How to eliminate the DC motor back emf?

Perhaps you're confusing the inductive voltage spike with back-EMF? Back EMF occurs because a DC motor acts like a generator, so when it's turning it generates a voltage that reduces the drop across the windings and hence the motor current. Or perhaps you're talking about the radio frequency energy generated by the commutator and brushes?

Inductive kickback occurs with any coil, whether it's within a mile of a DC motor or inside.

Do you mean a single capacitor? A large cap in parallel with the motor terminals will reduce the inductive spike (_not_ the back EMF) when you turn off the motor, but would consume a great deal of current on startup. Smaller caps, however, suppress RF quite nicely.

When I see the motor signal in CRO, they show out this tpe of waveform:

emf?

that looks like the effects of both the brushes and maybe flyback effects of non filtered rectifier between the zero crossing periods. you may want to try a small cap coupled across the + & - terminals.

In article , Electronic Swear wrote: [...]

That spike is due to the inductance of the motor windings. If this is a moderately large motor, there can be a fairly large amount of energy stored in the inductance. All this stored energy must be dealt with.

I assume your circuit looks like this:

V------------- ! [Motor] ! +------------- Scope ! !- ! !- ! GND--------------

In the DC case, the easiest answer is to put a diode across the motor so that the voltage on it can't reverse. In the AC case, you are going to need a capacitor with prehaps a resistor in series.

If you just use a capacitor, the energy that ends up in the capacitor is what was in the inductor so:

E = 1/2 * I^2 * L

emf?

I have already added a reverse biasing diode at the terminals of the motor. The motor is a very high power DC motor, so it has high inductance. I am using a relay to switch on motor. And for the input signal of the motor, I use a triac to chop in different angle to control the speed. Because the chopped signal is in AC, then i rectify it by a bridge and use a filter capacitor to smooth it.

I just use a small capacitor (0.015uF 1000Vdc) on ripple filtering, I don't know the capacitor is small or not. But I want to know if I increase the filtering capacitance of the cap. will have a better performance or not.

Or using a R-C filter before the motor input terminial to suppress the large surge at start-up. I am afraid the R must in a very small value with high power range in order to not reduce the performance of the motor.

Any suggestion?

i can only assume you must be using a DC PM or shunt more ? or maybe even a compound/universal type ? any ways. unless your reversing the direction, you should be able to use a large cap. the best thing to do is use both large electro caps for filtering and low ESR & ESL non polorized cap coupled together. also. using a reactor is a plus.

In article , Electronic Swear wrote: [....]

I think you can do an LC filter

From rectified SCRs ---+---)))))---------+---- Motor ! ! --- --- ^ --- ! ! ----+-----------------+----

At low SCR conduction angles the current from the SCRs is a bunch of think spikes. The motor isn't really intended to have those high frequencies on it. The inductance needs to be just enough to keep the SCRs safe from their di/dt rating and their max current with whatever capacitor you use.

The capacitor doesn't have to be all that huge to make a fairly major impact. You don't need to spread a pulse out that much to really lower its high frequency content.