Yes...but there is a race condition between trigger and power-up. The trigger happens too fast.

IF that's the case TS might try connecting the start switch across the FET.
SG

 

Yes...but there is a race condition between trigger and power-up. The trigger happens too fast.

Datasheet say that it powers up not triggered. The RC network on the trigger pin is meant to delay trigger by 100ms to enable stable power up. It is triggering so doing what I need in that respect. What's the implication with "too fast"?

 

The RC network on the trigger pin is meant to delay trigger by 100ms to enable stable power up

Reverse connections on RC. Capacitor from pin6 to ground.
SG

 

Will do. FYI when low the output happily drives >11mA into an LED connected to supply.

0.02V

 

0.02V

That's good below .1 volt
SG

 

That's good below .1 volt
SG

I'm seriously thinking of ditching the FET and using an opto-coupler to control the power. The Output can clearly drive it.

 

I'm seriously thinking of ditching the FET and using an opto-coupler to control the power.

I was getting to that myself. Did you reverse the RC network?
SG

 

I was getting to that myself. Did you reverse the RC network?
SG

Not yet. I was worried I might be not able to get the timer started but it is starting so I'm not to concerned about that bit!

I'll go and redo things for an opto-coupler and see how that goes.

Will post an update when I have more news.

Many thanks

 

I'll go and redo things for an opto-coupler and see how that goes.

Sounds good. This is my take on using an opto.
SG

 

Thanks,

What is the diode doing?

 

What is the diode doing?

Isolates the SET switch so only the chip activates the load.
SG

 

Isolates the SET switch so only the chip activates the load.
SG

Built your circuit using a Vishay TO14642 opto which has very similar characteristics to the CPC1008N (but lower Ron).

Timer aspect works fine but getting much the same problem as we had with the FET - When timer o/p goes high it doesn't go high enough to turn off the diode so the derived Vcc goes down but not off and we are still drawing 1.6mA from supply.

Playing around, I tried two and then three diodes in series. Each addition reduces the derived Vcc slightly.

Interesting, with three diodes in I'm seeing Vcc of 0.87V, 1.17V on pin 3 (o/p) and 2.26 on the cathode of the opto LED (3.33V on the anode).

If I remove the link from o/p to opto and hold SET to provide Vcc to cct o/p returns to 3.33V after the timing period.

I'm mighty confused now!

 

Built your circuit using a Vishay TO14642

I think you meant the VO14642.
SG.

 

I think you meant the VO14642. Any case the wiring should be as below.
SG.

Yes, I did mean VO14642.

Did you mean as per previous diagram?

 

Did you mean as per previous diagram?

I believe the schematic below is the correct wiring.
SG

 

Started off like that. currently have it in the AC/DC config with pin 6 to supply and pin 4 to circuit.

 

config with pin 6 to supply and pin 4 to circuit.

Try pin 4 and 6 to supply and 5 to circuit.
SG

 

Have tried that too. I went to the AC/DC connections to eliminate any possibility of being wrong

I usually use these with GND to pin 5 so seems logical that in high side use pins 4 and 6 would be the more +ve.

 

I usually use these with GND to pin 5 so seems logical that in high side use pins 4 and 6 would be the more +ve.

OK I think the problem is in the ICM7242. When no power is applied there is some residual resistance on the output pin which is causing the opto to stay on. Seen this before in other chips. Give me a few minutes and I'll come up with something.
SG

 

In the meantime I've just removed the link from the opto o/p to the cct and powered the circuit direct from PSU.

The timer now correctly resets the opto led if directly attached (cathode to 3.33V). With the diode in cct the opto cathode only goes to 2.4V so I suspect will still be partially on.