When is a flyback diode necessary?

[Trisorion]

I am making an electromagnet that will have a pack of AA batteries powering it. I expect it provide 30-40 amps through magnet wire that will be wrapped around a horseshoe shaped core. I will put a switch in the circuit so that I can turn it on and off. I am not very skilled with electronics but I thought this would be pretty straightforward. I thought that until I read the Wikipedia article on flyback diodes (**broken link removed**

the voltage is therefore so high that the air conducts...the near-instantaneous dissipation which occurs without a flyback diode is often observed as an arc across the opening mechanical contacts

Will I create a high voltage arc every time I try to turn the thing off? If so, how do I avoid this (I don’t think they make 40 amp flyback diodes)?

 

[MrAl]

You can find 40 amp diodes if you like, but with that shape core the
energy will probably die down in a short time period so you may get
by with say a 10 amp diode.
They do make diodes 100 amps and more.

 

[Leftyretro]

In fly back protection the diode does not have to be rated for continuous current. The fly back pulse is just microseconds long and most diodes have a surge current rating far in excess of there continuous ratings. However the reverse voltage rating should be looked at closely as the fly back pulse can reach 4X the coil voltage rating. You didn't state how much voltage your AAA pack will be running so I can't give you a for sure suggestion but the common 1N4007 is rated at 1000V PIV.

Lefty

 

[Trisorion]

1.2 volts, it is all in parallel. (It is part of the rules of the contest I am doing)


3 more questions:

1) What type of diode would be best suited to be a flyback diode in this application? (meaning, what should I look for on digikey)

2) Is there any downside to a diode? (Would it limit the strength of the magnet?)

3) Could I use a double pole double throw switch in place of a diode? ON position=connect the battery to each end of the coil Off position= connect both ends of the coil together. If it only takes a fraction of second to pulse a switch might be too slow…

 

[crutschow]

The diode should be connected in parallel with the coil with the cathode connected to the positive coil connection and the anode to the negative connection, thus the diode is reverse biased when applying voltage to the coil. It provides a path for the inductive current when you open the switch. This current direction causes the diode to be forward biased, so the voltage spike will be no more than the forward drop of the diode when carrying the coil current (a volt or so). It thus has to have a voltage rating only slightly larger than the voltage you are applying to the coil.

1) Any diode with a ≥10V and ≥10A rating should work.

2) All the diode will do is delay the collapse of the magnetic field for a short period after you remove the voltage. It will have no effect on the magnetic field otherwise. You can speed up the collapse, if needed, by adding a small resistor in series with the diode. This will increase the spike voltage by an amount equal to the coil current times the resistor value.

3) A relay will only transfer the arcing from the switch to the relay contacts. You still need the diode across the coil.

 

[MrAl]

1.2 volts, it is all in parallel. (It is part of the rules of the contest I am doing)


3 more questions:

1) What type of diode would be best suited to be a flyback diode in this application? (meaning, what should I look for on digikey)

2) Is there any downside to a diode? (Would it limit the strength of the magnet?)

3) Could I use a double pole double throw switch in place of a diode? ON position=connect the battery to each end of the coil Off position= connect both ends of the coil together. If it only takes a fraction of second to pulse a switch might be too slow…

Hi again,

The time the diode conducts is dependent on a few things:

dt=di*L/v

where

dt is the time it is conducting
di is the change in current (your max current, to be safe)
L is the inductance of your coil
v is the voltage of your batteries

The power dissipation of the diode is also based on how often the
coil is turned on and off. If you turn it on once and turn it off and
leave it for an hour without turning it back on again then we can
probably assume that the diode cools down enough to take another
shot. If not, the diode could eventually overheat.

Because of this and because we dont know the inductance of your
coil i think it would be best if you purchased a rather large diode,
just to make sure it doesnt blow out after one or a few uses.

I would get at least a 10 amp rating, and because the price doesnt
always change too much for the voltage rating i would look for a
50v rated unit, unless of course you can find a 20v rated unit much
cheaper. A standard rectifier type should be ok unless you plan
to switch this thing on and off very fast. If you have trouble
finding one i'll check it out for you, no problem.

The diode should also be mounted close to the coil rather than
far away.

The diode will have no negative effects on the working of the coil.
It's strength will be the same with or with out the diode. The
diode will help the coil in that it will limit the reverse voltage to
a safe level so the coils wire insulation doesnt break down.
This means it keeps the coil working longer.

You should not use a double pole switch because it can not switch
fast enough and it bounces too, and if you use a make before
break contact it will short out the battery bank. If you use a
break before make contact it will allow a high voltage spike
to occur for some time, and this is also no good.
The diode has a very fast on switching time so it clamps
the spike before it can reach a damaging level.

 

[Willbe]

If don't mind short switch lifetime you don't need to suppress the -L(dI/dT) spike.

