Harvesting energy from 3-phase

[Clive Arthur]

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

The power needed will have to end up as a 30V 300mA supply, so 10 watts
or so which is a lot for a normal CT.

This is not something I've looked at before and sounds quite difficult
- a quick Google shows that this can be done for very low power devices.

Any ideas? Multiple toroids with secondaries in parallel? Long sausage
toroid? The energy has to come from the wire, obviously, so the
impedance of the wire will have to increase a little which implies that
some length will be needed, but now I'm rambling.

Is the following even in the same county as the ballpark?..
30V @ 300mA is 100R
50A -> 300mA is a turns ratio of 1:167.
Inductance goes with the square of the turns, so the wire must see 3.6mR
3.6mR at 60Hz is 10uH
10uH/N^2 cores are doable, maybe multiple
3 watts per core isn't too big

Cheers
--
Clive

 

[Jan Panteltje]

On a sunny day (Thu, 8 Aug 2019 09:55:47 +0100) it happened Clive Arthur
<cliveta@nowaytoday.co.uk> wrote in <qigo2o$ge2$1@dont-email.me>:

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

OK, 50 A, to get 300 mA (AC) you need 50 / .3 = 167 turns secondary,
to get 30V @ 300 mA your load needs to be 100 Ohm.
What is the problem?

You will need a decent iron core current transformer.

If you use all 3 phases 1/3 per phase, 3 phase rectifier bridge makes things simpler
and transformers smaller?

Clamp on cores?

The power needed will have to end up as a 30V 300mA supply, so 10 watts
or so which is a lot for a normal CT.

This is not something I've looked at before and sounds quite difficult
- a quick Google shows that this can be done for very low power devices.

Any ideas? Multiple toroids with secondaries in parallel? Long sausage
toroid? The energy has to come from the wire, obviously, so the
impedance of the wire will have to increase a little which implies that
some length will be needed, but now I'm rambling.

Is the following even in the same county as the ballpark?..
30V @ 300mA is 100R
50A -> 300mA is a turns ratio of 1:167.
Inductance goes with the square of the turns, so the wire must see 3.6mR
3.6mR at 60Hz is 10uH
10uH/N^2 cores are doable, maybe multiple
3 watts per core isn't too big

Cheers
--
Clive

 

On Thursday, 8 August 2019 11:52:47 UTC+1, upsid...@downunder.com wrote:

If this is a one off project, locate some current transformers rated
for 100 A. Standard current transformers have either 1 A or 5 A
nominal secondary current,

Then use some ordinary voltage up-transformers to lift the voltage to
20-30V before rectification. Depending of the connection (wye/delta)
select the connection for correct voltage.

Do you need the second transformer to increase the voltage? Wouldn't
the current transformer output voltage rise to a sufficient level
if the burden resistance is higher than the specified value. A
bridge rectifier and big zener or shunt regulator should work.

John

 

On Thu, 08 Aug 2019 09:08:47 GMT, Jan Panteltje
<pNaOnStPeAlMtje@yahoo.com> wrote:

On a sunny day (Thu, 8 Aug 2019 09:55:47 +0100) it happened Clive Arthur
cliveta@nowaytoday.co.uk> wrote in <qigo2o$ge2$1@dont-email.me>:

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

OK, 50 A, to get 300 mA (AC) you need 50 / .3 = 167 turns secondary,
to get 30V @ 300 mA your load needs to be 100 Ohm.
What is the problem?

You will need a decent iron core current transformer.

If you use all 3 phases 1/3 per phase, 3 phase rectifier bridge makes things simpler
and transformers smaller?

Clamp on cores?

If this is a one off project, locate some current transformers rated
for 100 A. Standard current transformers have either 1 A or 5 A
nominal secondary current,

Then use some ordinary voltage up-transformers to lift the voltage to
20-30V before rectification. Depending of the connection (wye/delta)
select the connection for correct voltage.

If all three phases are used, use three 5 or 10 VA current
transformers. If the voltage transformers do not load the current
transformer sufficiently for 1 or 5 A secondary current, you can
always use some burden resistors in parallel with voltage
up-transformer primary.

