Not to mention the fact, again, that a 741 is more than adequate here.

So let's focus on helping this guy, rather than telling him how inadequate he is in using the 741 here.

 

Use another 555 as a photo switch like this,,

Should never use it to power another chip. Use it to control the reset pin instead.

 

Hello DD,

I have straitened the picture a bit with the GIMP:



Bertus

 

Should never use it to power another chip. Use it to control the reset pin instead.

i use it many times to power relays directly,,its no problem powering another 555, plenty of current.

 

The simplest would be to take the ldr plus a resistor to the reset line - it is tricky in that while the data sheet says that that pin is the base of a PNP, it's current flow is bi-directional. An easier and more consistent approach would be to go through a MOSFET or high beta bjt.

A simple approach, almost identical to what he has, is to route the opamps out to the reset pin. Unfortunately, the 741 cannot get to the ground enough for this approach to work. A comparator with oc output or lm358 like device would.

 

it wont work on the reset line, as when its off, pin 3 will be held low and one of the leds will be lit continuously

 

Use another 555 as a photo switch like this,,it has built in hysteresis.



View attachment 103962

Why reinvent the wheel!

The chip in a solar garden light has a pin for the solar cell that senses daylight and does all the switching with very little extra parts.

The colour changing variety generate about 4 - 5V, which would do for the original circuit if a CMOS 555 was used.

The TS mentioned an LDR - they're getting hard to come by because the toxic cadmium doesn't sit well with RoHS. That's probably why the abundantly available garden light chips use solar cell charge current to sense daylight.

A discount store local to me is offloading colour changing garden lights half price - ATM: I'm savaging the LEDs for PC PWR LEDs, the solar cells, inverters and any other useful bit go in a box till they come in handy.

 

Why reinvent the wheel!

The chip in a solar garden light has a pin for the solar cell that senses daylight and does all the switching with very little extra parts.

The colour changing variety generate about 4 - 5V, which would do for the original circuit if a CMOS 555 was used.

The TS mentioned an LDR - they're getting hard to come by because the toxic cadmium doesn't sit well with RoHS. That's probably why the abundantly available garden light chips use solar cell charge current to sense daylight.

A discount store local to me is offloading colour changing garden lights half price - ATM: I'm savaging the LEDs for PC PWR LEDs, the solar cells, inverters and any other useful bit go in a box till they come in handy.

that was suggested earlier on Ian posts# 7,12....this guy wants to use a USELESS 741,,,

 

that was suggested earlier on Ian posts# 7,12....this guy wants to use a USELESS 741,,,

Sorry Dodgydave, but I do not necessarily WANT to use a 741, it was purely the first device I chose to use as a comparator, it could have been a LM393 or even perhaps a LM311 or some other comparator but it was a 741.

 

So let's focus on helping this guy, rather than telling him how inadequate he is in using the 741 here.

Thank you dannyf for your support. I am very grateful with the comment you left.

 

Sorry Dodgydave, but I do not necessarily WANT to use a 741, it was purely the first device I chose to use as a comparator, it could have been a LM393 or even perhaps a LM311 ot some other comparator but it was a 741.

The 741 is plenty adequate for such a simple job, but a comparator would be technically a better choice. A little positive feedback to give it some hysteresis and you wouldn't know the difference from using a comparator. Maybe try about 1M from output to the + input, if it takes too much change in light level to switch - use a higher resistance.

The LM311 is pretty old too - I didn't know you could still get them.

 

Should never use it to power another chip.

Actually, unlike most opamps, the 555 output stage is good for a few hundred mA, and can easily power 25 or 50 mA of downstream stuff. (with appropriate decoupling, or course)

Apparently - its possible to use the741 as a noise generator.

Outstanding response.

So what? Any device made by man, including your super duper opamp, can be used as a noise generator.

That's not the kind of noise he is talking about.

The whole 741 discussion is meaningless

That is.

Ok, you are clearly a hater of the 741 and consider it should be made obsolete

Actually...if you read my responses to you and others in this thread, you will see that I agree with you that the 741 is not obsolete, is still being designed into new products, and can work well in this application with a little help. What I disagree with are your reasons for supporting the 741. The fact that it is old definitely does not make it obsolete or instantly unqualified for this application. However, the fact that it is old and still in production does not make it a good choice for your circuit in post #1.

"hater" - really? Are you 12?

ak

 

Here is an alternate reality, with another 40+ year old part. Looks busy, but when you add hysteresis and decoupling to the circuit in post #1, the circuit below has fewer connected device pins. And remember that the circuit in post #1 still needs to have 555 power switching added, while that function (disabling both LEDs) is included here with even lower static power consumption.

