Mains transformer goodness

[N_Cook]

Something to do while watching some less engaging TV. Unwound another large
toroidal mains transformer. Suspecting doubled up winding of the primary and
then for UK use joining opposite ends and so relying on 2 thicknesses of
lacquer to resist high voltage. Of course somewhere near the middle at some
point it fails catastrophically. Yes, burnt spot weld buried in the middle
of the primary. Anyone know what this duff winding technique is called ?
(reduces the number of shuttle passes by 2 must be the reason). Is there a
way of testing an unknown , but good , transformer for this winding pattern
?
How come this wiring procedure is not outlawed ?
More generally, someone in production told me that a "goodness" test for a
mains transformer is an open secondaries, no load, monitoring of the primary
current is useful, any truth/rationale in that. ?

 

[[email protected]]

Something to do while watching some less engaging TV. Unwound another large
toroidal mains transformer. Suspecting doubled up winding of the primary and
then for UK use joining opposite ends and so relying on 2 thicknesses of
lacquer to resist high voltage. Of course somewhere near the middle at some
point it fails catastrophically. Yes, burnt spot weld buried in the middle
of the primary. Anyone know what this duff winding technique is called ?
(reduces the number of shuttle passes by 2 must be the reason). Is there a
way of testing an unknown, but good, transformer for this winding pattern?

Not that I know of.

How come this wiring procedure is not outlawed?

Most of the transformers wound in this way don't burn out?

More generally, someone in production told me that a "goodness" test for a
mains transformer is an open secondaries, no load, monitoring of the primary
current is useful, any truth/rationale in that. ?

It tells you the inductance of the transformer. It doesn't tell you
how much heat is being dissipated in the windings and the core - for
that you have load the seconaries and watch how fast the output
voltage falls off with incresing load.

Since the permeability of the core declines with increasing
temperature, the inductance falls away as the transformer heats up, so
that magnetising current - and the power dissipation - goes up as the
transoformer gets warmer.

 

[Phil Allison]

"N_Cook"

Something to do while watching some less engaging TV. Unwound another
large
toroidal mains transformer. Suspecting doubled up winding of the primary
and
then for UK use joining opposite ends and so relying on 2 thicknesses of
lacquer to resist high voltage. Of course somewhere near the middle at
some
point it fails catastrophically. Yes, burnt spot weld buried in the middle
of the primary. Anyone know what this duff winding technique is called ?

** " Standard practice " ??

(reduces the number of shuttle passes by 2 must be the reason). Is there a
way of testing an unknown , but good , transformer for this winding
pattern
?

** Only by testing to destruction.

How come this wiring procedure is not outlawed ?

** It works, is not a recognised safety hazard and saves cost.

BTW: I have a 300VA toroidal with EXACTLY the same kind of failure here
at the moment, waiting for a replacement to be made. But it is lasted 8 or
9 years before shorting.

More generally, someone in production told me that a "goodness" test for a
mains transformer is an open secondaries, no load, monitoring of the
primary
current is useful, any truth/rationale in that. ?

** Yes.

If you checked a run of transformers and found one with a significantly
higher I mag than the others - then the core assembly or the primary
winding has been bodged. For example - an E core tranny where the
laminations are not fully interleaved or tightly enough packed to eliminate
air gaps between the Es and Is will show high I mag on test.

Toroidals normally have very low I mag figures, 10 to even 100 times lower
than similar sized e-cores. However, if the primary voltage exceeds the
maker's design value, expect to see I mag go through the roof. It will also
go through the roof if you try to use a toroidal made for 60 Hz mains power
on 50Hz.



....... Phil

 

[N_Cook]

"N_Cook"

** " Standard practice " ??



** Only by testing to destruction.


** It works, is not a recognised safety hazard and saves cost.

BTW: I have a 300VA toroidal with EXACTLY the same kind of failure here
at the moment, waiting for a replacement to be made. But it is lasted 8 or
9 years before shorting.



** Yes.

If you checked a run of transformers and found one with a significantly
higher I mag than the others - then the core assembly or the primary
winding has been bodged. For example - an E core tranny where the
laminations are not fully interleaved or tightly enough packed to eliminate
air gaps between the Es and Is will show high I mag on test.

Toroidals normally have very low I mag figures, 10 to even 100 times lower
than similar sized e-cores. However, if the primary voltage exceeds the
maker's design value, expect to see I mag go through the roof. It will also
go through the roof if you try to use a toroidal made for 60 Hz mains power
on 50Hz.