The power needed will have to end up as a 30V 300mA supply, so 10 watts
or so which is a lot for a normal CT.

This is not something I've looked at before and sounds quite difficult
- a quick Google shows that this can be done for very low power devices.

Any ideas? Multiple toroids with secondaries in parallel? Long sausage
toroid? The energy has to come from the wire, obviously, so the
impedance of the wire will have to increase a little which implies that
some length will be needed, but now I'm rambling.

Is the following even in the same county as the ballpark?..
30V @ 300mA is 100R
50A -> 300mA is a turns ratio of 1:167.
Inductance goes with the square of the turns, so the wire must see 3.6mR
3.6mR at 60Hz is 10uH
10uH/N^2 cores are doable, maybe multiple
3 watts per core isn't too big

Cheers
--
Clive

 

[John Larkin]

On Thu, 8 Aug 2019 09:55:47 +0100, Clive Arthur
<cliveta@nowaytoday.co.uk> wrote:

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

The power needed will have to end up as a 30V 300mA supply, so 10 watts
or so which is a lot for a normal CT.

This is not something I've looked at before and sounds quite difficult
- a quick Google shows that this can be done for very low power devices.

Any ideas? Multiple toroids with secondaries in parallel? Long sausage
toroid? The energy has to come from the wire, obviously, so the
impedance of the wire will have to increase a little which implies that
some length will be needed, but now I'm rambling.

Is the following even in the same county as the ballpark?..
30V @ 300mA is 100R
50A -> 300mA is a turns ratio of 1:167.
Inductance goes with the square of the turns, so the wire must see 3.6mR
3.6mR at 60Hz is 10uH
10uH/N^2 cores are doable, maybe multiple
3 watts per core isn't too big

Cheers

You could buy a standard 100:5 CT, and then drive a standard
transformer used as a step-up, then a bridge rectifier and shunt
regulator.



--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[Jon Elson]

On Thu, 08 Aug 2019 09:55:47 +0100, Clive Arthur wrote:

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

The power needed will have to end up as a 30V 300mA supply, so 10 watts
or so which is a lot for a normal CT.

Well, consider the cases where the motor is off -- no output.
And, when the motor starts, you get maybe 6X the current for several
seconds. Your circuit needs to handle this, as the output of the current
transformer will be 6X normal running output. Your current transformer
will be outputting 300 mA X 6 = 1.8 Amps!

Why not just power the device from an ordinary single-phase transformer
from a convenient outlet?

Jon

 

On Thursday, 8 August 2019 10:55:57 UTC+2, Clive Arthur wrote:

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

The power needed will have to end up as a 30V 300mA supply, so 10 watts
or so which is a lot for a normal CT.

This is not something I've looked at before and sounds quite difficult
- a quick Google shows that this can be done for very low power devices..

Any ideas? Multiple toroids with secondaries in parallel? Long sausage
toroid? The energy has to come from the wire, obviously, so the
impedance of the wire will have to increase a little which implies that
some length will be needed, but now I'm rambling.

Is the following even in the same county as the ballpark?..
30V @ 300mA is 100R
50A -> 300mA is a turns ratio of 1:167.
Inductance goes with the square of the turns, so the wire must see 3.6mR
3.6mR at 60Hz is 10uH
10uH/N^2 cores are doable, maybe multiple
3 watts per core isn't too big

Hoping I do not miss it, but:

One wire of the mains, with a solenoid winding would around it. You can get whatever current you like, in this case 300mA. But the voltage developed across the single wire will be extremely small, even if you have 167 ratio it will not be 30V. To have the power you are talking about, you need to break the loop

Cheers

Klaus

 

On Thursday, 8 August 2019 23:01:16 UTC+2, klaus.k...@gmail.com wrote:

On Thursday, 8 August 2019 10:55:57 UTC+2, Clive Arthur wrote:
A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

The power needed will have to end up as a 30V 300mA supply, so 10 watts
or so which is a lot for a normal CT.

This is not something I've looked at before and sounds quite difficult
- a quick Google shows that this can be done for very low power devices.

