Development of a contactor, prototype manufacturing

[wingsofdesire]

I am designing a contactor for both DC and AC mainly for polarity independent circuit-breaker. It has a contactor unit, actuating solenoid, arc chute, Electromagnetic arc blow out coil Auxiliary contacts and base plate. Basically i got stuck in the Time constant measurement, ie the Bracking and Closing time. What does parallel connected diode to coil signify?
The time constant measurement without protective device, with parallel connected varistor to coil, with parallel connected diode to coil all are coming the same.

Attaching the drawing of the same for better understanding.

 

[jim hardy]

I cannot make out any detail on that drawing.

The question appears to be: "what is the purpose of connecting a diode across a coil? "
Is that right?

There are two reasons:
1. Suppress the "inductive kick" by allowing the coil current to decay to zero through the diode, dissipating the stored energy as heat in the coil windings and diode instead of arcing at the contact that interrupts the coil circuit.
2. Similar to 1, but to introduce a controlled delay between time of interrupting coil circuit and magnetic flux decaying enough to release the armature. Often one uses a zener(avalanche) diode or resistance in series with an ordinary diode to control decay rate.

That's the principle.
As I said the drawing is illegible.
Were i not seeing a difference in time constant when paralleling a coil with various energy absorbers, i would be suspect first of my diode polarity and second of my measurement technique.

 

[Windadct]

The diode is to handle the inductive kick when the contactor is de-energized.

However the time constant for both is based on mechanical and electrical properties. For the closing - the closing solenoid will exert a force ( some inductive rise time as the current starts at 0A and T-0 when power is applied to the solenoid) - and then the mass of all of the components - need to accelerate etc... not a simple issue.

For breaking time - there are really 2 critical times - normal braking time is from the trip ( for a contactor, this is removal of the power to the solenoid) to the point when the contacts break ( separate) - however the critical issue is how fast to they separate, and how long (time) will it take to extinguish some amount of current, in DC this is difficult to accomplish. So most device's published extinguish times are usually based on worst case - inductive loads - it is a balancing act of blowout coil strength, and opening spring speed ( note the opening spring is acted against by the closing solenoid - so to make the contacts open faster, bigger spring, bigger solenoid) also - this tends to increase the dI/dT across the contacts - making the apparent voltage across them increase - it is not a simple project.
If this is to operate frequently - the arcing surfaces should probably be tungsten and easily serviceable ( by filing) and replaceable.

 

[wingsofdesire]

Thank you for those informative inputs. I've been lately out of town, and back again to resume.
Well, i have placed a silver tip in the contacts.
The contactor is working fine. All the high voltage and insulation tests are upto mark except for the Closing and Interrupting Time.
I have to increase the closing time atleast by 20-30ms.
The interrupting time as specified is about 8-12ms without protective device.
Yet my test results give me somewhere about 53-55ms, where as it shout be 80-100ms.
And my braking time is just about 60ms.
Has the magnetization property of the coil got anything to do with it. I was thinking of increasing the gauge of the wire or the number of turns. I haven't yet checked on the inductance of coil. Though the resistance is around 460 ohms at 20C. I am just stalled with this and cannot find a way out.

Thank you.

 

[Carl Pugh]

To decrease the closing time, you have to have less resistance to closing (Weaker springs or lower mass) or a stronger solenoid. (For a stronger solenoid you need a larger solenoid or higher ampturns on closing or maybe a different designed solenoid or maybe a DC solenoid).

To decrease interrupting time you need faster moving contacts or a better arc chute.

Gigavac makes some vacuum and high pressure gas filled relays. You might take a look at their website to see if you can get some ideas.

I designed a DC relay for Ross Engineering Corporation Campbell California using a unique approach instead of an arc chute. The relay worked great and we talked about getting a patent on it. A patent was never applied for and one relay was sold, so the idea is not patentable.
If you are interested in a different approach instead of using an arc chute, let me know.

 

[wingsofdesire]

Actually I have to work on some set parameters, in which the arc chute is the basic design criteria. Rated voltage being 3000v and max operating voltage 4500v.
Switching frequency is 2 switchings per second.
The main reason for the arc chute is that the arc cools down and produces necessary over voltage.

 

[Carl Pugh]

3000 volt DC is very high for a contactor using contacts that open in air.
What current are you operating at?
Have you considered a vacuum contactor?

 

[wingsofdesire]

The tests are basically conducted on no load basis. The rated current being 50amps. And closed current limit is 250 amps.
Did consider vacuum type. But again, the basic design criteria has to be followed.
Is the braking and opening dependent on the spring strength or mass?

 

[wingsofdesire]

The tests are basically conducted on no load basis. The rated current being 50amps. And closed current limit is 250 amps.
Did consider vacuum type. But again, the basic design criteria has to be followed.
Is the braking and opening only dependent on the spring strength or mass? Does alteration of any other factor help it?
Does the interrupting time also depend on it? What should be done to alter the interrupting time?

 

[Carl Pugh]

Am I reading this correctly, worst case the contactor has to break 250 amp DC at 4500 volt DC, in air at sea level?

 

[jim hardy]

Seeems you are having trouble with opening time.

A few questions i'd been hoping you'd get into:

How does this thing work? Particularly , what is the function of the coil?

