Will this simple RC circuit give a 1 sec delay?

Thread Starter

Mrdouble

Joined Aug 13, 2012
107
Pretty sure it will work but maybe others may offer better design.
My largest concern is the transistor not going into complete saturation (have a hard time reading data sheets :/). The transistor is a 3904. The load is a relay coil (720ohms)

View attachment 107715
 
Last edited:

MrAl

Joined Jun 17, 2014
11,389
Hi,

Why upload four times?

The theory there is that the transistor turns on in one second because the base emitter reaches the turn on voltage of the transistor, but that is oversimplified. The transistor STARTS to turn on at a certain voltage, it does not turn on completely at that voltage.

What you need is a Schmitt Trigger circuit, which would take more than one transistor.
There will still be some variation with temperature however, that's why the modern way is to use a comparator.

If you dont need an accurate trip point then the single transistor may work, but it will turn on somewhat slowly as the input voltage rises little by little. To get it to turn on fast at least one more transistor is added in what is called a Schmitt Trigger configuration.
 

Thread Starter

Mrdouble

Joined Aug 13, 2012
107
Thank you
I uploaded pics from phone and forum not very friendly. I switched to comp. deleted all pics. Now I'm back on phone lol. Either way. You answered my question
 

crutschow

Joined Mar 14, 2008
34,280
Here's a simple turn-on delay using a BJT and a 6.2V zener diode to give a longer delay for a given RC time-constant and also to give a sharper turn-on.
It should work fine if you don't need a precision delay time.

Note that the data sheet segment you originally posted listing the transistor gain is for linear operation (I can't open your present attachment).
For minimum BJT switch saturation voltage the rule-of-thumb is to use a base current of at about 1/10th of the collector current (although I used a little less than that here since minimum saturation voltage is not needed to drive the relay).

upload_2016-6-15_1-0-17.png
 

Attachments

MrAl

Joined Jun 17, 2014
11,389
Here's a simple turn-on delay using a BJT and a 6.2V zener diode to give a longer delay for a given RC time-constant and also to give a sharper turn-on.
It should work fine if you don't need a precision delay time.

Note that the data sheet segment you originally posted listing the transistor gain is for linear operation (I can't open your present attachment).
For minimum BJT switch saturation voltage the rule-of-thumb is to use a base current of at about 1/10th of the collector current (although I used a little less than that here since minimum saturation voltage is not needed to drive the relay).

View attachment 107720

Hi there Carl,

That's a cute circuit, but will it turn on fast enough to properly operate a relay? We want the contacts to close as fast as possible, so why not design him a little Schmitt Trigger circuit with two transistors? Should be easy, and will work very nicely.
 

AlbertHall

Joined Jun 4, 2014
12,345
will it turn on fast enough to properly operate a relay?
Relays generally switch snappily even with a slowly rising coil voltage as, as soon as the armature starts to move the air gap changes and this gives a positive feedback effect which accelerates the motion. This works both opening and closing the relay.
 

crutschow

Joined Mar 14, 2008
34,280
Relays generally switch snappily even with a slowly rising coil voltage as, as soon as the armature starts to move the air gap changes and this gives a positive feedback effect which accelerates the motion. This works both opening and closing the relay.
I agree.
I think the rise-time of relay coil signal has only a small effect on the closing or opening transition time of the contacts.
For example a relay coil often has a diode across it to suppress transients, which significantly slows down the coil current fall time, and I've never heard that that's a problem for the contacts, only that it delays the time for the relay to open.

Which reminds me, the relay coil in the circuit I posted should have a diode (1N4148 or similar) across the coil (cathode to V+) to protect the transistor as shown below:

upload_2016-6-15_7-23-40.png
 

MrAl

Joined Jun 17, 2014
11,389
hello again,

The question is not whether or not it will work, we know it will work, but without knowing the relay specifics we cant be perfectly sure of how well it will work. With a Schmitt trigger we can be sure regardless, and it seems even simpler than what i have seen here so far so why not. A better circuit must mean better operation.
Of course it the application does not suffer then who cares, but since the alternate is simple too, i dont see why anyone in their right mind would reject such an idea. The choice is an "ok circuit maybe" vs a "works good circuit", who would choose the former when the complexity is similar?

