Electronics > Beginners

AC leakage, does it matter?

(1/2) > >>

sjl:
I am trying to understand AC leakage from 'floating' power sources when compared to earth/building ground.

I have 4 devices that I am testing:

- Cheapo 5v DC output 2 prong SMPS
- Rigol DP712 Linear DC Power Supply
- Siglent SPD1168X Linear DC Power Supply
- Tripp Lite IS250 Isolation Transformer (converted into a 'technician' isolation transformer: no continuity between input & output, ground & output, chassy & output)

The SMPS was measured leaking 52 VAC RMS to ground, I measured AC current flow to ground at 82 uA

The DP712 was measured leaking 25 VAC RMS to ground, AC current flowing at 15 uA

The Siglent would not leak any to ground. (I guess the SPD1168X wins over the DP712 in this regard, even though the 712 is a more expensive PSU)  I wonder if there is an additional filtering circuit to remove this from the output.

With The Tripp Lite I measured 98 VAC leaking to ground, with an AC current flow of 32 uA
(would be using this when probing mains-powered DUT or for isolating a variac to work with lower voltage AC)


Questions:

1. Is the correct name for this phenomena simply 'AC leakage'? Would this be at all considered ghost/stray voltage?

2. Is there any danger to sensitive components (micros, ICs) from this voltage?  For example if the DP712 was powering a breadboard circuit without ground, and then a ground was introduced via an oscilloscope probe, could a microchip be damaged or affected by the leaking 25 VAC-- even though it is low current?

3. Why do some transformers appear to suffer like SMPS do-- even when there are no capacitors involved. I have seen leakage from transformers out-of-circuit.  Would this come down to more insulation between primary and secondary windings?  I read somewhere one side might be better shielded from the other-- why don't most bench PSU's and audio power supplies seem to do this? Other than the Siglent I have not been able to find a PSU or transformer that does not suffer from some form of leakage (I saw some with a linear audio transformer, torroidal and others).

4. Is it correct to say that in the SMPS situation it is a capacitive coupling issue via an input/EMI capacitor, and in the transformer situation it is also a capacitive issue between primary and secondary?

5. Ultimately should I be concerned with this phenomena in terms of building circuits, testing equipment, or am I wasting my time worrying about it?

Thank you!

Ian.M:
2. Mains leakage current can be deadly to sensitive components.
e.g. https://www.eevblog.com/forum/beginners/simple-fet-circuit-for-onewire-protocol-doesn_t-work-as-expected/msg1277272/#msg1277272
I wouldn't worry much about a few tens of uA leakage applied to components that have internal ESD protection (e.g. the input protection diodes present in most CMOS logic), but it could well be an issue working with highly sensitive components e.g. many types of RF FETs,  unprotected MOSFETs, etc.

AngraMelo:
Are you using a low impedance meter to check that? That could be "ghost" voltage

sjl:

--- Quote from: AngraMelo on December 06, 2018, 04:03:40 pm ---Are you using a low impedance meter to check that? That could be "ghost" voltage

--- End quote ---

I was using a Fluke 87V, so I believe that would be 10 megohm impedance

Gyro:
Welcome to the forum. From your questions, you already have a pretty good handle on this stuff.


--- Quote from: sjl on December 06, 2018, 06:26:43 am ---Questions:

1. Is the correct name for this phenomena simply 'AC leakage'? Would this be at all considered ghost/stray voltage?

--- End quote ---

The correct name is 'leakage current' (AC or otherwise) - the voltage you see is determined by the source voltage available, the impedance (resistive / capacitive) of the 'leak', and the input impedance of the measuring device. For SMPSs, the term 'Y Cap leakage' is in common use as it results, mainly, from the impedance of the Y rated safety caps typically connected between the input lines and the output / chassis.


--- Quote ---2. Is there any danger to sensitive components (micros, ICs) from this voltage?  For example if the DP712 was powering a breadboard circuit without ground, and then a ground was introduced via an oscilloscope probe, could a microchip be damaged or affected by the leaking 25 VAC-- even though it is low current?

--- End quote ---

Very little on a linear bench PSU. Definitely some danger on a 2-pin SMPS mains adapter though.

The leakage current will be as you measured for a normal rate of voltage change you get on the input mains sine wave. The problems can come from mains spikes or at switch-on. With fast rise-time edges like this, capacitance can couple much more current.


--- Quote ---3. Why do some transformers appear to suffer like SMPS do-- even when there are no capacitors involved. I have seen leakage from transformers out-of-circuit.  Would this come down to more insulation between primary and secondary windings?  I read somewhere one side might be better shielded from the other-- why don't most bench PSU's and audio power supplies seem to do this? Other than the Siglent I have not been able to find a PSU or transformer that does not suffer from some form of leakage (I saw some with a linear audio transformer, torroidal and others).

--- End quote ---

All transformers have some degree of capacitance between their primary and secondary windings. They are either wound side by side in split-bobbin construction, or one on top of the other with insulation between them (including toroidals). There is no real way of getting away from this stray capacitance, some transformers have an inter-winding conductive screen which is connected to ground to shunt most of the capacitance to ground. Some high spec bench meters even have two screens, one ground screen and one guard screen, connected to the floating input guard terminal. Even with an inter-winding screen, the secondary of the transformer will have capacitance to the screen, so assuming, say, a 50V secondary on a variable PSU, you could potentially see 25V leakage current on the secondary - completely unrelated to anything from the primary.


--- Quote ---4. Is it correct to say that in the SMPS situation it is a capacitive coupling issue via an input/EMI capacitor, and in the transformer situation it is also a capacitive issue between primary and secondary?

--- End quote ---

Yes exactly. Y-Cap leakage and capacitive leakage across the high frequency transformer windings.


--- Quote ---5. Ultimately should I be concerned with this phenomena in terms of building circuits, testing equipment, or am I wasting my time worrying about it?

--- End quote ---

With Linear bench PSUs you are probably worrying a little too much - I don't know the two particular models you're talking about, but many include a ground terminal that you can connect the (normally negative) output to, to remove any remaining leakage when working on sensitive parts. 2 Pin SMPS adapters, you should definitely worry about when working with sensitive parts.

Navigation

[0] Message Index

[#] Next page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod