grounded outlet. Whatever current there is its below the .1mA my meter can reliably measure.
---
I asked my appartment manger about getting the outlets near my PC properly grounded, but its not going to happen.

I can't even convince them to enlarge the slot on the outgoing mailbox so that my netfix envelopes will fit.

If I want properly grounded outlets I will have to move to an apparment complex with modern wiring.

 

[blue68f100]

The only way to get protection with a 2 wire (No earth ground) is with a GFCI. If you have access to the breaker for that circuit, you could take a look. The ground may have just been clipped and not connected. It was common pratice 30 yrs ago.

If they used metal conduct, as in commercial buildings, the conduit can be used for earth ground.

Without a earth ground a UPS unit will not work correctly. Has no path to ground. What a bummer.

Some old UPS units that were not switchers generated their own ground. Don't see these units any more.

 

[Pain]

Sorry I didn't mean to imply I was getting an accurate measurement of what the PSU was actually leaking to ground.

I ment that my method is giving me an accurate measurement of the voltage and current actually passing through my body I when touched grounded metal on my case on my case. (grounded in the sense that it is connected to the ungrounded grounding wire)

If you want an acurate measurment of the voltage on the case, measure to the case from ground. Your method isn't measuring anything.