Something I haven't brought up yet is the fact that long R-C time constant timer circuits do not perform well. Besides the basic inaccuracy of large electrolytic capacitors, they have very poor temperature coefficients, and the capacitor leakage current, which has its own temperature coefficient, can equal or exceed the charging current from the timer circuit. There are several oscillator-plus-divider chips to chose from, and any of them will give more reliable performance with only 7 connected pins, one fewer than a 555.

EDIT: Adjusted timer components. Should be about 2.5 min for each LED. R9 might need adjustment. The goal is for R8-R9 to set up 86% hysteresis (two time constants) to get more stretch out of a smaller value for C2.

ak

 

Actually, unlike must opamps, the 555 output stage is good for a few hundred mA, and can easily power 25 or 50 mA of downstream stuff. (with appropriate decoupling, or course)

ak


Outstanding response.

ak


That's not the kind of noise he is talking about.

That is.

ak


Actually...if you read my responses to you and others in this thread, you will see that I agree with you that the 741 is not obsolete, is still being designed into new products, and can work well in this application with a little help.

ak

The 741 is a well established functional block that has stood the test of time - but there are more recent parts optimised for specific advantages, the list of applications where a 741 just won't do is slowly but steadily growing.

 

"hater" - really? Are you 12?

ak

"Hater" - Did you not understand this? It is an English word. Look it up and read the definition.
But I do not understand what you mean by "Are you 12?" 12 what?
If you mean 12 years old - No, I'm 11 years old!

 

Many thanks everyone for all your suggestions - 5* 's to all of you. I am still not certain which circuit to choose but I'll be doing some more research and reading together with taking your advice I'm sure I will work something out.
My observations are:
The circuit in #40/#43 using 2x 555's (or 1x 556) looks good and is a nice idea but it appears to have quite a wide hysteresis. I wanted the circuit controlling the LED's to switch on and off more or less at the same lux level but obviously slight different so hysteresis would have to be incorporated.
The circuit in #53 - which I'm sure would be fine apart from the large number of additional components and could be difficult for me to produce a PCB. I was hoping to keep the design as simple as possible seeing as I am just a hobbyist, I am not an expert.
The idea of the MOSFET or transistor in #2 still appeals to me and also changing the 741 for a LM393 or LM324 or some other op-amp / comparator.
Thanks again and regards.
PS. The model log cabin is coming along a treat!

 

use you original circuit with a Lm324 , you get four chips in one, much easier to build the circuit by AK in post#53,

PS he's missed a 470k resistor from pin 5 to pin 11 ground

 

I am still not certain which circuit to choose

It is really quite simple. You have two key elements to your circuit that you need to determine:

1) the switch: you are using the opamp to switch on / off the 555 + leds. Not really preferred. The high side switch idea proposed earlier is the way to go and you can use a pnp or a p-ch mosfet there;

2) the light sensor: you are using the 741 for that. It works, but cannot turn off a pnp switch, or a p-ch with low Vgs (<1.5v, most of them are not in that category). You can use a 555 for it (too complex in my view), or a logic gate / n-ch or high beta npn, or even a tl431.

the final outcome with depend on the parts you have and your comfort level with each combination. It seems to me that the p-ch switch + 741 is the closest one to what you already have and is the way to go.

For me, I would have gone with a mcu: it allows you to drive multiple lights at random intervals. but then that solution requires some pre-requisites that not everyone has.

 

Here is how the switch + light sensor can be done:

It is a pnp driven by a tl431-based light sensor. R2 is the LDR (on resistance around 5K and dark resistance over 100K). R1 is pikced so that the switch turns on at LDR = 14K.

It works with mosfets too but does require 5Vcc or over for most mosfets, and works down to 3.3Vcc if pnp is used. You can replace the tl431 with a n-ch or a npn - need to readjust R1 however.

Once you understand how it works, no need to worry about the minor stuff, like if a 741 is too obsolete to work here.

 

Here is how the switch + light sensor can be done:

It is a pnp driven by a tl431-based light sensor. R2 is the LDR (on resistance around 5K and dark resistance over 100K). R1 is pikced so that the switch turns on at LDR = 14K.

It works with mosfets too but does require 5Vcc or over for most mosfets, and works down to 3.3Vcc if pnp is used. You can replace the tl431 with a n-ch or a npn - need to readjust R1 however.

Once you understand how it works, no need to worry about the minor stuff, like if a 741 is too obsolete to work here.

Thank you dannyf. Can I ask which software you use to design circuits? .. as something like this would help me a great deal.