EveryCircuit - LED Test

Robert_Kidd

5 years ago

It’s a shame since 3/4 of the current is wasted by the potential divider. There’s probably a better way.

Gerkl

5 years ago

if I put higher resistor than it is not waste I guess. I am playing with numbers.

Gerkl

5 years ago

I guess I got it. This looks good. Now time to test it.

Robert_Kidd

5 years ago

That’s much better.

Robert_Kidd

5 years ago

Top resistor 5.6 Ohm.

Issacsutt

5 years ago

Check this out, it’s a pretty good solution: http://everycircuit.com/circuit/5594636192841728

Gerkl

5 years ago

Thank you Issacsutt. I am trying to limit amount of components since my project does not have enough space for extra pieces.

Issacsutt

5 years ago

No problem. I went back and altered a few things so that it’s a little simpler now, let me know if this one will work for you or if it still needs to be even simpler: http://everycircuit.com/circuit/5594636192841728

Issacsutt

5 years ago

There’s also this awesome solution which uses a current source instead, it works very well is very simple and DOESN’T require using any LED resistors, (in fact, unless your using a high enough supply voltage, it won’t work with resistors unless there’re really small). Also, it will still regulate the current even over a pretty wide supply range too. Here’s the link for it...

Issacsutt

5 years ago

http://everycircuit.com/circuit/6336844659425280

Gerkl

5 years ago

Thank you again Issacsutt. I think I will stick with your first option because it keeps 3V when switch is OFF. In second option voltage before switch jumps up to 3.5V when switch is OFF. It may damage my touch sensor. All I know the touch sensor works max on 3V. That is why I have to drop voltage before switch (my touch sensor) to 3V.

Gerkl

5 years ago

And I already start designing PCB boards.

Gerkl

5 years ago

What am I doing wrong? Why I can't duplicate it on my curcuit? Gives me error - Can't find solution.

Gerkl

5 years ago

finnaly make it working lol

Issacsutt

5 years ago

Yeah not a problem I love helping people with this stuff! As for the 3.5v jump with switch open, that’s because with switch open there is no load other than the voltage divider of the regulator, which means it is trying to supply max voltage to maintain a 120mA current (which of course can’t happen with switch off). But honestly, all you would have to do is just wire the touch sensor after the switch and it should be completely fine.

Issacsutt

5 years ago

I was talking about the currently mirror one, which I think would give you the most accurate results, but make sure you know what the current and volt rating of you LEDs are... and i would strongly recommend building these circuits on a breadboard first to fine tune the feedback resistor before soldering it.

Issacsutt

5 years ago

Also, you can easily make your own touch switch using just 2 transistors, only drawback is it would require 2 touch plate/wire contacts, but you could place the two contacts so close to each other that it still only requires one finger to touch... if your interested in making your own touch switch instead (that wouldn’t have any special voltage requirements) than just let me know and I can send you a circuit diagram with description.

Issacsutt

5 years ago

Correction: ***I meant current mirror; not currently mirror***

Gerkl

5 years ago

Like a said before I am limited with space for my project. I already got the touch sensors. https://evandesigns.com/products/touch-sensor-switch?_pos=1&_sid=56502e7c1&_ss=r . They work fine. On Circuit that switch is representing my touch sensor. Unfortunately there is no suitable device from a list I can use to simulate that sensor on circuit. Can you help me choose right NPN and PNP transistors for this project please. I just ordered some generic one but I am not sure if they are right for this application. Thank you.

Issacsutt

5 years ago

Ok. Really almost any general purpose transistor should be fine for this one, such as 2N3904 for the npn and 2N3906 for the pnp are the ones I use most often. You can even get them from Amazon and for pretty cheap. But also, is the current in your simulation actually the same current as what the real circuit is going to see? Because if it draws more than that, then you may have to adjust some of the resistors.

Gerkl

5 years ago

my battery is 3.7V rechargeable 850mA. LED max current 20 mA - total 120mA when it is ON. I don't know how much touch sensor will use.