...... Phil

I'm amazed they last as long as they do , 400 turns or so of paired primary
wires layed together touching and thermal cycling / chaffing possibility
along the lacquer . Certainly not produced in a clean-room so any amount of
microscopic grit between the wires, what is the thickness of 2 layers of
lacquer to resist pk-pk voltage of 350V , let alone mains surges?

 

[PlainBill]

Something to do while watching some less engaging TV. Unwound another large
toroidal mains transformer. Suspecting doubled up winding of the primary and
then for UK use joining opposite ends and so relying on 2 thicknesses of
lacquer to resist high voltage. Of course somewhere near the middle at some
point it fails catastrophically. Yes, burnt spot weld buried in the middle
of the primary. Anyone know what this duff winding technique is called ?
(reduces the number of shuttle passes by 2 must be the reason). Is there a
way of testing an unknown , but good , transformer for this winding pattern
?
How come this wiring procedure is not outlawed ?
More generally, someone in production told me that a "goodness" test for a
mains transformer is an open secondaries, no load, monitoring of the primary
current is useful, any truth/rationale in that. ?

Way back in my dark past I worked in a plant where we wound our own
transformers. Indeed, the doubling of the primaries is standard
practice. Note that if your primary voltage is 240VAC the voltage
between the adjacent wires will be approximately 180V maximum.

Of course, these were using 'I and E' laminates. We tested the wound
bobbin for the correct number of turns and leakage between the wires
of the paired windings. After the laminations were added the entire
transformer was vacuum impregnated with varnish. Not much chance for
movement between the windings after that!

PlainBill

 

[Eeyore]

I'm amazed they last as long as they do , 400 turns or so of paired primary
wires layed together touching and thermal cycling / chaffing possibility
along the lacquer .

You are describing all transformers.

Certainly not produced in a clean-room so any amount of
microscopic grit between the wires, what is the thickness of 2 layers of
lacquer to resist pk-pk voltage of 350V , let alone mains surges?

The lacquer will resist plenty more than that.

Graham

 

[N_Cook]

Way back in my dark past I worked in a plant where we wound our own
transformers. Indeed, the doubling of the primaries is standard
practice. Note that if your primary voltage is 240VAC the voltage
between the adjacent wires will be approximately 180V maximum.

Of course, these were using 'I and E' laminates. We tested the wound
bobbin for the correct number of turns and leakage between the wires
of the paired windings. After the laminations were added the entire
transformer was vacuum impregnated with varnish. Not much chance for
movement between the windings after that!

PlainBill

Did you ever metal saw across any of them to actually check on the
impregnation. ? Not transformers, but I've seen such a check on a similar
process with much the same viscosity of fluid and most of the interior was
not impregnated, vacuum gave out to viscosity down fine pathways.

 

[Eeyore]

Did you ever metal saw across any of them to actually check on the
impregnation. ? Not transformers, but I've seen such a check on a similar
process with much the same viscosity of fluid and most of the interior was
not impregnated, vacuum gave out to viscosity down fine pathways.

You're worrying about stuff that's not important.

Graham

 

[N_Cook]

Way back in my dark past I worked in a plant where we wound our own
transformers. Indeed, the doubling of the primaries is standard
practice. Note that if your primary voltage is 240VAC the voltage
between the adjacent wires will be approximately 180V maximum.

Of course, these were using 'I and E' laminates. We tested the wound
bobbin for the correct number of turns and leakage between the wires
of the paired windings. After the laminations were added the entire
transformer was vacuum impregnated with varnish. Not much chance for
movement between the windings after that!

PlainBill

In the ideal yes, but this Tx and another similar there is no interlayer
insulation between each of the 3 or 4 primary "layers".
Compact winding only on the inside diameter, not the outside , where
windings
could easily slip a "layer". So worst case voltage could be 240/0.707 , peak
to peak, without any mains spikes.

 

[Ecnerwal]

I would really worry about inhaling/handling the slightest amount of PCBs.
Especially if the unit was manufactured before the '70s. (@_@)

Then you'd be somewhat less than brilliant. No mention of transformer
oil was involved, and nobody is likely to uncan an oil filled
transformer in their living room. A dry, varnished/lacquered transformer
has no oil, so no PCBs.

 

[N_Cook]

Then you'd be somewhat less than brilliant. No mention of transformer
oil was involved, and nobody is likely to uncan an oil filled
transformer in their living room. A dry, varnished/lacquered transformer
has no oil, so no PCBs.

I did think elf 'n' safety a week ago, opening a pack of silicone rubber
sleeving. All this fine glistening deposit everwhere , will I get silicosis
?

