fast switch-off for transistor switch

[vajirkar]

Hi guys,
I have a simple transistor switch driving a relay. See the first figure (without diode D1). I have two such transistors driving two relays. The relays switch a 110V load. Ideally only one relay should be switched on at a time, so only one transistor should be switched on. The driving circuit seems to be doing this correctly, however, when switching from one transistor to the other (i.e. switching from one relay to other), the transistor being switched off is probably not switching off fast enough. So there is a small window in time when both relays could be switched on. I'm thinking about using a diode D1 at the base of the transistor as in figure (with diode D1) to speed up the switching off process.
I haven't tried it (hopefully this weekend), but I wanted to know if this could lead to some other problem that may not be obvious to me.

View attachment 35669

 

[GPG1]

Most of the lag will be caused by the diode across the relay. Have you got a spare contact on each relay

 

[audioguru2]

Hi Vajirkar,
Welcome to our forum.
Never mind trying to speedup the transistor. It is already over 1000 times faster than a mechanical relay.
The protection diode across the relay coil is very important, even though it slows the release of the relay. Without having this diode, the collector of the transistor gets a spike of hundreds of volts when the magnetic field in the relay coil collapses.
If your driving impedance is low enough (to discharge the capacitor through your input diode), try delaying the turn-on of each transistor with an RC network. Another diode must then be added to the transistor's emitter.

View attachment 35670

 

[vajirkar]

GPG, I don't have any spare contacts on the relays. If I did, what did you have in mind? Just curious.

Audioguru, I like that idea. Why is the diode at the emitter needed? The discharge will occur from the diode at the base, no?

Thanks for your responses,
Sid

 

[audioguru2]

Hi Sid,
The diode is needed at the emitter to be certain that the transistor turns-off. Without that diode, if the voltage drop of the input diode is less than the voltage drop of the transistor's base-emitter junction, or if the input voltage doesn't reach exactly 0.0V, then the transistor will have more than enough base-emitter voltage (supplied by the capacitor) and it will stay turned-on. The diode "lifts" the voltage at the emitter so that the transistor will turn-off even if the capacitor's voltage is not completely discharged.

 

[vajirkar]

Thanks, nice explanation audioguru.

Thanks again for the suggestion. I'll give it shot over the weekend.

- Sid

 

[GPG1]

I was thinking of connecting a contact from relay 1 in series with relay 2 and vice versa so yhat only one relay could be "on" at a time

 

[MP1]

Why the 1M resistor? Start by changing it to 100K.
This resistor value is generally used for stability and to insure that the transistor switch is completely turned off. This resistor insures that the base of the transistor does not go slightly negative which would cause a very small amount of collector current to flow.
The value of this resistor is not critical but a value about 10 times the base resistor is normally chosen.

MP

 

[vajirkar]

GPG.. yes that's exactly what I have done, and that's why I don't have spare contacts. For some reason I still see a spark. See the attached image.

MP, you are probably right. I did have 100K and 1M there first I may have made a mistake in switching to 10K. I'll try using the 10K and 100K as audioguru suggested.

- Sid

View attachment 35673

 

[vajirkar]

Let me add that the load is a motor of a linear actuator. Could it be possible that its not good to suddenly change the direction of rotation for a motor (inductive) load? Maybe there is nothing wrong with the transistor switching circuit at all!

- Sid

 

[GPG1]

Seems that way. Go solid state.

 

[hotwaterwizard2]

Too bad there wasn't some way to incorperate a
Set / Reset Flip-Flop into the circuit so that only one relay is on at a time.

 

[hotwaterwizard2]

Here are some useful Links

http://ourworld.compuserve.com/homepages/Bill_Bowden/r_ctrl.htm

http://www.reed-electronics.com/ednmag/archives/1996/020196/03di2.htm

http://www.tfs.net/~petek/autoelec/simple/tog3pos.htm

 

[hotwaterwizard2]

I combined some circuits and here is what I came up with!

View attachment 35681

 

[vajirkar]

Thanks for all the detailed info guys.
I tried all the different techniques, but the main problem seems to be that sudden switching of the large inductive load isn't a great idea.
All I actually needed was a way to stop the actuator motor, so I just added an extra button to switch off both transistors and stop the actuator.
I have another question. I'll just start a separate thread for that.

Thanks again,
Sid

 

[hotwaterwizard2]

Here is another View of the circuit I came up with just in case anyone may need it for something. I can come up with a faster relay for you that will not have a spike if you like.

View attachment 35683

 

[hotwaterwizard2]

I believe this should do the trick!

http://www.amarkham.com/andrew/light_master.htm

 

[audioguru2]

Let me add that the load is a motor of a linear actuator. Could it be possible that its not good to suddenly change the direction of rotation for a motor (inductive) load? Maybe there is nothing wrong with the transistor switching circuit at all!

- Sid

Hi Sid,
That's it!
I think that it is normal for a relay contact to spark if the inductive motor is disconnected from the mains supply at the moment that the AC voltage is at its peak of 170V. The resulting hundreds of volts of back-EMF will also cause arcing at the bouncing contact of the reversing relay, which a moment later is trying to pass a huge current to slow, stop then reverse the motor.
The arcing will be fixed by adding a zero-crossing detector circuit.

 

[trigger]

I think TRIAC + DIAC can do the high voltage switching without a spark.......

I have done projects in 100W light bulb switching using LDR as sensor at 220Vac. The switch on and off depends upon the DC level to the DIAC.

So if needed the circuit I can draw them out........

(or you can google search..... should find in many free schematic sites )