Issacsutt

5 years ago

Ok it should all work fine then...

Issacsutt

5 years ago

Let me know how well it works after you get it all put together

Issacsutt

5 years ago

If you don’t mind that is

Gerkl

5 years ago

yes. I will. It might take some time.

Issacsutt

5 years ago

Ok 👌

Gerkl

5 years ago

well. First board was total disaster. After some errors and trials finally fix it and made it run. But it drops to 2.5V with open Switch and it is not enough to run touch sensor. I guess it has something to do with transistors setting but I couldn't find what to change. That is bad that this builder is not allowing select generic transistors.

Robert_Kidd

5 years ago

Are you absolutely certain you have all components and connections correct? Double check.

Gerkl

5 years ago

yee. I did check several times.

Gerkl

5 years ago

I even simulated on different Circuit Simulator and getting same result as real circuit.

Gerkl

5 years ago

There are some settings for transistors which I don't understand what to change (adjust). I checked datasheet for 2N3906 and did not find those parameters. Like Saturation Current or Collector Resistance and etc.

Gerkl

5 years ago

Now this simulator stop working

Gerkl

5 years ago

Source 4-3.9V. Basically first PNP drops 0.6-0.7V than second drops 0.7-0.8V. Output 2.5 V.

Robert_Kidd

5 years ago

Right, although you’ve altered a couple of resistor values slightly from Issac’s original (which I screen shot) that doesn’t seem to be an issue. However, the original circuit has the switch closed at startup. If you replace the switch with a normally closed switch then it starts fine. It clearly needs a load as power supply circuits quite often do. I realise you probably want to conserve power but if you add a 1k across output to ground, before the switch you’ll see it works fine and you can start and stop without problem.

Robert_Kidd

5 years ago

It’s a bit deceptive, if you add that 1k it works fine, repeatedly. With it running, still with switch open, you can increase 1k up to, say, 20k and circuit still starts and stops. HOWEVER, if you close simulator, reopen the circuit and add a resistor higher than 1k it fails again. So it looks like it needs a few milliamperes load current to work reliably.

Robert_Kidd

5 years ago

If you can live with the extra few mA current consumption you may have a solution. There might be other ways to resolve this but as it’s 05:40am here I’m going back to sleep:-)

Robert_Kidd

5 years ago

Note also that if you add a n/o switch in series with your source (4V) then the sim doesn’t throw up the error and when you flick the switch it starts fine without any extra load resistor. If I get the time I’ll try to knock this circuit up tomorrow.

Robert_Kidd

5 years ago

I mean later today!

Gerkl

5 years ago

Thank you Robert for your contribution. Switch that showed over here actually represents my touch sensor (I don't know how to replicate it here) which works off 3V and connects to positive and ground. I though it would have enough Amp consumption but I guess I was wrong. I will try your suggestion on actual board tomorrow.

Robert_Kidd

5 years ago

OK.

Robert_Kidd

5 years ago

Out of interest I’ve just emailed Evan Designs asking what the current consumption is in the off condition.

Robert_Kidd

5 years ago

Evan’s have replied saying the touch sensor current consumption is a ‘nice low 1.5uA’ which explains why it’s an insufficient load for the power source.

Gerkl

5 years ago

I had to change resistors values to match my resistor's kit.

Gerkl

5 years ago

Unfortunately 1K resistor across power out did not work on actual board.

Gerkl

5 years ago

I gonna try to break this touch sensor down to see how it is made. Took a close look (microscope) on details. it is using A0LA MOSFET, RH6030 chip and some resistors. will work on a circuit later.

Issacsutt

5 years ago

Sorry I haven’t been on here in a few days... I just altered the circuit a bit hopefully this fixes everything... just let me know if it doesn’t and I’ll keep working on some more solutions. http://everycircuit.com/circuit/6156893498900480.

Issacsutt

5 years ago

Also, in case you were interested I left a description within the post on the above link, explaining the idea behind the improvement.

