EveryCircuit

Reviews

Home

modified 6 years ago

What makes it start to blink

5

47

274

06:15:52

I hope someone can help me out, I can't figure out why after a while, the circuit will choose one of the LED's to start the blinking cycle. Why don't they all stay on all the time.

published 6 years ago

6 years ago

Isn't that what oscillators meant to do!?

6 years ago

importante is the correct time setting in simulation. This configuration 30kOhm 10uF is good for 3Hz. I set a AC source as dummy and tune it 3 times higher in frequency then i expect it to run. Here about 10Hz. Then i highlight one node on the scooe to follow the voltage. Swipe with your finger to the right and you get for 10Hz the perfect simulation speed. Restart simulation and it will start. Sure, such a full symetric design actually has no chance to start, cuz its symetrie would avoid it. But there is always some asymetric numeric rounding in computer. Even its very little it does increase over time exponential. Hope this helps. Fastest way in this config here is to place a push button from base of one the transistors down to ground, but this is not what we have in reality 😀 http://everycircuit.com/circuit/6472282550304768

6 years ago

because its oscillator and this is matematical software with some equation corrections.... If you want to keep both LEDs shining withouth blinking, just add one SWITCH between BASES of both transistors...... Just conect the bases to same line stop the oscilation

6 years ago

in ideal world, oscilation cannot start, because its perfectly symetric... But in reality, components have different values even due to humidity and ait temperature..... so it will every time oscilate.

6 years ago

just try to change the value of one resistor by only 1 to 99 ohms :-) and you will see imidiatly result

6 years ago

ok not imidietly but maybe 3times shorter delay time :-)

6 years ago

@Calgerion, as you noticed its not the case you made it asymetric, its more the fact you have changed something. It doesn't matter if its 1 ohm up or down. It does the same effect if your restart with 99 and then change to 100 by 1 up, and it will start. Asymetric itself is equalised by the Dc operating point everycircuit calculates at startup and thats why nothing happens. In reality we have thermal noise which helps to bring things out of balance, anyway in reality there is no Dc op! So already at power on, things are very unbalanced. While here in simulation the first steps everycircuit does are to bring everything into balance to get a stable DC Operating point. After that EC have it hard to bring it out of balance, sometines more and sometimes less.

6 years ago

In these simulations the computer program does the maths left to right, top to bottom, in this way voltages and currents in the first part of the circuit have started to change before the same calculations have been done on the other part of the circuit, thus when the program does the next set of calculations the circuit the values of the first part of the circuit are different from the the values of the second part of the circuit. Thus the second part of the circuit will re-act accordingly to this change. This emulates a real life circuit, but in that case the circuit component values selected are never quite the same, thus one part of the circuit will start to change before the other, and the other part of the circuit will re-act in accordingly with that change.

6 years ago

this would imply everycircuit does work sequenciell throught all components step by step. But thats probably not the case. Actually a circuit can be sees as a big equation you can put in a matrix and by transforming this matrix you can solve for any node the voltage, that means you calculate a while but have the results for all nodes at one point in time, so no sequentiell woking. This explains why some matrix combination get stuck and and needs help in "change" anything only a little.

6 years ago

@hurz, this maybe so, but matrix type of simulator would need a lot more computing power and a grossly more complicated program than the line scan or sequential method. The line scan method is used in most PLC's and DCS's, and some tests that I have performed seem to indicate that it is being used by this program.

6 years ago

How good are your programming skills?

6 years ago

Anyway, we can only speculate, but im not with you about the performance thing. Computers in mobile phone can do much more then you probably can imagin. E. g. there are even HW support to do matrix calculations you cn make use from graphic chips open GL and even inside the central core. But so such a little everycircuit the performance is not coupled with a size of a matrix which is in dimension sqrt(nr of node) As long we dont know we can only speculate but would be interessting to get an inside view of eveycircuit and how its implemented. I speculate it does not make use of HW accelerators and can easily solve the matrix from SW libs.

6 years ago

I can agree about the power of new HW & software. One of the tests that I have performed was was a simple comparison of two identical circuits, like the circuit above, I tuned one example up and got it performing perfectly, checking all the perameters as much as could be done, then in one frame I placed two of these circuits, they were side by side. In the other frame they were one above the other, if they were running in a matrix the results should have been similar for both examples, they weren't, it was only the pair that was one above the other that was.

6 years ago

Am i understand correct, you saying or what you have found out, the order/positioning of circuits relative to each other does change its behaviour?

6 years ago

That is correct

6 years ago

i will dig into it. I have never noticed this kind of behaviour. You do have a good example circuit? Good in sense of minimal

6 years ago

probably a more simple circuit will do? Just RC or RCL?

6 years ago

Try, http://everycircuit.com/circuit/5532346137444352 & http://everycircuit.com/circuit/6130826929242112

6 years ago

but this is just one circuit per sheet, suppose to be two circuits per sheet? The first one in 305ms/s simspeed while tbe second is at 219ms/s so if you compare them they might react different over time!

6 years ago

if i set both circuits with equal simspeeds they behave totaly identical, i cant see any differences

6 years ago

You asked for simple circuits, so I tried to simplify it for you. If you look carefully, there is differences in circuit starting.

6 years ago

i was thinking about a much smaller circuit. Yes, there is a difference, but only if the simspeed is different. Make them equal and both circuits are identical. I dont see any differences.

6 years ago

I started out with them the same, must have messed them up. Ok, can you explain the following circuit for me, http://everycircuit.com/circuit/5953153879769088

6 years ago

No i cant😂 But this is exactly the question, why does a fully symetric circuit decide to go left or right, actually it should not start at all, only if little very little differences happen. I found an even simpler circuit where the question is why its going left and not right. wait...

