Switching on giant electromagnet

[Terry Pinnell]

During a recent visit to the Huygens Science Museum in Leiden, one of
the most impressive exhibits was what was described as the world's
'second largest electromagnet'. It weighed 14 tons and I reckon its
coil wire was about 1 cm in diameter. I got to thinking about just how
you'd switch on a beast like this - or, more crucially, switch it off.

I've subsequently found a few answers here
http://www.sciencenews.org/articles/20010623/timeline.asp

but would be interested to hear from anyone with first-hand experience
of working with stuff on this scale, particularly before the
widespread use of computer-controlled switching. I have images of
someone ducking behind heavy reinforcements and pulling a large lever
for the first time...

 

[Rheilly Phoull]

During a recent visit to the Huygens Science Museum in Leiden, one of
the most impressive exhibits was what was described as the world's
'second largest electromagnet'. It weighed 14 tons and I reckon its
coil wire was about 1 cm in diameter. I got to thinking about just how
you'd switch on a beast like this - or, more crucially, switch it off.

I've subsequently found a few answers here
http://www.sciencenews.org/articles/20010623/timeline.asp

but would be interested to hear from anyone with first-hand experience
of working with stuff on this scale, particularly before the
widespread use of computer-controlled switching. I have images of
someone ducking behind heavy reinforcements and pulling a large lever
for the first time...

Heh heh, it's the switching OFF that would bother me
Personally that would definately be the 'Assistants' job

 

[Dejan Durdenic]

During a recent visit to the Huygens Science Museum in Leiden, one of
the most impressive exhibits was what was described as the world's
'second largest electromagnet'. It weighed 14 tons and I reckon its
coil wire was about 1 cm in diameter. I got to thinking about just how
you'd switch on a beast like this - or, more crucially, switch it off.

I've subsequently found a few answers here
http://www.sciencenews.org/articles/20010623/timeline.asp

but would be interested to hear from anyone with first-hand experience
of working with stuff on this scale, particularly before the
widespread use of computer-controlled switching. I have images of
someone ducking behind heavy reinforcements and pulling a large lever
for the first time...

I had to design a current source supplying a huge electromagnet. When
switching
off, we ramped-down the current at the precalculated rate (keeping voltage
at the
safe level). Still, as the precaution, antiparallel diode was installed in
case power
goes off abruptly...

Dejan

 

[John Devereux]

During a recent visit to the Huygens Science Museum in Leiden, one of
the most impressive exhibits was what was described as the world's
'second largest electromagnet'. It weighed 14 tons and I reckon its
coil wire was about 1 cm in diameter. I got to thinking about just how
you'd switch on a beast like this - or, more crucially, switch it off.

You need a diode across the coil, to protect the transistor from the
switch-off spike. A 1N4148 would probably be OK

 

[Frithiof Andreas Jensen]

but would be interested to hear from anyone with first-hand experience
of working with stuff on this scale, particularly before the
widespread use of computer-controlled switching.

The crude way is to have a dump resistor across the coil - the resistance of
the coil will be small so the resistor does not cause to much of a power
loss. One could also store the energy in a capacitor bank.

 

[Hank]

You need a diode across the coil, to protect the transistor from the
switch-off spike. A 1N4148 would probably be OK

perhaps you meant 4,148 1N4148 diodes?

 

[Klaus Bahner]

During a recent visit to the Huygens Science Museum in Leiden, one of
the most impressive exhibits was what was described as the world's
'second largest electromagnet'. It weighed 14 tons and I reckon its
coil wire was about 1 cm in diameter. I got to thinking about just how
you'd switch on a beast like this - or, more crucially, switch it off.

