# Controlling repulsion of a magnet with electromagnet and SCRs.

#### Tonyr1084

Joined Sep 24, 2015
5,059
@cmartinez Here's what picture I have in my mind the TS is attempting to do: The pendulum has swung away for whatever reason, doesn't matter. The pendulum is going to swing back the opposite way with equal force, minus any friction losses. Under such condition the pendulum would eventually come to rest. With a periodic pulse of magnetic (repulsive) force a small amount of energy can be put back into the swinging pendulum and thus keep it swinging. The timing of the pulse needs to be very close to the period of the swing of the pendulum. Probably doesn't need to be highly accurate as long as the repulsive force (RF) comes just before the pendulum begins to swing away. To wait till after the pendulum has started swinging away means the RF will have loose effectiveness and the pendulum may come to a stop. The timing of the RF may also help to fine tune the swing of the pendulum, sending it back away before it has exhausted it's approaching energy (swinging toward the electromagnet).

Well, that's what I envision for this project. If I'm wrong I'm sure someone will correct me.

Think of a baseball bat swinging on a rope. Every time the bat swings toward you - you hit it with a basket ball. Just enough force to keep it swinging. That's what I'm thinking is the project at hand. Is it practical? What purpose are we wanting to control the pendulum? Would be nice to have this answered as well. And then, yes, we can give good answers as opposed to arguing over what our opinions are regarding the intended purpose of this thread may be.

#### Bernard

Joined Aug 7, 2008
5,649
Might we drop every thing after opening statement " Controlling repulsion of a magnet with electromagnet and SCRs down to @TeeKay6, post # 17. I would also drop U2 & D1 as SW1 can
do the same. I designed, built & used identical circuit back in 1954 +-, to fire blasting caps when SCR first became available ( and before ). I needed a controllable switch to sync cap break to 1/10 sec timing line on oscillograph.

#### Bernard

Joined Aug 7, 2008
5,649
If this is just to swing a pendulum then we could go back to " The Cardboard Clock" , Wireless World 1982 as ref. in Pendulum timing conundrum, #13, Search Forums.

#### atferrari

Joined Jan 6, 2004
4,092
Might we drop every thing after opening statement " Controlling repulsion of a magnet with electromagnet and SCRs down to @TeeKay6, post # 17. I would also drop U2 & D1 as SW1 can
do the same. I designed, built & used identical circuit back in 1954 +-, to fire blasting caps when SCR first became available ( and before ). I needed a controllable switch to sync cap break to 1/10 sec timing line on oscillograph.
@TeeKay6 @Bernard @cmartinez

Please refer to my circuit above.

With C charged (no matter how) I start by firing SCR1.

With SCR1 conducting, x msec later I trigger SCR2.

Once conducting, SCR2 starts diverting current from C1, directly to ground.

Since SCR1 is now deprived of current, it enters in cut off and current stops flowing through L1.

Firing SCR2 at will, I can control the period the current flows through the inductor. Yes @Tonyr1084 , that is my interest; finely adjusting the time between trigger pulses.

With C1 fully discharged, SCR2 deprived of current also enters in cut off. This is the end of the process I started by triggering SCR1.

Can we follow from here? I still would appreciate the questions below answered. Gracias.

1 - The diode in parallel with the inductor, how should be it connected?

2 - A bigger magnet against the same coil will result in a more effective repulsion?

#### cmartinez

Joined Jan 17, 2007
7,083
@TeeKay6 @Bernard @cmartinez

View attachment 191369

Please refer to my circuit above.

With C charged (no matter how) I start by firing SCR1.

With SCR1 conducting, x msec later I trigger SCR2.

Once conducting, SCR2 starts diverting current from C1, directly to ground.

Since SCR1 is now deprived of current, it enters in cut off and current stops flowing through L1.

Firing SCR2 at will, I can control the period the current flows through the inductor. Yes @Tonyr1084 , that is my interest; finely adjusting the time between trigger pulses.

With C1 fully discharged, SCR2 deprived of current also enters in cut off. This is the end of the process I started by triggering SCR1.

Can we follow from here? I still would appreciate the questions below answered. Gracias.
Ok, Agustín ... I think I'm beginning to understand.

My answer to your first question remains the same. Install the diode in parallel with the coil, with its arrow (cathode) pointing towards SCR1. That is, SCR1's arrow and D1's arrow should be pointing against each other.

As to your second question, my guess is that once the coil equals the magnet's field it won't matter how much more current you pump into it, since it will have reached saturation relative to the magnet. I do not know what the necessary math used to describe said situation is, though.