Doing locked rotor tests on a large motor we'd get a grapefruit-sized fireball, in the air, above the switch.

 

[300zxev]

Hi All

Can you please apply your knowledge to another scenario ...

I am looking to assemble a 144V 1000Ah battery pack to power a 196V DC Electric motor ...
I do have a speed controller in this circuit that will supply 300A to the motor.
But for brief instances I would like to bypass the controller and supply 500A direct to the motor ...

It has been suggested I include 2 flyback diodes ... one between the motor and controller to prevent the controller blowing ... and one between the motor and battery pack ...

The bypass will be activated by a 600A contactor.


What I was after was suggestions on what diode to use and where to place it ...

I notice below it was suggested to place the diode in parallel with the coil ... but wouldn't it be placed in series with my motor (coil) ?

Thanks in advance

 

[300zxev]

any ideas ??? suggestions ???

 

[crutschow]

I notice below it was suggested to place the diode in parallel with the coil ... but wouldn't it be placed in series with my motor (coil) ?

The inductive spike current is in the same direction as the normal current flow, thus the spike voltage will be in the same direction as any series diode. (Think of the inductor as a battery with the positive voltage at the current output terminal of the inductor). The series diode will thus have no effect on the spike. The diode must be placed in parallel across the inductance to suppress the spike.

 

[300zxev]

Will I be experiencing this inductive spike then in my scenario ...
(144v 500A power to the motor suddenly turned off)
It was damage to a paralleled 300A controller that I wanted to avoid.
I was told to install "flyback diodes" ... but i was fishing for info on which ones, where and why ?

 

[dknguyen]

Will I be experiencing this inductive spike then in my scenario ...
(144v 500A power to the motor suddenly turned off)
It was damage to a paralleled 300A controller that I wanted to avoid.
I was told to install "flyback diodes" ... but i was fishing for info on which ones, where and why ?

Of course.

With a bi-directional motor controller, blindly using flyback diodes no longer work because current can flow through the main inductance in the circuit (the motor) in either direction. Ideally they would be in anti-parallel with the inductance but because of current can flow in both directions through a bidirectional motor, this does not work since it could cause a short-circuit across the motor in one of the directions.

THat's why they have flyback diodes in anti-parallel with each power transistor in the controller (and possibly snubbers here and there) to deal with it. These provide a a circular path for current generated by the inductive spike to flow from one end of the inductor to the other end to try and supress the spike while not producing a short-circuit in normal operation.

Let's review: for a unidirectional inductive load, you put a diode in anti-parallel with the inductance to give the current produced by the voltage spike to flow. For a bidirectional inductive load you put diodes in anti-parallel with each power transistor. Don't confuse the two. If you mix them up, it will either offer no protection (using the bidrectional method for a uni-directional circuit) or give great protection in one direction, but cause a short-circuit in the other direction (using the uni-directional method for bi-directional).

These both work to protect the switch by giving the flyback current a path to flow to supress the voltage spike. However, the bidrectional case is not as direct, nor as effective in supressing the spike as the unidirectional method is. Examine how they both work carefully, and you will see that the unidirectional method will clamp the voltage spike to the forward voltage drop of the diode while the bidirectional method will clamp the voltage spike to the Supply Voltage + diode forward voltage drop. So not only does it let the voltage spike go much higher, it even requires the spike to exceed the supply voltage a little bit.

H-bridge secrets part 2

Any motor controller that is worth anything at all should be able to survive the voltage spikes it produces.

 

[300zxev]

Hi

Thanks for some more info ...
The DC motor control is only single direction ... reverse direction is done mechanically ...
The controller might be worth its weight ... but i know it's worth it's weight in gold so I don't want to let the smoke out ...
Flyback diode ... still no with single direction motor ?

 

[dknguyen]

What do you mean direction is reversed mechanically? You are going to have to describe this motor driver more.

Basically, if current can flow through the motor in two directions, do not place a diode in parallel with it or it will protect in one direction but produce a short-circuit in the other.

If your motor controller is only offers unidirectonal speed control, but you are using relays (or something else) to reverse the current flow through the motor to reverse direction then your motor controller is nothing more than one big transistor with some fancy control logic. I think you should be able to place two flyback diodes across two pairs of primary contacts on the relay. So that when the relay is turned on one way, one of the diodes is anti-parallel connected with the motor while the other id disconnected (and vice versa for the other direction). It's basically using the uni-directional method twice on one motor to try and protect it in both directions, but to prevent the diodes from producing a short-circuit, the relay is used to disconnect the diode that would produce a short-circuit. However, this does not protect the relay contacts from arching and the only way to protect against that is a non-polarized snubber (protects in both direction at once but won't cause a short-circuit in either direction) in parallel with the motor.

Just give us a schematic here. Describing how to connect things to a circuit which I do not know is hard for both me and you.