6 years ago

have you seen, Lenny tries to separate us both and is spreading hate. We both do have a twist about this topic. That guys is a nightmare for EC. To ne back to this issue here, check this, its less complex in nr. of components but has the same dilemma http://everycircuit.com/circuit/5459159139745792

6 years ago

Don’t underestimate BillyT, he is fully aware of the stupidity flowing between both of us, nether side willing to heed, akin to a bunch of kids playing in a sandpit throwing sand at each other.

6 years ago

But I’ll stop throwing sand at you if you stop posting denigrating comments (you think is “critique”) on my circuits, respectful and constructive critique is always welcome on anyone’s circuits my friend...😎

6 years ago

Lenny you do not deserve respect, as long you behave like a baby and publish crap. Stop that

6 years ago

BTW, dont try to split me and BillyT that will not work. Verbalisation is the only thing you can. One hand does not make applaus my little Lenny be careful.

6 years ago

If you have something consturctive tell us, if not shut up Lenny my little bastard.

6 years ago

Just stating the facts my friend...😎

6 years ago

you are a psycho and the show is over and out. fin

6 years ago

You’re entitled to your opinions my friend...😎

6 years ago

It certainly is a dilemer, http://everycircuit.com/circuit/5410371482157056

6 years ago

how about this, its not random there is a system behind, but i have no idea what makes it go left or right. Little numeric rounding. But why are they in the lower RS inverse to the upper on. Maybe Igor knows an answer, in case everycircuit sees a deadlock situation it solves it with an voltage/current injection. Cause its still not clear why this goes into a stable position, it should stay in an undefined state 2.5V in between HIGH and LOW for both RS outputs! http://everycircuit.com/circuit/5725180572467200

6 years ago

here a fully symetric RS latch with mosfets, there is no reason to leave the stable undefined state which is not HIGH and also not LOW, only if we make it at some point asymetric it does leave the confortable situation and get HIGH and LOW http://everycircuit.com/circuit/5051597294665728

6 years ago

It seems to depend upon the the degree of symmetry and the amount of time allowed for state change, (change the time base to speed it up.) http://everycircuit.com/circuit/6305955109928960

6 years ago

Thanks for the help guys. Just for clarification: This is the correct behavior that wanted for my circuit. I just didn't know why or how it decided on one side over another. I now realize that even in a digital simulation, when you get down to the calculations, some part gets calculated first, and rounding does happen, which seems like the most probable reason for the small difference that occurs. Difference in wire length didn't seem to be a considerable influence. Neither did complete repositioning of the components, it always seems to pick the original 'left side LED' to power off first no matter how much I manipulate the time or the location of the original 'left components`.

6 years ago

thats what we tried to find out, is there a prefered side, and i dont see a prefered side like you see. The example with two RS latches pulled down to zero and the given free to choose its position one might set left side while the other one choose the right side, so there must be some noise on the lowest level we cant see or better we have not checked so far. We know there is noise which i was thinking is just numerical calculation 12 digits after komma it is just nunerical noisy. but i have the feeling Igor might inject some pseudo random noise on a nano or pico ampere or volt level, maybe we can look for it.

6 years ago

And right, the wiring length is not part of this simulation. For positioning i also dont see any impact.

6 years ago

@BillyT, back to your observation the order how things are placed do impact its state, which i cant see. This circuit here proofs the order doesn't matter, first two latches are normal placed while the last two are swapped, but even they are swapped and should be calculated in a different sequence according to you observation they behave totaly equal to the first two latches !! check it http://everycircuit.com/circuit/5481961657991168

6 years ago

@hurz, the difference shows up better if they are vertically stacked, try running the attached example for 15ms after closing the switch, put a switch in yours &compare the results,, http://everycircuit.com/circuit/5914036575928320

6 years ago

can you demonstrate directly in my circuit what you mean. i ubderstand i shall copy the switch to my latest circuit and move the latches around, but why and how that is your knowledge, please do that and show me

6 years ago

BTW, you noticed dark grey components do have undefined inputs or at least one undefined input and are taken out from simulation, so as they were not present at all !!!

6 years ago

1. The effect now does not appear to happen for every event (test) and the test results are different, when originally tested, my Blue trace was constantly single step, as was your Green trace, this time both Blue traces some time differ. 2. The dark grey goes away when you close the switch. 3. If the switch was not closed and the observations made as instructed, the differences will not be observed. Here is a copy of the changes made to your circuit. http://everycircuit.com/circuit/4685191201947648

6 years ago

ok, maybe it wasn't clear enough, inrestart always by stop sim, rewind sim and then start it again. You come now with a new scenario of an "restart" which is more a warm start and not what i did a cold start. Anyway, what i see in your warm start its does get random and no theory seems to be prooven of order in circuit layout or matrix equal calculation in time. Somehow with your warm start methode it does somehow remind form previews operation something which gets more and more out of sequence and looks totaly random. I prefered to use the cold start to prove EC sim behaviour. What do you think is the best way. Remember the original question was does the order of the circuit impacts the result? Is this what we are hunting?

6 years ago

can we, or we might have to make the headline what we are looking for more clear and compact?

add comment in editor

EveryCircuit is an easy to use, highly interactive circuit simulator and schematic capture tool. Real-time circuit simulation, interactivity, and dynamic visualization make it a must have application for professionals and academia. EveryCircuit user community has collaboratively created the largest searchable library of circuit designs. EveryCircuit app runs online in popular browsers and on mobile phones and tablets, enabling you to capture design ideas and learn electronics on the go.

Copyright © 2026 by MuseMaze, Inc. Terms of use Privacy policy