Issacsutt

5 years ago

Oh and this new circuit doesn’t have any issues starting the simulation with the switch open

Gerkl

5 years ago

No worries. I appreciate your help. I am familiar with electronics in general but never build anything from scratch by myself.

Issacsutt

5 years ago

Ok.. So are you just used to relying on IC chips then I guess? You never try designing discrete circuits?

Issacsutt

5 years ago

And no problem, good luck

Gerkl

5 years ago

I am not used to relying on anything. By being familiar I meant I know what transistor, resistors, capacitors are and what they do but never build anything myself neither on chips or anything else. Lol

Gerkl

5 years ago

I am more doing woodworking projects. Just needed to add UV light to my project and that is how I end up designing circuit. Learning on the go. But still don't exactly understand how it works.

Issacsutt

5 years ago

Ohh ok cool I was just wondering. Sorry if I miss understood you.

Gerkl

5 years ago

unfortunatly nothing changed. It is still dropping to 2.5 V without load

Gerkl

5 years ago

I am about to give up. It is hard to troubleshoot something that you don't know how it works. On different simulators I am getting different results. That is frustrating. Using Circuit Simulator 1.1.0 and it is showing what am I actually getting from the board but I don't know what should I adjust to get result I need.

Gerkl

5 years ago

California

Robert_Kidd

5 years ago

I can understand why you feel like throwing on the towel. Don’t!

Robert_Kidd

5 years ago

I believe the problem is to do with the transistor gain and we can get around it by changing the 2k resistor to a 5k potentiometer (with the centre pin connected to one of its end pins). I appreciate you don’t have much space and you probably don’t have a 5k pot sitting around either.

Gerkl

5 years ago

It was going same thing to me. First time I installed battery it was giving me 4 V. After I took it off and put it back it dropped to 2.5V.

Robert_Kidd

5 years ago

I’ve tweaked values slightly and am going to build later tonight. Will update.

Robert_Kidd

5 years ago

Where do you live?

Gerkl

5 years ago

Usa. California.

Robert_Kidd

5 years ago

Haven’t got very far, yet. Been breadboarding it but can’t get it to go. Tomorrow I’ll build it on strip board as I’m always dubious about connections on breadboard. In theory it’s quicker but when you doubt what results you are getting it’s a time waster. I’m in UK.

Issacsutt

5 years ago

Ok, I think this circuit may still have a lot of gain, and it’s possible that it could actually be oscillating in the hundreds of kHz to low MHz range (I have built this before but with a different power supply range). What I had done back then was add a cap somewhere between 0.47uF and 100uF between positive supply and the base of the main driving pnp transistor. If your using a multimeter to measure the voltage out, then it is most likely displaying the average voltage of the oscillation if it is oscillating. And I suspect when you have a low enough Impedance load on it, the oscillations aren’t strong enough to sustain... if you have an oscilloscope try measuring it, in the mean time I’ll try building this too, but I’ll have to use a 5v supply instead.

Issacsutt

5 years ago

Also the voltage drops and gain of the transistors vary immensely compared to the ones in simulation. Personally, I always build it on a breadboard and fine tune the voltage divider while testing with various loads before I ever trust it in an actual project. The output voltage should remain relatively constant within a few tenths of a volt and be able to maintain that over various loads as well as various supply voltages, and if it doesn’t do that then first adjust the voltage divider (I usually only alter the lower half value; r2), if that doesn’t work I would then check wiring as well as orientation of the transistors (the PNPs are usually oriented 180 degrees backwards compared to the NPNs on a breadboard, but that also depends on weather you have the circuit flowing from left or right) and I’d also check values of all the resistors. And also feel for anything hot. And if none of that resolves anything, then that’s when I might start questioning some things... it also helps tremendously to have an oscilloscope, (for years I never had one, but then I got a $25 on Amazon, which later on was not good enough for most of my projects so I got a real one for like $300 I think, but anything helps honestly). Sorry for the super long text, I just get carried away sometimes

Gerkl

5 years ago

To be honest I just wanted to build something simple and make it work. I was not expecting to dive in such details like measure signals using oscilloscope or designing and troubleshooting electrical circuits. Simply I don't have enough time for it. Thank you guys for all you help.