Here's just one example:

Schmitt_npn.gif

That's just an example, there may be better circuits. See if you can design one.
 

crutschow

Joined Mar 14, 2008
34,280
hello again,

The question is not whether or not it will work, we know it will work, but without knowing the relay specifics we cant be perfectly sure of how well it will work. With a Schmitt trigger we can be sure regardless, and it seems even simpler than what i have seen here so far so why not. A better circuit must mean better operation.
Of course it the application does not suffer then who cares, but since the alternate is simple too, i dont see why anyone in their right mind would reject such an idea. The choice is an "ok circuit maybe" vs a "works good circuit", who would choose the former when the complexity is similar?

Here's just one example:

View attachment 107733

That's just an example, there may be better circuits. See if you can design one.
My motto is KISS, so a one-transistor circuit is better than a two transistor circuit, unless that second transistor is really needed, and I know of no reason that a slow rise-time will have any negative consequences for a standard mechanical relay operation.

Your circuit will require an additional transistor to drive the relay coil since RE will reduce the voltage across the relay.

And I rather enjoy being in my wrong mind. :p
 

AlbertHall

Joined Jun 4, 2014
12,345
Alternatively, could write a simple program for a RPi to control the relay and then you would only have to reprogram it to change the timing as required :D
 

hellifino

Joined Jul 2, 2015
19
Here's a simple turn-on delay using a BJT and a 6.2V zener diode to give a longer delay for a given RC time-constant and also to give a sharper turn-on.
It should work fine if you don't need a precision delay time.

Note that the data sheet segment you originally posted listing the transistor gain is for linear operation (I can't open your present attachment).
For minimum BJT switch saturation voltage the rule-of-thumb is to use a base current of at about 1/10th of the collector current (although I used a little less than that here since minimum saturation voltage is not needed to drive the relay).

View attachment 107720

** Unrelated to the OP **

But I have to comment here just to thank @crutschow since I have been messing around with a similar circuit on the side for a few days now, using RC to shutoff another load... and I could never get the ******************* to work in the simulator (including the circuit he drew above). Turns out that STARTUP option fixes everything. ( /'-')/

So frustrating lol, but much appreciated crutschow. :)
 

crutschow

Joined Mar 14, 2008
34,280
........................ and I could never get the ******************* to work in the simulator (including the circuit he drew above). Turns out that STARTUP option fixes everything. ( /'-')/
..................
Yes, it may not be apparent but Spice simulators normally do a DC operating point analysis before doing a transient analysis which places all the DC bias voltages at the steady-state values with power supplied. This is the circuit voltages after all delays have occurred.
To avoid that you use the Startup or Skip Initial Operating Point Solution (UIC) options which skips the operating point calculation with the power applied. Then you can see the delay time from power ON.
 

MrAl

Joined Jun 17, 2014
11,389
My motto is KISS, so a one-transistor circuit is better than a two transistor circuit, unless that second transistor is really needed, and I know of no reason that a slow rise-time will have any negative consequences for a standard mechanical relay operation.

Your circuit will require an additional transistor to drive the relay coil since RE will reduce the voltage across the relay.

And I rather enjoy being in my wrong mind. :p
Hello,

First, that's not my circuit, just an example, so design to fit the application would be better.
Second, "My god, a two transistor circuit, that's way too complicated for me...i cant stand it".
Third, reliability is not directly related to the number of components.
Fourth, armature inertia is also a factor in the relay closure. I rarely make a point of anything unless i have a good reason, and experience with high power AC synth converters that have to switch into action when the power goes out tells me that there is a finite delay and issues with closing and opening relay contacts more slowly than usual, especially those with double throw contact action where one set is opening while another set is closing.

As i said in the past, the one transistor circuit will "probably" work, i just like to make sure and one small transistor is not a very high price to pay. Small transistors are cheap and reliable.

So if you want to work with the OP to get this working with just one transistor, that's fine with me :)
 
Last edited:
Top