Power transformer grounding

[catalin gramada]

I really don.t know how much the core is completly isolated, from a safe point of view.There is a capacitive coupling in between. Sure a crimp washer between core and end bell will assure firm contact, although saw a lot of PT which did not used it and end bells covered with a thick layer of unconductive paint. Even so,should conduct enough to blow the fuse and not keep the frame under voltage potential.

late edit: is not a "research", this thinks should be very basic and in consequence important. Unfortunately I haven.t a electrical formation and infos found are not in conformity with what I tested. The electrostatic screen tied together with HT center tap did less noise ,in my case, than tied to chassis to one of the transformer bolt. I haven.t a earth contact . So, should I use the electrostatic screen in this condition , or not ? IF Yes, where You consider should I tie it if I don.t benefit by earthing, please ?Thanks

[SoulFetish]

As far as safety is concerned, wouldnt it be advisable to at least use a Ground fault interupt?

[catalin gramada]

Hi. Why I insist to use the electrostatic shield even I don,t benefit by earthing ? Because I found the current between transformer frame and HT center tap is reduced by a 9:1 ratio when I used it. The imediate consequence was noise reduction on my board and I supposed closing loop in this way affect how internal windings are inductive coupled with end bells and reduce its emmision...perhaps. Also found the minimal noise moving the electrostatic shield wiring from transformer bolt (eg. as Hammond internal connected did) to HT center tap.
That is uncommon wiring so I was curious by Your oppinion if something wrong or short explanation about. Thanks.

[nickb]

Hi. Why I insist to use the electrostatic shield even I don,t benefit by earthing ? Because I found the current between transformer frame and HT center tap is reduced by a 9:1 ratio when I used it. The imediate consequence was noise reduction on my board and I supposed closing loop in this way affect how internal windings are inductive coupled with end bells and reduce its emmision...perhaps. Also found the minimal noise moving the electrostatic shield wiring from transformer bolt (eg. as Hammond internal connected did) to HT center tap.
That is uncommon wiring so I was curious by Your oppinion if something wrong or short explanation about. Thanks.

I believe I explained why it works in #9 above.

What worries me from a safety perspective is that you are reducing the amount of insulation between yourself and the primary winding when you connect the screen to your circuit ground. It would be safer not to connect the screen at all. You might try a connecting it through a capacitor that is small enough to conduct the noise but which still presents a decent impedance at your local mains frequency (50Hz?)

[catalin gramada]

HI. You direction from /9/ was well received and actualy you are right spot on what I intended to do. Thanks

[trobbins]

I'm not sure you are aware of the difference between electrostatic coupling arising from electric fields that allow current to flow through stray capacitances, and electromagnetic coupling arising from magnetic fields that allow current to flow around closed loop circuit paths.

A typically implemented transformer has bell ends, and is chassis mounted, with the core, bell ends and chassis all electrically bonded. A winding on that transformer has stray capacitance paths all around the extremity of the bulk winding to the nearest closest metal.

For a HV secondary winding in a valve amp, the many stray capacitance paths include to the heater windings (which are normally the outermost winding layers), to the primary winding (normally the innermost winding on the core), to the core (especially in the core window regions), to the bell-ends (which are physically a bit further away, but are sort of a good surround shield), and to the chassis (typically where the bell ends may be 'open' for certain types of bell-end format). If you electrically separate/isolate any of those surrounding parts and then apply voltage to the windings, and then connect an AC voltmeter to that isolated part and any other part you will get a reading (which depends on the voltmeter impedance and the parasitic capacitance path that form the measurement loop).

The thread started with I think sort of wanting to discuss what you may have thought was electromagnetic coupling. A transformer exudes stray magnetic fields that don't end up coupling directly from a winding to the core. Exotic winding layouts can reduce the level of stray field. Totally enclosing conducting loops (such as a belly band, and even bell ends and chassis) can attract that stray field in the form of an electrical short circuit path. Nothing is ever perfect, so stray fields still eek out, and can affect other circuitry (the old technique of physically orienting the location and direction of power and output transformers on a chassis top by listening to an output transformer winding with headphones whilst the PT was powered is a good example).

So typically the best thing to do for audio ends up being the best thing to do for safety - don't float metallic parts, but rather keep them all tightly and electrically bonded.

The electrostatic shield acts as an electrical screen between primary and secondary windings. For most equipment, that screen is used to bypass stray capacitance currents from going between primary and secondary circuits, and rather shunting them to say the chassis - those parasitic capacitance currents still flow, but as effectively separate loops now. Where the ES tap is connected is usually a benign issue (ie. you won't hear or notice any difference), but if you are keen you can try and rationalise what capacitive current noise flows through the screen tap as a result of the voltage waveforms on each side of the screen, and then determine if it is better to connect the screen to the 0V audio star, or the chassis.

As an aside, equipment working at very low noise levels would typically use transformers with two electrostatic screens - one associated with the primary and the other with the secondary, and the screens may be more exotic than just a simple foil open winding. Some power transformers must insert a protective earth screen that is rated at the full fault current of the primary winding, as an additional safety requirement.

[catalin gramada]

Hello. Thank for Your useful infos. I will establish an order of priorities starting with safety condition.
You guys are amazin. In one test I did it fooling around this EL shield I even get spark between transformer bell and one closed positioned can cap with both tied firm to my floating chassis. That.s somethin.no one wants to happen , specially at 800v. In absence of effective earthing I decided to suspend this shield for safety consideration. Thanks