Issacsutt

5 years ago

Ok GOT IT TO WORK PERFECTLY!! I just scoped and measured everything under various conditions with the added capacitor, (which I found the best location for it was a little different then I originally had in mind)...

Issacsutt

5 years ago

Ok, here’s the new circuit; please don’t give up, at least make this your last attempt. http://everycircuit.com/circuit/6156893498900480

Issacsutt

5 years ago

Without the capacitor it was oscillating at over 200kHz with no load! And with the right size load, it actually oscillated up to 1.8MHz!!!

Robert_Kidd

5 years ago

Still get CFS.

Robert_Kidd

5 years ago

With your new V4

Issacsutt

5 years ago

What do you mean by CFS??

Robert_Kidd

5 years ago

Can’t find solution

Issacsutt

5 years ago

Oh that’s just EC being weird, I would pay no attention to it, cause it didn’t even show the oscillations in the original design.

Issacsutt

5 years ago

Actually, it’s only doing that because I changed the 500 ohm resistor to 100 ohms 😂😂. Change it back to 500 momentarily and try it!

Gerkl

5 years ago

Thank you. I will test it once i get capacitors from amazon.

Gerkl

5 years ago

LOL. I have another idea I want to put together and now I am afraid to even start it.

Issacsutt

5 years ago

Haha, what you got... we will make it happen! ;)

Robert_Kidd

5 years ago

@Issacsutt, why only change it back to 510 Ohm momentarily? It consumes quite a bit more current at 100 Ohm, why the change?

Issacsutt

5 years ago

That was just an example to show why EC showed “no solution”. This particular circuit relies quite a bit on gain, and the fact that EC is not displaying correct results is because it’s trying to model every bjt using just one transistor, but in reality no two transistors are the same. So please disregard the simulation on this one 👌

Gerkl

5 years ago

Lol. "WE will make it happen". Most likely you will be doing it and I will be complaining that I'd doesn't work. Lol. That is not fair to you.

Issacsutt

5 years ago

To be honest, I feel bad about that, and like I kinda owed you anyway because I never tested it first before suggesting it, and I’m sorry about that, it would have sped things up a lot had I just built and tested it earlier. I’m Really sorry you had to spend so much time troubleshooting. At least now though, I’ve learned some from that; so from now on I’m gonna try to test things more often unless I 100% already know it will work

Robert_Kidd

5 years ago

@Gerkl, interested to hear back when you made these changes.

Robert_Kidd

5 years ago

@Issacsutt, so what is the value of the 510/100 Ohm resistor in the final circuit? How about posting it as a ‘new’ or ‘solved’ circuit as it’s a pain keep nipping of to the unlisted circuit on an iPhone?

Issacsutt

5 years ago

Final value is 100 Ohms, but EC is not gonna be able to simulate it properly. Best thing I can do to fix the “no solution” issue is probably gonna be to modify it by added an extra component or something... but would that confuse anyone because it’s not intended to be part of the actual circuit?

Issacsutt

5 years ago

Oh I see, your saying I should post it publicly??

Robert_Kidd

5 years ago

Yes

Issacsutt

5 years ago

Ok cool I will work on that. I’ll let you know on here when it’s been posted.

Robert_Kidd

5 years ago

Thanks.

Issacsutt

5 years ago

Ok I just posted it, here’s a link to it also: http://everycircuit.com/circuit/4833339868184576

Issacsutt

5 years ago

Unfortunately I had to completely redo it on a new post in order to make it work cause otherwise EC said it was a duplicate and wouldn’t give me the option to list it publicly.

Issacsutt

5 years ago

I put two sources on there, one of which was a sine wave source so that you can visualize its voltage regulation, which I’m pretty sure it does a much better job of in real life, but at least the simulation works.

Robert_Kidd

5 years ago

That’s great. Thanks.

Issacsutt

5 years ago

Yeah no problem