Any ideas? Multiple toroids with secondaries in parallel? Long sausage
toroid? The energy has to come from the wire, obviously, so the
impedance of the wire will have to increase a little which implies that
some length will be needed, but now I'm rambling.

Is the following even in the same county as the ballpark?..
30V @ 300mA is 100R
50A -> 300mA is a turns ratio of 1:167.
Inductance goes with the square of the turns, so the wire must see 3.6mR
3.6mR at 60Hz is 10uH
10uH/N^2 cores are doable, maybe multiple
3 watts per core isn't too big

Hoping I do not miss it, but:

One wire of the mains, with a solenoid winding would around it. You can get whatever current you like, in this case 300mA. But the voltage developed across the single wire will be extremely small, even if you have 167 ratio it will not be 30V. To have the power you are talking about, you need to break the loop

To develop a voltage across the magnetising inductance, you need a very big inductance, enough that you develop voltage a 50Hz. So the ferrite you need to use for a 30W transformer would be very very big

Cheers

Klaus

 

[Jon Elson]

On Thu, 08 Aug 2019 14:01:12 -0700, klaus.kragelund wrote:

One wire of the mains, with a solenoid winding would around it. You can
get whatever current you like, in this case 300mA. But the voltage
developed across the single wire will be extremely small, even if you
have 167 ratio it will not be 30V.

No, a current transformer is not a solenoid around the mains wire, it is
a toroid magnetic core around the mains wire, with a multi-turn winding
around the toroid.

You can design it so it will saturate during motor starting, so you don't
have to dissipate a great excess amount of energy during the start.

For a 50/60 Hz motor, one would not use a ferrite, but a laminated iron
core of some design.

Jon

 

On Thursday, 8 August 2019 19:39:42 UTC+1, Jon Elson wrote:

Why not just power the device from an ordinary single-phase transformer
from a convenient outlet?

Jon

clandestine power abstration? I'm guessing one could light a house & listen to the radio on 10w

I assume it's more a rules & regs thing, or letting it be installed by a non-sparky.


NT

 

On Thu, 08 Aug 2019 13:39:33 -0500, Jon Elson <elson@pico-systems.com>
wrote:

On Thu, 08 Aug 2019 09:55:47 +0100, Clive Arthur wrote:

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

The power needed will have to end up as a 30V 300mA supply, so 10 watts
or so which is a lot for a normal CT.


Well, consider the cases where the motor is off -- no output.
And, when the motor starts, you get maybe 6X the current for several
seconds. Your circuit needs to handle this, as th,e output of the current
transformer will be 6X normal running output. Your current transformer
will be outputting 300 mA X 6 = 1.8 Amps!

While a starting induction motor can take up to 6 times nominal
current, you are not allowed a start a 40-80 kW motor by directly
power it on. You need a soft starter or a classical wye/delta starter
(with a short break when going from we to delta).

Why not just power the device from an ordinary single-phase transformer
from a convenient outlet?

Jon

 

[whit3rd]

On Thursday, August 8, 2019 at 1:55:57 AM UTC-7, Clive Arthur wrote:

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid.

[intended for 10W of output power]

Why? You can (six diodes) rectify the three-phase, and filter it,
and make a switchmode isolated power supply, where does the
'electrical contact' become a problem?

A 'current transformer' gives very low voltages, into a short
circuit, it's a terrible choice of power-delivery component.
It's useful for METERING, where power transfer
(and impedance matching) is not required.

 

On Thu, 8 Aug 2019 21:15:03 -0700 (PDT), whit3rd <whit3rd@gmail.com>
wrote:

On Thursday, August 8, 2019 at 1:55:57 AM UTC-7, Clive Arthur wrote:
A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid.

[intended for 10W of output power]

Why? You can (six diodes) rectify the three-phase, and filter it,
and make a switchmode isolated power supply, where does the
'electrical contact' become a problem?

A 'current transformer' gives very low voltages, into a short
circuit, it's a terrible choice of power-delivery component.
It's useful for METERING, where power transfer
(and impedance matching) is not required.

There are current transformers from 1.5 VA to 60 VA and standard
secondary currents of 1 A and 5 A. Thus the 1.5 VA version could
deliver 1.5 V resp. 300 mV. The 60 VA models could deliver 60 V resp.
12 V.

