JLCPCB Prototype for $2(Any Color): https://jlcpcb.com
Connecting an LM324 quad op-amp as a LED flasher... weirdly.
Connecting an LM324 quad op-amp as a LED flasher... weirdly.
By Julian
Youtuber, shed dweller, solar charge controller aficionado
..and no Arduino needed.
What's a common application for this type of circuit?
Fun, you may want to play with the oscilloscope and maybe with LTSpice a bit to see how it all compares. I think there are no real surprises. Relaxed video.
Might be worth converting the buffer to a voltage controlled current source, then the LED current would be dependent on the ramp voltage, making the brightness increase and decrease in a even more linear fashion.
Hey it's a new year, you know that means. Time to flash some LEDs, 6502 style. I enjoyed you z80 LED flasher and built one myself but now it's time for a new chip
Yes Julian, that is weird where the only normal part of the circuit is the ‘unity-gain-follower/buffer’ on the output. What’s strange is the comparator operation value swaps in each Op-Amp with capacitors in place. So yes, a strange but delightful configuration overall.
What happens to the current draw in the new configuration? More or less than the standard configuration?
Sadly email failed, so inserted here
Julian,
I hope this is able to reach you via EMail as comments in the YouTube vidoes always seem too short. Yet I will try to keep
this short as well. I keep this to basic text format only, thus no pictures/diagrams, keeping it cleanest possible for
virus etc issues.
As already mentioned in the comments your "Wierd Oscillator" is using the LM324 in many respects outside of
specifications. Many OpAmps would actually already be dead, and the LM324 many work but deterirorate over time.
Critical parameters related to this system operation include the following
For ESD protection, and often not well detailed in data sheets are diode, resistor networks on input pins, that absorb
some level of Electro Static Discharge. That can be peak currents reaching amps, but of very short, sub microsecond
duration. Typically data sheets will include a maximum average current. In the LM324 data sheet there is no secific limit
but similar to normal can be expected.
maximum input terminal current = 10mA,
That level is to be ensure that any source voltage applied will have sufficient source resistance to limit current below
that 10mA. Of course depending on pulse duration higher current could be applied, total energy should remain well below
equivalent of clipping voltage times max current. The higher the energy to more likely long term detrioration and failure.
In this LM324 case the diagram at sect 8.2 shows only negative side clipping diodes, but also stated is input must not
exceed Vcc-1.5V first paragraph sect 8.1. Any higher and the input transistor base emitter junctions will be destroyed.
These are tiny junctions, designed transistors are only uA collector current and input bias current in 10's of nA, and
most certainly will not safely take even 10mA.
Differential input voltage is stated in the data sheet at up to maximum Vcc voltage, sect 6.1, but caution the above
mentioned max input must be less than Vcc-1.5V so what ever supply voltage, inputs must not go above Vcc-1.5, at 5V
supply that will be 3.5V. Negative side is listed at -0.3V, that is so those clipping diodes mentione dbove, do not get
active condition. These limits are however mainly for linear OpAmp operation, and that is not yet breakdown of the double
base-emitter junction, or negative clipping byhe diodes. So for typical silicon devices that B-E breakdown would likely
not be for up to 10V higher. But also the equivalent circiuit in 8.2 is not true complete circuit, with transistors on the
chip having isolation junctions (diodes) etc. Beyond Vcc the chip is in potential danger of various interconnects and
possible through conduction supply to ground. Some older devices suffer SCR latchups but modern designs try to minimise
that risk.
Finally one critical parameter to consider is the output drive capability. As the output is switched at Vcc/2 threshold,
the capacitor has about Vcc/2 on it, and being driven by change of state on output will try to go to Vcc+ Vcc/2 on low-
high and Gnd-Vcc/2 on High-Low transistions. The indicated negative end diodes will clip at about GND-0.7V with whatever
peak output drive is available. What happens on the positive side is not certain but definite beyond the Vcc-1.5V spec,
but maybe still under B-E breakdown is supply is <12v.
Iout MAX = +-60mA SHORT and around 20-60 source and 10-30mA sink.
———
PROPER OSCILLATOR USING AN OPAMP IS USING ONLY ONE AMPLIFIER.
Place a Capacitor to ground from -ve input to ground, and oscillation rate resistor from output back to -ve input and
capcitor. Now to oscilate we need to use positive feedback hysteresis using the +ve input. This is done with two
resistors giving the 1/2 supply level to the +ve input a divider much as you inlcuded in your circuit, and a feedback
resistor from output to +ve input and supply divider. The feedback resistor in ratio to the equivalent divider resistance
will set the hysteresis, taht now will give an upper and lower switching pioint of the charge volatge on the capacitor.
Say we use 10K resistors across the supply, that gives 1/2 supply level and equivalent resuistance of two 10K in parralel,
thus 5K. To get upper and lower threshold 75% and 25%, other levels equally feasible, and assuming full supply swing out
the output, calculate Vcc*5K/(Rf+5K) =0.25Vcc, clearly Rf =3* 5K. Note the LM324 does NOT drive full supply level at
lower loads, a load 2K only to Vcc-1.5V, the output for >10K load is not well specified but probably still a bit short
of full Vcc. Anyhow for practical purposes affects the actual threshold on high side so not quite 75% attained.
Vcc ——–
|
——-| 15k |—–—– | ——– |
| | | |
10k | | | | \ |
| | ———————| + \ |
—– | | \ |
| | | >—————-
| | ——–| – / |
———- | | / |
| | / |
| | |
—– | ——– |
| | ——–| |——-
10k | | | ———
| | | Rt
—– _______
| _______ Ct
| |
===== |
======
TRY IT, As the capacitor charges, from 0V, the output will of course be positive a -ve is less than +ve input, and the
threshold will be upper limit as output is high. Once the capacitor has reached high enough, the +ve threshold is crossed,
forcing output to go low, causing the threshold on +ve also to go to low limit. As output is low and thus discharging the
capacitor, the -ve input will move downwards till the low threshold on +ve is reached and the cycle reverses. And thus is
born a oscillator that has frequency controlled by Rt, Ct and the threshold chosen for switching. At no time will the
device limits be exceeded, as lomng as the OpAmp can acces larger input differentials on input like the LM324. NOT ALL
OPAMPS can acces large differential inputs! A comparator can also work fo this oscillator but usually is open collector
output, but the hysteresis ntwork can be used as load, and carefull design uses no more resistors, though more complex
calculations.
good luck, leave the weird circuit it is not using devices within specification an depending on capacotor stored energy
may long term fail or even immediate fail for manyopamps. as the LM324 has 10's nA input bias the resistor values can
crtainly go to 100's K ohm even Meg Ohm if some bias inflince is allowed. There by usin much lower power to run, than usin
K ohm range values, and smaller capaciors for longish periods.
hope this is of use to you. Have a Merry Christmas and Happy New Year.
another channel ":Paul" over at "Learn Electronics" WEB- SITE asked to check out your MOSFETS experiments HELP elp get a better grasp me.
Merry Christmas Julian… Can you put a scope on the circuit?
Merry Christmas.
Is there a way to make "flasher" but instead of LED it would turn on/off small heating element?
hello good video of what book you used to make the osculator