I think I have a neat, simple and easily implemented idea for a soft start circuit

hifinerd

Member
2010-08-25 3:08 am
Hi All,

I have been trying to come up with a simple soft start circuit that routes power initially through a resistor or thermistor and then activates a shorting relay after a few seconds. I think I have come up with a workable circuit that is easier to implement than most discreet designs and should be relatively immune to noise and switch on hard.

A rough conceptual schematic can be seen attached below, but it will need refinement. Essentially I use a 556 timer, which is a dual 555 timer IC in the one package, as a high voltage, high current schmitt trigger. It activates sharply when the charge in the cap reaches 2/3 of the supply voltage, or 8v in this case, and won't change state again until the voltage drops to 4v. This makes it ideal for our application and hopefully makes it less prone to false triggering.

It is necessary to use two 555s (or one 556), as the 555 schmitt trigger is inverting and the first 555 will initially be in the high state, which is no good as this would turn the relay on from the outset. However, if we feed it directly to the second, the second will initially be in the low state. When the cap reaches 8v the first 555 will go low, which in turn will make the second 555 go high and activate the realy..

The 555 can source/sink 200mA and should be able to easily drive most relays directly.

I think this is a neat design that should be pretty foolproof and easily thrown together on a breadboard. Obviously some supply bypassing is needed for the 555 and the RC network needs to be tuned to suit your turn on delay. Oh and the 555 might need some zener protection to protect it from the inductive kick of the relay.

For more info on the 555 schmitt trigger have a look at this:

555 and 556 Timer Circuits

So what do you think?
 

Attachments

  • soft start.pdf
    9.8 KB · Views: 435
Last edited:

hifinerd

Member
2010-08-25 3:08 am
Last edited:
This was my thinking, too.
This delay thing needs only 1 timer.

You take the output to relay from DISCHARGE Pin.
I should take it from OUT.
Even if most relay could be triggered by DIS,
we have the OUT Pin for SOURCE and SINK a lot of current.

The formula for calculate TIME, as I have got from my SPICE Sim
seems to be:
T(sec) = RC x 1.11

RC = T/ 1.11

Here you can see the result of my testrun,
with RC set to 8 seconds ( 47uF and 153k )
 

Attachments

  • LM556_delay_1.png
    LM556_delay_1.png
    11.3 KB · Views: 693
Hi,
your final circuit proposal is good. It is worth posting for all to see.

Far better than the RC time delay that we see so many recommending after getting the circuit all wrong and fiddling with values to try to achieve some semblance of timer control.
I wonder how wide the timing tolerance becomes with variations in mains supply voltage?
I'll bet some of them never activate. Just like MyRefC when component tolerances go skiwhiff.
 
Last edited:

hifinerd

Member
2010-08-25 3:08 am
OK, here is the final circuit pending your suggested improvements

For your perusal...

EDIT: Ah crap! forgot the supply decoupling. Will add it later after hearing your additional feedback.

Also, one might want to use 15v as most people will be dealing with 12v relay coils.

 

Attachments

  • soft start.pdf
    11 KB · Views: 191
Last edited:
Won't vary at all if you use a regulated supply...I think.
not your 555 use.
It is a timer that is tolerant of supply voltage variations. It compares internal voltages and does not rely on absolute voltage.

Whereas, my comment was directed at the RC triggering a relay driver transistor. It is a flawed timer method.

Move (but invert it) the top diode to be in series with the Zener. The Zener Vf is similar to a standard diode Vf. It needs a reversed Diode to prevent the Zener shorting out the supply to the relay coil.
 
Last edited:

hifinerd

Member
2010-08-25 3:08 am
not your 555 use.
Move (but invert it) the top diode to be in series with the Zener. The Zener Vf is similar to a standard diode Vf. It needs a reversed Diode to prevent the Zener shorting out the supply to the relay coil.

Well actually that thing across the coil is a 12v TVS and as I understand it shouldn't conduct in either direction unless the voltage exceeds 12v, but ok I'll change it for clarity.
 
It is getting better.
This is how I would do it.
I always like to connect stuff like relays to GROUND.
And this is probably what we all like if possible.

One standard small signal transistor INVERTS LM555 out signal.
Resistor R2, 4k7, is important. When OUT gets low
there will be like 11.35V across R2.
And so the base current is limited to 2.4 mA and transistor is not destroyed.

Here is schematic.
 

Attachments

  • LM555_delay_2.png
    LM555_delay_2.png
    11.6 KB · Views: 357

hifinerd

Member
2010-08-25 3:08 am
It is getting better.

I always like to connect stuff like relays to GROUND... one standard small signal transistor INVERTS LM555 out signal.

Your schematic looks good lineup.

This is exactly why my first schematic in my first post had 2 555s, to invert the signal.

Initially sounds more complicated than your version until you realise you can get two 555s in the one chip, ie the 556. I changed it to one on advice from others here.

I think both versions would work well (555 with inverting resistor or dual 555).
 
Last edited: