[neurofen]

im not much of an electrician but i think i have a problem houston .

the psu ive just built (which i avent turned on yet thankfully) is creating continuity between the earth plate and almost every connector that there is.

bit stumped..

the earth comes from the ac power input straight to a metal plate that the various components sit on, there is not direct link to earth from any of the 3 gecko drivers/breakout board, toroid, capacitors or rectifier? yet they all create a circuit if tested with the voltmeter.. i have lifted up the various components away from the plate and put them on little non conducting blocks. but theres still loops

help.

[Al_The_Man]

If you are testing with a modern electronic meter it is not unusual to measure stray voltage currents due to the high impedance of the meter, at least that is what I suspect is happening, it is better to check with a resistance check, an insulation tester is even better but not always available.
If you use a voltmeter, use a resistor between 500 & 100 ohms, depending on how high the apparent voltage is, across the meter and see if the voltage collapses? If it remains the same then you have a problem, otherwise it indicates stray inductive or capacitive leakage.
This is just ONE of the reasons I prefer to build a system with all commons bonded to earth ground rather than the 'isolate everything' method.
Al.

[neurofen]

thanks again al, i have an electronic meter and was using the setting on it which looks like a diode.. continuity i thught.

im going to have to hunt around and see if i can find a resistor, but if i test for voltage, on the dc side of the transformer... i get a reading of 0.16v between the earth and all positive terminals (nothing anywhere else )... i think i might have to learn to use my voltmeter. does this indicate a normal reading?

[Al_The_Man]

I reread you post and when you said you used a voltmeter, I took it to be in the voltage mode, not the diode test range, you really need to try it in the low ohms range on the other side of the meter.
So in that case you do not need the resistor unless you power on and use the voltage range.
If you show a definite low resistance on the ohms scale, then you would need to lift connection one at a time and isolate the point it is grounding.
BTW, your meter shows a battery sign, is you meter battery flat, My Fluke indicates crazy reading when the battery is low.
Al.

[neurofen]

i'll do a battery change, hadn't noticed that. thanks

[neurofen]

ok so with a battery change and using the resistance setting marked 200ohms, theres is zero reading between earth and any of the connections except the negative terminals on the dc side where i get a reading when it settles of 0.3 (it peaks monentarily at around 100 then drops to 0.3)

[Al_The_Man]

If you mean by Zero reading, zero Ohms then it indicates a ground connection somewhere, I would first disconnect the load from the power supply and test the various sections of the power supply first.
If still a low reading, start separating the various components/connections to isolate the offending connection, once you have the power supply satisfactorily isolated, then check the load the same way, before connecting to the power supply.
Al.

[neurofen]

i should maybe explain that it isnt plugged into the mains and never has been

[Al_The_Man]

I realized that, the post #7 still applies.
Al.

[mmcp42]

surely when a PSU is switched off (like yours is) it's a direct short to ground all by itself!

try disconnecting the PSU and measuring

then try just turning on the PSU and see if it generates a voltage

I'm not convinced that you actually have a problem

(but I won't replace your kit or house if it all goes 'orribly wrong )

[Al_The_Man]

surely when a PSU is switched off (like yours is) it's a direct short to ground all by itself!

Please Elaborate!
Al.

[neurofen]

ok, so ive disconnected the load from the power supply, and then dissonnected the capacitors/ without them attached theres a reading of about 12 on the voltmeter in resitance mode on the primary side of the toroid (between live and neutral) and then about 0.5 on the secondary side.

should there be continuity between the bases of the 2 capacitors? if so they should be electrically isolated form the earth plate that they sit on right?

Is it safe to plug in the toroid to the mains, without the capacitors attached and only the ac side of the rectifier wired to it (the dc side is unconnected)

does the toroid in anway hold charge?

[Al_The_Man]

I though we were checking for continuity to ground on your power supply?
On the primary the resistance is going to be low between the supply, and you will also get continuity to ground via the grounded neutral, thats normal.
It is the secondary onwards that you need to be concerned about if you need isolation from ground.
The toroid secondary is not normally grounded unless done intentionally, this goes for the DC output of the bridge and capacitors, I would doubt the case of the electrolytic's are at ground, but anything is possible.
If one was to ground either the secondary or bridge output, the other MUST be above ground, other wise you would short circuit the rectifier.
It would appear from your meter reading that the cases ARE common. so there could be your problem?
Check continuity between both + and -ve connection to case.
Al.

[neurofen]

yes, i had continuity between the 'earth' plate and most of the connections on the various components. there is no continuity at the moment between earth and either the positive or negative poles of the toroid or the rectifier without the caps attached.. but as soon as the capacitors are attached and sat on the aluminium plate the bottoms of them make a circuit with earth..

al, when you say ground-- do you mean neutral/ground common 0v - or do you mean earth ground?

as for the readings between neutral(common ground) and live, i was just checking they were as they should be

i think the problem may be that the bottoms of the caps are positive(neutral, common ground.. whatever teh f@@k its called ) and create a cirucit with earth

[EPSfoamCutter]

Hi, If your trying to see if you have DC power or not the thing to do at this point is to just set your meter to DC Volts and see if you have the voltage you should be getting out of your supply. If not check the ac input side.
If you have AC in then you should get DC out.

[neurofen]

the problem seems to be this (as a note i've wired everything up again except the capacitors bases are now sitting on the table top)...

the bottoms of the capacitors are part of their positive circuit. so as soon as i screw the capacitors to the aluminium base plate (which is wired stright to earth) eveything shorts to earth. i will have to mount the capacitors on something seperate.

[EPSfoamCutter]

I see. That is not a good thing. Can you give me the part number for the caps and the size of the caps?
Also can your metter test the caps to se if they are good. you may have a shorted cap.
I would check the caps first before taking the cans off the eath ground. The caps are cheap in cost to try and just replace them.

[neurofen]

epsfoamcutter.. the caps are EPCOS B41580A7159M

both of them have continuity between the top positve screw hole and the very bottom of the capacitor itself.

I'm not sure how i'd take a capacitance reeading..


http://www.farnell.com/datasheets/624147.pdf

[EPSfoamCutter]

Ok the microfarad of marked uf rating for this cap is 15000UF and 40V.
Just checking the cap with continuity or ohms is not a good check you need to check them with an LCD meter so you can check the microfarad of the cap, BUT you should not short the + tap to the can here is a link to Allied. LINK: EPCOS - B41580A7159M - Allied Electronics they only cost 10.67. You should replace them. When caps short they can go bag sometimes and mess up thinks in the box with them. You should check the power to then too, to see if it fried the caps.

[neurofen]

ok, so i did some hunting and from the spec sheets supplied by the capaictor manufacturers.. they say this

"The bases of types with threaded stud are not insulated"


so i'm going to mount them on some nylon blocks. that'll solve that one

thanks for all your input everybody.. i'm going to give the suppliers a ring in the morning see if they can tell me why the bases should be uninsulated? there must be a reason.

sleep easy