0
\$\begingroup\$

I was a silent reader for a long time and could solve all the problems on my own, but this problem is really freaking me out. I was messing around until 3 AM today with this "feature" and still don't know what is going on...

Initial Situation

There is a LED strip, that is blinking with 1.5 Hz (50% on time) and +12 V. I have to create a circuit, that can be placed between voltage source and the LED strip that delays the blinking by a few hundred milliseconds. It does not have to stop blinking these hundred milliseconds later, delaying the on time is enough. Interrupting the ground connection is no option, because there are a few more more LED stripts that share the same ground potential.

My current circuit

Due to the fact that my LED strip is connected to ground and I have to delay the incoming +VCC signal, I think (am I right!?) the easiest way is to use a P channel MOSFET as high side switch. I know that I could use a N channel MOSFET with a bootstrap driver, but i read a lot of articles that say using a P channel MOSFET is just fine, if you don't care about the higher Rdson (I am switching ~200 mA, so i don't really care about Rdson). In addition my PCB space is limited and a DIP8 driver would not fit well (Whole pcb should be designed as through hole).

The gate signal for the MOSFET is coming from the output pin of a NE555, which threshold and trigger pin is connected between a resistor and a capacitor.

Problem

I have really strange things going on with my circuit. For test purposes, I connect a single LED with a resistor to VCC, to test the functionality, but I just cant get the expected result. First, my circuit was working a little bit, the led blinking was delayed for the expected time, but went on for a few milliseconds when +VCC was applied to the circuit. I think the reason for this is, that the MOSFET is conductive by default, and my NE555 needs a little time to turning the mosfet off.

The other, much stranger problem is that after a few blinking cycles the LED is on the whole time!? First, it is glowing just a bit but then starts flickering/changing its brightness...

I have attached my current circuit. I would be so happy if somone could tell me what I messed up with this circuit design.

Or if there are much better solutions (maybe attiny13 as gate control!?) I would appreciate every recomendation.

Thank you in advance!! And sorry for my bad english :X

enter image description here

\$\endgroup\$
  • \$\begingroup\$ ”In addition my PCB space is limited and a DIP8 driver would not fit well” Choose a smaller capsule! \$\endgroup\$ – winny Jun 28 '18 at 17:44
  • \$\begingroup\$ How did you connect your led strip? If I understand it right it is connected between VCCOUT and GNDOUT. Don't forget to put a resistance in series to it, otherwise!! . Further I don't understand why you put in D1 and R2. It makes none sense. \$\endgroup\$ – abu_bua Jun 28 '18 at 18:04
  • \$\begingroup\$ What happens if you disconnect R3 from the 555 (with the LED strip connected)? The strip should be dark. Then what happens if you ground the end of R3 that was connected to the 555? The strip should work as if connected directly to VCC. If not, the problem is not in the 555 circuit. Also, is the VCC stable when the LEDs blink? \$\endgroup\$ – Elliot Alderson Jun 28 '18 at 18:18
  • \$\begingroup\$ A much better solution is a Schmitt Inverter Relaxation Osc \$\endgroup\$ – Sunnyskyguy EE75 Jun 28 '18 at 18:20
1
\$\begingroup\$
  1. You should add a resistor from VCCIN to GNDIN to discharge the capacitor C2 more-or-less fully when power is removed.
  2. It's been a long time since I've seen an NE555 let alone played with one, but check that the output pin is going close enough to the positive rail with your 11K pullup to VCC. If it doesn't get within 1.5 to 2V of the rail the MOSFET can slightly turn on. The CMOS versions do not have that problem, but the NE555 is the original bipolar design by Hans Camenzind (RIP).
\$\endgroup\$
  • \$\begingroup\$ Won't R2 and D1 discharge C2 quickly if VCC falls? I don't see that another resistor is needed for that. \$\endgroup\$ – Elliot Alderson Jun 28 '18 at 20:48
  • \$\begingroup\$ @ElliotAlderson Yes, it will if Vcc falls. There is no guarantee Vcc will fall all the way to zero quickly once input power is disconnected, unless it is shorted or there is a resistor across the input. Especially if that's actually a CMOS 555 it could take some time. \$\endgroup\$ – Spehro Pefhany Jun 28 '18 at 20:50
  • \$\begingroup\$ What bad thing will happen if Vcc falls slowly? As long as the voltage on C2 is not (significantly) greater than the 555's Vcc, isn't the 555 protected? \$\endgroup\$ – Elliot Alderson Jun 29 '18 at 11:39
  • \$\begingroup\$ @ElliotAlderson If the OP removes power and restores it, the desired delay may not occur because C2 has not been fully discharged. \$\endgroup\$ – Spehro Pefhany Jun 29 '18 at 11:42
  • \$\begingroup\$ My understanding of the question was that the OP only wanted to delay the turn-on time. I don't think they care about a turn-off delay. \$\endgroup\$ – Elliot Alderson Jun 29 '18 at 11:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.