I've subsequently found a few answers here
http://www.sciencenews.org/articles/20010623/timeline.asp

but would be interested to hear from anyone with first-hand experience
of working with stuff on this scale, particularly before the

In fact it is not so difficult as one might think. I work with several
large scale magnets, the biggest one rated at a maximum current of 600
A, which gives a maximum field of 2T. The power supply was designed
some 30 years ago, so this certainly qualifies for no computer control
/soft start technologies
The power supply is a "simple" linear one with 100+ water cooled 2N3773
as pass transistors. It's basically just a scaled up version of a
standard linear supply - nothing really fancy. The fanciest thing is
probably a servo controlled variac for the 3-phase main transformer to
reduce power dissipation. Turning it on just switches the main
transformer on - after that the control voltage is increased to the
desired value. The most primitive way of controlling this supply is just
turning a potentiometer which sets the control voltage. So this really
works just as any simple, say 0-30V, 1A linear power supply.
The switch off procedure consists of decreasing the output voltage to
0V, then shutting it off. However in case of a power cut or a
overtemperature failure, the supply is immediately switched off even at
high currents. In this case a couple of large rectifier diodes (Can't
remember the exact type right now - these large stud mounted ones)
protect the supply. Furthermore several varistors are added to cut off
transients.

This works amazingly well. Once in a while one of the 2N3773 is
destroyed during a hard switch off/on cycle, but it really happens
rather rarely - maybe in the order of one per year in alltogether 5
supplies of this kind. (Most annoying thing is to find the broken one,
if you have more than 100 to check, replacing it takes a few seconds ;-)

In fact the biggest challenge in this kind of supply is the water
cooling (clogging, leaking pipes) and to achieve the required stability.
The magnetic field has to be kept constant to an order of 10E-5 or
better, which means you have to stabilize the current in the range of a
few milliamps, although your current is in the order of several hundred
amps. Another challenge arises when you try to change the magnetic field
rapidly, in fact it's not the inductivity of such a beast which causes
so much trouble, but the eddy currents in the several tons of iron core.


Cheers,
Klaus

 

[Terry Pinnell]

In fact it is not so difficult as one might think. I work with several
large scale magnets, the biggest one rated at a maximum current of 600
A, which gives a maximum field of 2T. The power supply was designed
some 30 years ago, so this certainly qualifies for no computer control
/soft start technologies
The power supply is a "simple" linear one with 100+ water cooled 2N3773
as pass transistors. It's basically just a scaled up version of a
standard linear supply - nothing really fancy. The fanciest thing is
probably a servo controlled variac for the 3-phase main transformer to
reduce power dissipation. Turning it on just switches the main
transformer on - after that the control voltage is increased to the
desired value. The most primitive way of controlling this supply is just
turning a potentiometer which sets the control voltage. So this really
works just as any simple, say 0-30V, 1A linear power supply.
The switch off procedure consists of decreasing the output voltage to
0V, then shutting it off. However in case of a power cut or a
overtemperature failure, the supply is immediately switched off even at
high currents. In this case a couple of large rectifier diodes (Can't
remember the exact type right now - these large stud mounted ones)
protect the supply. Furthermore several varistors are added to cut off
transients.

This works amazingly well. Once in a while one of the 2N3773 is
destroyed during a hard switch off/on cycle, but it really happens
rather rarely - maybe in the order of one per year in alltogether 5
supplies of this kind. (Most annoying thing is to find the broken one,
if you have more than 100 to check, replacing it takes a few seconds ;-)

In fact the biggest challenge in this kind of supply is the water
cooling (clogging, leaking pipes) and to achieve the required stability.
The magnetic field has to be kept constant to an order of 10E-5 or
better, which means you have to stabilize the current in the range of a
few milliamps, although your current is in the order of several hundred
amps. Another challenge arises when you try to change the magnetic field
rapidly, in fact it's not the inductivity of such a beast which causes
so much trouble, but the eddy currents in the several tons of iron core.

Thanks for all the replies.

Klaus: I assume you empty your pockets of keys and coins when at work?
<g>.

 

[Klaus Bahner]

Klaus: I assume you empty your pockets of keys and coins when at work?
<g>.