#### Alec_t

Joined Sep 17, 2013
11,641
Re the second question, it is the magnetic field strength of the magnet, rather than its physical size, which is important. The force is proportional to the product of the magnetic field strengths of the magnet and coil.

#### TeeKay6

Joined Apr 20, 2019
572
@TeeKay6 @Bernard @cmartinez

View attachment 191369

Please refer to my circuit above.

With C charged (no matter how) I start by firing SCR1.

With SCR1 conducting, x msec later I trigger SCR2.

Once conducting, SCR2 starts diverting current from C1, directly to ground.

Since SCR1 is now deprived of current, it enters in cut off and current stops flowing through L1.

Firing SCR2 at will, I can control the period the current flows through the inductor. Yes @Tonyr1084 , that is my interest; finely adjusting the time between trigger pulses.

With C1 fully discharged, SCR2 deprived of current also enters in cut off. This is the end of the process I started by triggering SCR1.

Can we follow from here? I still would appreciate the questions below answered. Gracias.
Depending on the resonance characteristics of L1-C1, SCR1 could turn itself off prior to SCR2 turning on. SCR1 will be much easier to trigger if the inductor (and its flyback protection diode***) is moved to the anode side; circuit operation will be the same. With the revised connection the SCR1 trigger can occur between ground and the gate terminal. No matter what configuration you use, the SCR1 gate signal must remain active until the inductor current has risen above the holding current of SCR1. If a flyback diode is used, note that inductor current can flow for a time after SCR1 is off; eventually the inductor current will drop to zero as the inductor's stored energy is consumed by the resistance of the inductor and the diode. ***Connect diode cathode to + side of C1; diode anode to anode of SCR1 (and - side of inductor).

Last edited:

#### atferrari

Joined Jan 6, 2004
4,092
SCR1 will be much easier to trigger if the inductor (and its flyback protection diode***) is moved to the anode side; circuit operation will be the same. With the revised connection the SCR1 trigger can occur between ground and the gate terminal.
Simpler; I like that! Thanks @TeeKay6

#### Janis59

Joined Aug 21, 2017
1,224
Diode is needed in all situations. When pulse is having the di/dt maximum (opening and closing) then dU=L*di/dt. For example, 1 Henry coil, with average for igbt 10 000 V/microsecond makes 10 000 000 000 Volts spike. Okay, probably coil quality factor is not yet so high, but sooner some 100, but even then 300 Volts*100=30 000 Volts. Its killing everything around without of diode.

#### Bernard

Joined Aug 7, 2008
5,649
Some miscellaneous information. Coil , 7 parallel strands # 26 Cu wire, 1/2 ohm on 22 mm ID X 40 mm OD, 53 mm long, air core coil; magnet 1 oz, 16 mm L X 15 mm D, strong,? Cap 10,000 uF charged thru 120 ohms from 12 V supply. Magnet jumps 1 in. on discharge, Expotential discharge with no under shoot. Magnet passing over coil produces a single AC cycle, 50 mV rise, 30 mV fall.

#### atferrari

Joined Jan 6, 2004
4,092
Some miscellaneous information. Coil , 7 parallel strands # 26 Cu wire, 1/2 ohm on 22 mm ID X 40 mm OD, 53 mm long, air core coil; magnet 1 oz, 16 mm L X 15 mm D, strong,? Cap 10,000 uF charged thru 120 ohms from 12 V supply. Magnet jumps 1 in. on discharge, Expotential discharge with no under shoot. Magnet passing over coil produces a single AC cycle, 50 mV rise, 30 mV fall.
Thanks Bernard

#### redrok

Joined Aug 27, 2010
11
I'm not sure if you need the diode... although I would definitely place it as a precaution. Place it in parallel with the coil, in a direction opposite to ground. That is, with the diode's "arrow" pointing towards SCR1.
The snubber diode is commonly used.
However, the diode has the effect of prolonging the flow of current in the inductor.
If this is undesirable a suitable resistor in series with the diode could be used to more rapidly disipate the inductor's energy. Choose the highest resistor value that doesn't exceed the maximum withstanding voltage of your SCRs.
redrok

#### shortbus

Joined Sep 30, 2009
7,936

#### atferrari

Joined Jan 6, 2004
4,092
@ atferrari I don't know if you saw this link at ETO or not, but thought it may help with your project.
http://sound-au.com/clocks/free-pendulum.html It was posted by I think his name is "rfjenkins". While not exactly done like you're trying it has a similar outcome.
I know that site from long time. I intend to control timing.

Thanks shortbus anyway.

#### shortbus

Joined Sep 30, 2009
7,936
Just thought of your project when I saw it, and didn't know if it would help.