Does the coil hold the breaker closed, and it opens the contacts by a spring when electromagnet releases it?
Or--- Does the coil pull on some kind of trip lever, like a mousetrap, initiating mechanical actions by other parts to open the contacts?

In other words is this trip sequence begun by energizing or by de-energizing the coil?
Is the coil powered by DC or by AC?

....

Are you testing a brand new prototype or are you trying to reverse-design some obsolete gizmo for the replacement market?
In other words are we troubleshooting a design or troubleshooting an old piece of equipment?

Reason I ask these things is this -
An old DC coil with a few shorted turns will still actuate a solenoid.
But it will be slow to respond because the shorted turns oppose the changing flux whenever it is energized or de-energized. Just a few shorted turns don't change its DC resistance very much so you likely won't notice them with an ohm-meter, but an inductance measurement reveals them with good sensitivity.
I had coils the size of a basketball. Driving a nail into the side of one gave a 20% drop in inductance as soon as the nail hit copper, but less than 1% resistance change.
The coil lost a surprising amount of its inductive "kick".

Since you seem puzzled go back to the very basics...
You said early on there's something peculiar about this coil's insensitivity to parallel devices intended to absorb its inductive 'kick', ,,, so maybe there's something wrong with it.
Does it have a shaded pole? That's another timing trick magnetic designers use.

One solves a lot of mysteries by visual inspection with his mind's eye open to very basic physics.

old jim

 

[wingsofdesire]

Mr. Hardy, Carl thanks for those inputs.


I will try to break it down and be as descriptive as possible. We are actually trying to reverse design and upgrade.


Now about the basic construction, It has Contact Unit, Actuating Solenoid, Arc Chute, Electro magnetic arc blow out coil auxiliary contacts and a base plate.


The Contact unit consists of a fixed contact holder and a moving contact lever. Both parts are fitted with silver tips. There are spring mounted to avoid contact bounce.

The movement of the contact is a gliding movement. Actuation of the moving contact is done using te solenoid plunger and actuating lever. It is opened by armature resetting spring. The actuation of the moving contact is achieved electromagnetically and it is opened by resetting the spring after holding current has been switched off.

It is just the energizing and de energizing mechanism. The blow out coil ensures rapid and reliable extinction of arc. The auxiliary contacts are basically snap switches NO or NC. Voltage for drive 110 VDC.


Now my problem is with the opening and closing time of the fixed and mobile contact when energized or energized.

I am unable to achieve the closing time of 100ms. and interrupting time of 8-12ms. without protective device.

I have to increase the closing time atleast by 20-30ms, because am getting somewhere around 53-55ms, where as it shout be 80-100ms.

And my braking time is just about 60ms, whereas the interrupting time mentioned is about 8-12 ms.

 

[jim hardy]

So do I understand correctly:

Solenoid's magnetic pull draws contacts into closed position and "cocks" the opening spring;
and that happens faster than you expect ?

Spring pushes contact open when released by solenoid upon cessation of current (through solenoid)
and that happens slower than you expect ?

Again, back to basics.

Closing time is affected by airgap and spring tension. To what value must flux increase to let magnetic pull overcome spring tension (+ weight of any parts) ? Flux is proportional to current/airgap .

Opening time similarly affected by airgap and spring tension - to what value must flux decrease to let spring overcome magnetic pull?

Flux is proportional to current/airgap .
Old relays and voltage regulator coils had stops at both ends of armature travel and they had spring tension adjustments. This was so that pull-in and release currents could be individually set. This allowed for natural timing in sequential circuits, and precise operating points.

I have no idea whether the old contactor you are redesigning used such 'tricks of the trade' .
Study it carefully.

old jim

 

[wingsofdesire]

Exactly right, sir.

The closing time is faster than required and the opening time is slower than expected.

The coil that we are using is about 450-470 ohms.

We will start with tial and error procedure by increasing the tension of the spring first. The air gap between the contacts also need to be modified.

Sir, can you please enlighten me about the voltage regulator coil stops that you have mentioned.

Yes sir, the old contactor also was based on the same principle but there was no time constant requirement in that. We just had to make sure that the contacts closed when the current was applied.

So as far as I can get, we need to fiddle around with the air gap and the spring tension to get the required time constants.

 

[jim hardy]

Sir, can you please enlighten me about the voltage regulator coil stops that you have mentioned.

here's a couple of thumbnail descriptions that demonstrate the principle.

http://avstop.com/ac/apgeneral/regulation.htmlhttp://automotiveenginemechanics.tpub.com/TM-9-8000/TM-9-80000303.htm

So as far as I can get, we need to fiddle around with the air gap and the spring tension to get the required time constants.

Yes, that is where I would start. But i'd try to calculate the target operating and release currents from magnetic pull vs spring force, and the inductance versus airgap, so i'd know what are the L/R time constants I'm working with.

Should be an interesting physics problem. In my day i'd write a Qbasic program that printed out results over a range of input variables so as to optimize mechanical adjustability, today's youth use spreadsheets instead.

Have fun, guys !

Also there's that old unanswered question from first posts about the observed insensitivity to parallel suppression device - do you know there's no shorted turns on your sample solenoid? How does its measured inductance compare to calculated?

old jim