 

[Phil Allison]

"Eeysore"

You are describing all transformers.

** Totally off with the fairies bollocks.

The lacquer will resist plenty more than that.

** Totally off with the fairies bollocks.


....... Phil

 

[Phil Allison]

"N_Cook"
PlainBill

Did you ever metal saw across any of them to actually check on the
impregnation. ? Not transformers, but I've seen such a check on a similar
process with much the same viscosity of fluid and most of the interior was
not impregnated, vacuum gave out to viscosity down fine pathways.

** Nevertheless, vacuum impregnation works very well with transformers.

The basic idea is to remove all the air spaces between windings and fill
them with insulation material - so the unit will never suffer from "corona"
discharge.

http://en.wikipedia.org/wiki/Corona_discharge#Problems_caused_by_corona_discharges

Corona discharge is the biggest killer over time of transformers that handle
AC mains or higher voltages - particularly valve output transformers and
EHT transformers.

Until fairly recent times, vacuum impregnation was standard practice for all
mains voltage transformers - toroidal types were wound with porous cloth
tape insulation to allow this step to be done.

Now makers use polyester tape insulation along with tougher, flexible
enamels and impregnation is impossible.

IMO - a mains toroidal with dual primaries ought to have them wound
separately and with at least one layer of poly tape in-between.



...... Phil

 

[Jasen Betts]

I'm amazed they last as long as they do , 400 turns or so of paired primary
wires layed together touching and thermal cycling / chaffing possibility
along the lacquer . Certainly not produced in a clean-room so any amount of
microscopic grit between the wires, what is the thickness of 2 layers of
lacquer to resist pk-pk voltage of 350V , let alone mains surges?

peak to peak is half of that between series-connected windings.

 

[Adrian C]

I did think elf 'n' safety a week ago, opening a pack of silicone rubber
sleeving. All this fine glistening deposit everwhere , will I get silicosis
?

Though if the deposit had been from something dowsed down from a fire at
over 300degC, it's likely to be bloody painful amounts of Hydrofluoric
acid. Be ready to lose fingers :-(

 

[Fred Abse]

I did think elf 'n' safety a week ago, opening a pack of silicone rubber
sleeving. All this fine glistening deposit everwhere , will I get silicosis
?

Probably French chalk, AKA talc. Put there as a lubricant.

 

[Eeyore]

Probably French chalk, AKA talc. Put there as a lubricant.

Yup, totally normal.

Graham

 

[Eeyore]

"Eeysore"

** Totally off with the fairies bollocks.

Well not ALL, but lots.

** Totally off with the fairies bollocks.

5 - 10 kV according to this site.
http://www.awtad-fzco.com/ecwire.htm

Graham

 

[whit3rd]

Was this PCB stuff overblown about being dangerous?  I worked for
Westinghouse motor and transformer repair in the 40's. Was up to my elbows
some times in PCB oil on transformers. I am still here.

There's three issues with PCB oils. First, they're fireproof (a
safety plus).
Second, the oil can overheat and decompose, into dioxins and other
toxins/carcinogens. This, is a safety minus. I've heard of a
contaminated
animal feed episode that got LOTS of political attention. The third
one,
is kinda scary: like DDT, the compound persists and spreads in the
natural environment, so it can cause problems in future decades or
centuries if it isn't collected and neutralized. No one wants to
take the risk of causing the next species extinction and/or lawsuit
flood.

 

[N_Cook]

Anyone care to speculate on the timeline of failure?
There was no local hotspot and no lacquer damage more than a mm from the
"spot weld", and a bit of very localised smoke staining travelling a cm or
so along the affected wires in each direction. Nothing to suggest that the
initial bridge was between the 2 bifilar wires of the primary, so not
running a 120 volt primary in effect on 240V ac for any time. For 240V use
the 2 primaries seriesed to give about 2.8 ohm originally , after failure
then about 0.4 ohm.
The 2 primaries broke into 5 lengths
9.6,14.6,20.1,20.1 and 25.3m long , measured to about 0.2m accuracy.
So originally probably 2 x 45m. Don't know for sure as did not think to
check but the weld was probably 20.1m from one end, but I would suggest that
bridge occured after an arc to another layer (higher pd) and then localised
heating to bridge across to the bifilar fellow wire.

In summary , no evidence that bifilar wiring itself was the reason for
failure but more due to the lack of any interlayer insulation. Because of
the uneven wire spacing between inside and outside faces of the toroid it is
too easy for the layers to be jumbled.