The high power and hence high voltage is useful if the distance
between the current transformer to the actual measurement device is
large due to the voltage losses due to the wiring.

Anyway sturdy wires must be used to handle any short current peaks due
to short circuit in the high power line. Long wires are typically
encountered when th CT is outside and the instrument is indoors.

 

[Jasen Betts]

On 2019-08-08, Jan Panteltje <pNaOnStPeAlMtje@yahoo.com> wrote:

On a sunny day (Thu, 8 Aug 2019 09:55:47 +0100) it happened Clive Arthur
cliveta@nowaytoday.co.uk> wrote in <qigo2o$ge2$1@dont-email.me>:

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

OK, 50 A, to get 300 mA (AC) you need 50 / .3 = 167 turns secondary,
to get 30V @ 300 mA your load needs to be 100 Ohm.
What is the problem?

You will need a decent iron core current transformer.

If you use all 3 phases 1/3 per phase, 3 phase rectifier bridge makes things simpler
and transformers smaller?

Clamp on cores?

yeah get three current transformers and three bridge rectifiers
connect the three DC supplies made by that in series and feed
that to a high-power zener, or to some sort of current-fed switcher,
to get your target voltage. (or target current whichever one he really
needs)

--
When I tried casting out nines I made a hash of it.

 

[Clive Arthur]

On 08/08/2019 19:39, Jon Elson wrote:

On Thu, 08 Aug 2019 09:55:47 +0100, Clive Arthur wrote:

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

The power needed will have to end up as a 30V 300mA supply, so 10 watts
or so which is a lot for a normal CT.


Well, consider the cases where the motor is off -- no output.
And, when the motor starts, you get maybe 6X the current for several
seconds. Your circuit needs to handle this, as the output of the current
transformer will be 6X normal running output. Your current transformer
will be outputting 300 mA X 6 = 1.8 Amps!

Why not just power the device from an ordinary single-phase transformer
from a convenient outlet?

Jon

Power off doesn't matter, and the max current is limited to 100A.

Thanks for all the input, this is probably easier than I'd thought
albeit completely outside my experience.

[The nearest outlet isn't convenient, being possibly a few miles away at
the top of a very deep hole!]

Cheers
--
Clive

 

On Friday, 9 August 2019 00:34:10 UTC+2, Jon Elson wrote:

On Thu, 08 Aug 2019 14:01:12 -0700, klaus.kragelund wrote:


One wire of the mains, with a solenoid winding would around it. You can
get whatever current you like, in this case 300mA. But the voltage
developed across the single wire will be extremely small, even if you
have 167 ratio it will not be 30V.

No, a current transformer is not a solenoid around the mains wire, it is
a toroid magnetic core around the mains wire, with a multi-turn winding
around the toroid.

Toroid or solenoid, I do not care, it's a transformer

You can design it so it will saturate during motor starting, so you don't
have to dissipate a great excess amount of energy during the start.

For a 50/60 Hz motor, one would not use a ferrite, but a laminated iron
core of some design.

Please tell me how you will get 10W out, with out a VERY big transformer?

The voltage on the primary will be arbitrarily low....

Regards

Klaus

 

On Fri, 9 Aug 2019 09:08:06 +0100, Clive Arthur
<cliveta@nowaytoday.co.uk> wrote:

On 08/08/2019 19:39, Jon Elson wrote:
On Thu, 08 Aug 2019 09:55:47 +0100, Clive Arthur wrote:

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

The power needed will have to end up as a 30V 300mA supply, so 10 watts
or so which is a lot for a normal CT.


Well, consider the cases where the motor is off -- no output.
And, when the motor starts, you get maybe 6X the current for several
seconds. Your circuit needs to handle this, as the output of the current
transformer will be 6X normal running output. Your current transformer
will be outputting 300 mA X 6 = 1.8 Amps!

Why not just power the device from an ordinary single-phase transformer
from a convenient outlet?

Jon

Power off doesn't matter, and the max current is limited to 100A.

Please note that the short circuit current (and hence CT secondary
current) can be much larger, so hopefully your DC load survives this.