Actually no - it's sometimes a bit dissapointing, but you don't see much of
the field on the outside. One of the reasons such magnets are so huge is
that the core has to concentrate the field in the gap. Due to the huge core
(think of it as a good "magnetic conductor") the stray field is rather low.
Not even a screwdriver sticks to the outside of the core - if you on the
other hand put it into the gap, there is really a strong field. However the
field is comparable to those rare earth magnets found in hard disk drives or
better loudspeakers. The difference lies just in the dimensions: In a
loudspeaker/hard disk drive magnet the field is maintened only over a small
gap, a few millimeters wide and over an area in the order of a few square
centimeters. In our biggest bending magnet, the gap is a couple of
centimeters and the gap area approaches probably one squaremeter ...

Klaus

 

[Kevin Aylward]

During a recent visit to the Huygens Science Museum in Leiden, one of
the most impressive exhibits was what was described as the world's
'second largest electromagnet'. It weighed 14 tons and I reckon its
coil wire was about 1 cm in diameter. I got to thinking about just how
you'd switch on a beast like this - or, more crucially, switch it off.

I've subsequently found a few answers here
http://www.sciencenews.org/articles/20010623/timeline.asp

but would be interested to hear from anyone with first-hand experience
of working with stuff on this scale, particularly before the
widespread use of computer-controlled switching. I have images of
someone ducking behind heavy reinforcements and pulling a large lever
for the first time...

Well, my 2 cents worth. I was at the SuperConducting supercolider,
magnet division, for a couple of years before the big axeman came.

I wasn't directly involved in the design of the developmental test
magnet's power supply, it was externally built, but I did do quite a bit
on the instrumentation systems.

The power supply was a monster. 6ft tall by about 4' by 4', or
thereabouts. 10,000 amps at around 2ppm stability. I think it was
something like 14, 3cm cables in || connecting the PS to the magnet.

The err.. interesting bit, was the quench protection, that is when the
superconducting magnet goes normal. The PS needs to be immediately shut
down when quenching is detected to avoid blowing everybody up. I did the
quench detector bit, but not the really interesting part, which was the
current extraction rack to dissipate the energy. Again, another monster.
Banks of SCR' hocky pucks (3KA continuous), large capacitors, inductor
(loop of cable) and ahmm a "resistor".

Kevin Aylward
[email protected]
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

 

[Terry Given]

Well, my 2 cents worth. I was at the SuperConducting supercolider,
magnet division, for a couple of years before the big axeman came.

I wasn't directly involved in the design of the developmental test
magnet's power supply, it was externally built, but I did do quite a bit
on the instrumentation systems.

The power supply was a monster. 6ft tall by about 4' by 4', or
thereabouts. 10,000 amps at around 2ppm stability. I think it was
something like 14, 3cm cables in || connecting the PS to the magnet.

The err.. interesting bit, was the quench protection, that is when the
superconducting magnet goes normal. The PS needs to be immediately shut
down when quenching is detected to avoid blowing everybody up. I did the
quench detector bit, but not the really interesting part, which was the
current extraction rack to dissipate the energy. Again, another monster.
Banks of SCR' hocky pucks (3KA continuous), large capacitors, inductor
(loop of cable) and ahmm a "resistor".

Kevin Aylward

I've worked with some fairly big dynamic braking resistors - a 250kW DB
resistor is a sight to behold. Spiral-wound thin metal strip about 2cm wide,
and lots of it. We used to use puny little 11kW (pulsed) 50 ohm DB resistors
(about 10mm diameter, 150mm long in a stainless steel tube) with multimeter
probe leads as bus discharge resistors to suck out cap charge post testing.
It got exciting when you forgot to turn the drive off before connecting the
discharge resistor across the (oops, still live) 700Vdc bus (right beside
the 2MVA transformer). Detaching the leads was a no-no as an arc would form,
and would invariably short the DC bus. The standard technique was to scream
loudly, in the hope an E-stop would be triggered by other workers.

cheers
Terry