thanks for your input !

does it matter how i wire the diode into circuit , in other words should + of the supply go to + of the diode...am going to use a dpdt relay and switch both + and - from supply to the coil ...is it better to put diode closer to relay or to coil ..or does it matter

bob drafts

If the diode is wired forward-biased, either the overcurrent protection device will operate, disconnecting the supply, or something else will melt/smoke/catch fire. The device may or may not survive the experience.

If it is reverse-biased, it will suppress the arc at the switch contacts in the manner desired, as already stated. Any arc at the switch is caused by the collapse of the magnetic field within the relay when the supply to it is switched off at that switch.

I am not up to date on your process. Could give more details. Why is it necessary to de-energize the motor? What would happen if it were left running?

Please download and study this excellent paper. i am sure it has what you need.

http://relays.te.com/schrack/pdf/c0_v4bg_4.pdf

i see that your question is about arc suppression.

where is the arc?

or is the question really about overheating / failure of your motor?

or is the question really about an arc in the switching device for the clutch?

1. when your machine is switched ON and OFF every three seconds - is it 3 seconds ON and 3 seconds OFF? or is it an OFF after every 3 seconds of ON period?

2. what is over heating? the switchgear? the drive circuitry? the motor?

3. what is the type of load on the motor? load cycle? inertia?

4. what is the rated speed of the motor?

5. in how much time does the motor accelerate to full speed?

please tell us more so we can tell you correctly.

don't over think the problem. use a solenoid that's suitable for the application.you may be able to find one at a junkyard

Quite correct that the exponential decay of the solenoid current increases the drop off time. If this delay of a 100ms or more is not acceptable, an Metal Oxide Varistor would be the answer. Surge voltage allowed may be more than a freewheel diode, but then increase in timing is likely to be no more than 2 or 3 ms.

The paper you referenced in the earlier post has, if I remember correctly, a very good discussion about this. Their suggested approach to keeping the time as short as possible while keeping the reverse voltage spike as low as possible was a zener diode in series with the "freewheeling" diode and to chose a zener whose voltage was approximately the same as for the supply circuit.

--JMM

What a memory ! You are quite right. It is on page 35, i am trying to put a screenshot of it here, may not be readable....