Thanks for all the input, this is probably easier than I'd thought
albeit completely outside my experience.

[The nearest outlet isn't convenient, being possibly a few miles away at
the top of a very deep hole!]

On Europe in mines often 400 V delta only (or 690 or even 1000 V)
cabling is used. To power low power devices and providing lighting
at each site a 400/230 V transformer primacy is connected between two
phases and the cold side of the secondary is grounded to a local
ground electrode. Control system voltages (e.g. 24 Vdc) are then
derived from the local 230 V network using standard components..


Apparently this is in the Americas (60 Hz). If the power feed is
several kilometers away, this has to be a medium voltage (MV) motor
(with 4 kV ? delta only). Thus a small transformer with 4 kV primary
and 120 V secondary would be needed. If this is an underwater site, I
can understand why galvanic connection to the MV lines are avoided.

But anyway, the cable outer shield of 3 phase cble must be broken, so
that only a single phase wire is running through the CT (or 3 phase
wires through 3 separate CTs). There is no point of running a complete
3 phase cable through a single CT, since the phase currents and hence
magnetic fields will cancel.

 

[Clive Arthur]

On 08/08/2019 10:08, Jan Panteltje wrote:

On a sunny day (Thu, 8 Aug 2019 09:55:47 +0100) it happened Clive Arthur
cliveta@nowaytoday.co.uk> wrote in <qigo2o$ge2$1@dont-email.me>:

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

OK, 50 A, to get 300 mA (AC) you need 50 / .3 = 167 turns secondary,
to get 30V @ 300 mA your load needs to be 100 Ohm.
What is the problem?

You will need a decent iron core current transformer.

If you use all 3 phases 1/3 per phase, 3 phase rectifier bridge makes things simpler
and transformers smaller?

Clamp on cores?

How about this (just thinking out loud)...

Wind a solenoid on a bundle of insulated iron wires, like a Rogowski
coil but with a 'laminated' core, and wrap this around the current
carrying conductor, maybe several turns.

One for each phase, of course.

Cheers
--
Clive

 

[Tim Williams]

"Jon Elson" <elson@pico-systems.com> wrote in message
news:wcidnfuJoN_FPtHAnZ2dnUU7-dednZ2d@giganews.com...

You can design it so it will saturate during motor starting, so you don't
have to dissipate a great excess amount of energy during the start.

Not if you're using a shunt regulator!

Tim

--
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Design
Website: https://www.seventransistorlabs.com/

 

[Jan Panteltje]

On a sunny day (Fri, 9 Aug 2019 12:24:32 +0100) it happened Clive Arthur
<cliveta@nowaytoday.co.uk> wrote in <qijl5i$mjb$1@dont-email.me>:

On 08/08/2019 10:08, Jan Panteltje wrote:
On a sunny day (Thu, 8 Aug 2019 09:55:47 +0100) it happened Clive Arthur
cliveta@nowaytoday.co.uk> wrote in <qigo2o$ge2$1@dont-email.me>:

A customer has a three-phase supply drawing from 50A to 100A at
nominally 60Hz. He wants to tap some power from this without any
electrical contact, so probably using say three current transformers
with each phase going through one toroid. Probably the wire is too thick
to make multiple turns.

OK, 50 A, to get 300 mA (AC) you need 50 / .3 = 167 turns secondary,
to get 30V @ 300 mA your load needs to be 100 Ohm.
What is the problem?

You will need a decent iron core current transformer.

If you use all 3 phases 1/3 per phase, 3 phase rectifier bridge makes things simpler
and transformers smaller?

Clamp on cores?

How about this (just thinking out loud)...

Wind a solenoid on a bundle of insulated iron wires, like a Rogowski
coil but with a 'laminated' core, and wrap this around the current
carrying conductor, maybe several turns.

Interesting, had to look that up
https://en.wikipedia.org/wiki/Rogowski_coil#/media/File:Rogowski_coil.png

One for each phase, of course.

Or perhaps UI cores, coil on U leg and shift across current carrying inductor,
then close with the I part.



Personally I would just cut a mains transformer ring core in 2 halves.
Has to be real iron, no ferrites.
Winding on one side.

?