Hi All this is my first post on this forum .
I am building a light sensing device using an Lm324.The circuit can be seen in the image I have posted.In it’s final design it will be used to trigger a DSLR camera to take pictures of lightning. Stage 1 which is sensing with the photodiode and output led works fine, and the voltage at the anode ( test point A) of the LED is 2.4 v however when I take a voltage reading at test point B ( with what I understand to be a unity gain amp 1k ohm X 1k ohm I get 3.5V approx which to me is ok .If I cover the photodiode so that the sense led at point A goes out, the voltage at Test point B increases to 3.7 volts.
This is not what I would have expected
Can anyone give me some pointers?
The final circuit will trigger a CD4001 monostable into a 4N25 to trigger the camera
Any advice would be of great assistance
Cheers
Nick

 

Read the datasheet. It's right there, that the maximum output voltage is Vcc-1.5.

 

It would help to draw a functional schematic as opposed to a chip wiring diagram. The reason is so you can get the function correct before you wire things up. I can't tell from your diagram if it is correct or not because I don't remember what each pin in the package actually does. It also appears that you have some unused input pins which need to be tied down. When wires cross each other you need to be very explicit about whether there is a connection there or not. What is the value of the unlabeled resistor next to the one labeled, which I take to be 10 MΩ.

 

Welcome to AAC!

Can anyone give me some pointers?

Circuits are easier to read if you use functional symbols instead of wiring diagram.

EDIT: Corrected D1.

You don't have IC1C wired as a voltage follower.

 

with what I understand to be a unity gain amp 1k ohm X 1k ohm I get 3.5V approx which to me is ok

That's not the correct way to make a unity gain amp or buffer. Should be as below.

 

It would help to draw a functional schematic as opposed to a chip wiring diagram. The reason is so you can get the function correct before you wire things up. I can't tell from your diagram if it is correct or not because I don't remember what each pin in the package actually does. It also appears that you have some unused input pins which need to be tied down. When wires cross each other you need to be very explicit about whether there is a connection there or not. what is the value of the unlabeled resistor next to the one labeled, which I take to be 10 MΩ.

Thank you I will tidy things up

 

Welcome to AAC!

Circuits are easier to read if you use functional symbols instead of wiring diagram.
View attachment 232330
EDIT: Corrected D1.

You don't have IC1C wired as a voltage follower.

Hi Dennis
What program did you use for the functional diagram please?

 

The LM324 has 4 opamps in it but you are using only two if them. You are not disabling the unused ones that can cause oscillations and interference. An LM358 has only 2 of the same opamps in it.

Are the old LM324 and LM358 fast enough? They are very slow.

 

The LM324 has 4 opamps in it but you are using only two if them. You are not disabling the unused ones that can cause oscillations and interference. An LM358 has only 2 of the same opamps in it.

Are the old LM324 and LM358 fast enough? They are very slow.

Thank you

The LM324 has 4 opamps in it but you are using only two if them. You are not disabling the unused ones that can cause oscillations and interference. An LM358 has only 2 of the same opamps in it.

Are the old LM324 and LM358 fast enough? They are very slow.

Thank you I have also designed a similar cut using a ca3140 which works well ,it’s just that the LM 324 was in my collection and has more than one amp in it to circuit design more compact

 

The LM 324 was in my collection and has more than one amp in it to circuit design more compact

The LM358 has two of the same opamps as the LM324 in it and is in a smaller 8 pins package.

 

What program did you use for the functional diagram please?

I use an old freeware version of Eagle. It won't magically make you able to draw more readable schematics. That takes years of practice and a good amount of common sense. That excludes some with engineering degrees...

EDIT: added missing word (version).

 

That's not the correct way to make a unity gain amp or buffer. Should be as below.

View attachment 232331

Thank you

 

Why all the pissiness and distraction? The answer is very simple and a common error by newbies. The opamp is not rail-to-rail. Done.

 

Why all the pissiness and distraction? The answer is very simple and a common error by newbies. The opamp is not rail-to-rail. Done.

That's not what the problem is.
The TS said, "If I cover the photodiode so that the sense led at point A goes out, the voltage at Test point B increases to 3.7 volts".
He means when the voltage at point A is zero the voltage at point B is 3.7 volts.
That's because he wired the output buffer amp incorrectly.

 

That's because he wired the output buffer amp incorrectly.

Fair enough. I fixated on the wrong problem.

 

The LM324 has 4 opamps in it but you are using only two if them. You are not disabling the unused ones that can cause oscillations and interference. An LM358 has only 2 of the same opamps in it.

Are the old LM324 and LM358 fast enough? They are very slow.

We may think of these as slow opamps but compare them with the time it takes to open a shutter.

Not: Forrect Mims has done this and published his design. You might want to have a look around the web (I saw this about 20 years ago on his personal website).

 

Hi Dennis
What program did you use for the functional diagram please?

One software worth the learning curve is the free LTspice. You can make schematic drawings with it that look familiar to people here. Then you get the added benefit that, once drawn, you can simulate your circuit to see how it performs. You can trace voltages and currents versus time, for instance. It's just an imperfect simulation, but still an incredibly useful tool. Its biggest downside is that you only get a schematic and cannot go directly to a board layout. That's what more elaborate tools like Eagle are for. But more power = more learning curve.

 

One software worth the learning curve is the free LTspice. You can make schematic drawings with it that look familiar to people here.

I use LTspice on occasion, but I don't like the the look of the schematics. Many of them look like they were drawn by a child. If I was going to publish or post something, I'd be more likely to redraw it with Eagle.

Then you get the added benefit that, once drawn, you can simulate your circuit to see how it performs.

I still have a strong preference to use the simulator between my ears...

 

Hi All
It is a little while since I have replied.So a status update is in order.
I have learn’t to use iCircuit and produced a modified design update (schematic). I have redesigned the circuit slightly with a gain of 11. Gain = 1 + Rf/Rin. I appear to have achieved what I wanted which is a variable output from the second stage which will be a significantly higher voltage than the output of stage 1, meaning that when there is insufficient voltage output in stage 1 to light a red LED there will be a substantial voltage at the output of stage 2. (Stage 1= 147mV = stage 2 = 1.67v)actual values.
This result hopefully means that when a weak bolt of lightning strikes(ie a long way off) I should get a response that can trigger the camera via a cd4001 monostable and 4n25 optocoupler
If you can see any further shortcomings in my design your comments are welcome
Thank you all
Nick
PS attached is also what I believe will probably be very close to my final circuit but I don’t know why the program calls it an unsolvable cct

 

You can eliminate the 1K resistor Rin and the
1K on the output of the op amp at point B
The max output at point B is likely 3.7 volts not 5.7