 

[300zxev]

This DC motor / Battery setup is for an electric vehicle.
I'm sure the motor is capable of bidirectional, but in most installations in a vehicle it is setup as unidirectional as it is coupled to a vehicle gearbox which has a reverse gear.

The controller does only offer unidirectional speed control ... The way you have described the controller sounds quite basic, but they are expensive even for a 300A capable unit.

I don't have any schematics on hand at present (at work) but I do have a basic wiring diagram to give you a picture of what this setup looks like.

You can ignore the "reversing contactor" because that is for gearbox free installs, but i will be retaining the gearbox.

My intention is to add an extra contactor (higher amp than pictured) to bypass the 300A controller and potentially send 500A direct to the DC Motor until it is turned off and power rerouted back through the controller.

But i wouldn't want any induced load from the motor killing the 300A controller.

 

[mvs sarma]

............
Will I create a high voltage arc every time I try to turn the thing off? If so, how do I avoid this (I don’t think they make 40 amp flyback diodes)?

There are transient suppressors in market, like p6KV..

 

[dknguyen]

Oh! I thought when you mean "mechanical" you mean electro-mechanical as in relays. Not actually mechanical-mechanical. So the current only ever goes through the motor in one direction. So it would seem that you can just put a very fast diode in anti-parallel with the motor. Low forward voltage does not matter as much here (as in the bidirectional clamping because it does not clamp the voltage to Vsupply+Vdiode) so long as the diode can handle the current pulse...

But...and this is a big one you must make sure the motor controller isn't ever applying reverse voltage to the motor. Some motor controllers during the PWM "off" time, the PWM isn't realy off. It actually applies reverse voltage to the motor to collapse the drain energy from the inductance faster to stress the system less. As you probably guessed, when you have a diode producing a short-circuit across the motor that same reverse direction...it's a big no no.

Or, like MVS Sarma suggested, you can just use a BIDIRECTIONAL transient supressor. It's like a two diodes in series facing opposite directions to block current in both directions. THey are pulse rated and very very fast. WHen the voltage gets too high the reverse-biased diode "breaks down" at a specified voltage conducting current through itself in the reverse direction clamping the voltage to it's breakdown voltage. There are other non-diode things. THat's just one kind but they all kind of work like that.

Is this a brushed motor? IN an electric car? Kind of strange...Why such a low voltage and such high current? It'd be more efficient with a higher voltage and lower current. I know of no 12V battery than can provide 500A without melting...then again I don't know how big lead-acids can get either.

 

[300zxev]

I'll drop a line to the controller manufacturer today and ask about the reverse voltage.

Had a quick look at these transient suppressors ... haven't found the p6kv yet but does this do the same thing ?

**broken link removed**

Is the 80ohm resistance going to add much less efficiency to the circuit ?
It appears it can be set to a voltage ... but will this affect lower voltages (if that is what the controller is supplying)
And also is it only reverse voltage that I should be protecting the controller from ? What about current ?

This is a brushed DC motor ... specs are here if you want to see
**broken link removed**

I believe if the voltage is pushed much higher in these things it causes arcing accross the brushes ... The higher voltage EVs normally run AC motors ...
I'll be usung an array of 120 - 144 12V VRLA Batteries to produce the 500A

 

[300zxev]

Here is a link to the controller manufacturer
**broken link removed**

He is located in Toronto

 

[paul_ma7]

flyback diode question

THat's why they have flyback diodes in anti-parallel with each power transistor in the controller (and possibly snubbers here and there) to deal with it. These provide a a circular path for current generated by the inductive spike to flow from one end of the inductor to the other end to try and supress the spike while not producing a short-circuit in normal operation.

Let's review: for a unidirectional inductive load, you put a diode in anti-parallel with the inductance to give the current produced by the voltage spike to flow. For a bidirectional inductive load you put diodes in anti-parallel with each power transistor. Don't confuse the two. If you mix them up, it will either offer no protection (using the bidrectional method for a uni-directional circuit) or give great protection in one direction, but cause a short-circuit in the other direction (using the uni-directional method for bi-directional).

These both work to protect the switch by giving the flyback current a path to flow to supress the voltage spike. However, the bidrectional case is not as direct, nor as effective in supressing the spike as the unidirectional method is. Examine how they both work carefully, and you will see that the unidirectional method will clamp the voltage spike to the forward voltage drop of the diode while the bidirectional method will clamp the voltage spike to the Supply Voltage + diode forward voltage drop. So not only does it let the voltage spike go much higher, it even requires the spike to exceed the supply voltage a little bit.

H-bridge secrets part 2

Any motor controller that is worth anything at all should be able to survive the voltage spikes it produces.[/QUOTE]


----------------



if i have inductors in an ac circuit, should i install two flyback diodes for every inductor,
to protect the other components from the kickback?