Help Deriving a Circuit

[driftlogic]

@QuietMan,

The 4017 Chip used in post #59 is a little different than the one in #53. I see that it hooks to the Vcc. Can it support the voltage. Is this chip better?

EDIT: Also, was just messing around in MultiSim. I recreated the circuit from Post #53... In my simulations the LEDs didnt flash , and a voltage of >2.7 seemedd to cause pin 3 (output) of the 555 chip to show an animated explosion. Not sure what's going on.

 

[QuietMan]

It is the same chip. I just showed all the connections, they left the power supply pins off the schematic. When I draw op amp circuits and some logic I tend to do the same thing, but it isn't a good thing.

You can add as many LEDs in parallel as you want. Each LED is around 700ma, so round off to an amp per LED. Most MOSFETs are major overkill for this application. This circuit as it stands will pull around 2.1A when the LEDs are on.

I am positive about the 555 circuit working, what did you do with pins 5 and 7 on your sim?

I can't help you much on the LEDs, though I'll go back and look at the links. With these suckers it is very much a case of buyer beware.

Putting the LEDs in series as shown in post #53 will dramatically reduce the power requirements of the resistors. It also makes the circuit a bit more power supply dependent, but it will work well.

 

[driftlogic]

Ok I understand what you're saying about the 4017.

(This is redo of the circuit in #53, not your version 1.1 just to be clear.) Pin 5 (control) was not connected to anything and Pin 7 (discharge) was wired like before. You can see below in the schematic the simulation explodes Pin 3 (output) for some reason... Up to voltage = ~2.7 it doesn't explode but nothing with the LED's happen. Here is the schematic as drawn in MultiSim (showing the burnt Pin 3):
**broken link removed**

EDIT: I see the R6 value is wrong, but it doesnt change the simulation.

 

[QuietMan]

They are showing the 555 chip is blown. Since I don't use simulators all I can say is my circuit works, it may be that R1 is too low by the simulators standards, pin 7 switches to ground and can be current sensitive. Part of the reason for my redo is the drivers for the MOSFETs are seriously flawed, a MOSFET is paradoxical in that it needs a really high current feed, but draws almost no current. This is because the speed you turn a MOSFET on/off is critical, and the gate looks very close to a capacitor (say 0.1µF). So you need something that can provide a quick surge of current feeding them. A large resistance such as R3 and R4 can cause unexpected issues, which is what I'm trying to address.

Try using my design, and replace R1 and R2 with 10KΩ, it will oscillate at 7Hz.

 

[driftlogic]

Okay I will simulate your design (with the changes of R1 and R2) and post back the results.

 

[driftlogic]

Okay I simulated version 1.1 and the 555 blew at the same pin. Also in recreating v1.1 i noted that pin 7 of the 555 (discharge) was not wired to anything unlike the circuit in post #53. Below is the schematic of 1.1 that I redid. Apologies for the cluttered wiring, and poor numbering of components. I think it is wired the same though as what QuietMan showed in his v1.1 circuit: **broken link removed**

 

[QuietMan]

You are getting tired. R2 and R11 are 10Ω, not 10KΩ.

Now you see how version 1.0 was born.

 

[carbonzit]

I don't think that's why the 555 is "blowing up". How about that lone LED with only a 10Ω resistor; isn't that drawing excessive current? (I take it that's for debugging purposes?)

 

[driftlogic]

Yes you are right I am tired .

...R2 and R11 are 10Ω, not 10KΩ.

I changed them to 10K ohm but the simulation pin blowing is the same. I do trust that your circuit drawn as it is would work however. I just thought it would be nice if the simulation would work though, so I could see it in action!
As for the Lone LED with the 10Ohm resistor... I'm not sure.

 

[carbonzit]

Well, take it out and see. It's not part of the project anyhow, right?

 

[driftlogic]

Took it out. Same blown pin. The pin does not blow when the Voltage is somewhat less than 3V. But the LEDs dont do anything...

 

[driftlogic]

I have simulated the circuit with other software and it does work as expected.

EDIT: Oddly in the simulation the switch did nothing until I moved it to be immediately after where the 555 chip got power. When placed here the LED banks will turn off but that LED off the 555 still blinks (not really needed though).

 

[QuietMan]

If you look at my schematic there is not one 10Ω resistor. The lowest LED is 15Ω, for the high power, while the small LED on the output of the 555 (R3) is 1KΩ. You are trying to pull almost an amp out of the 555, when with my schematic it is around 10ma.

 

[driftlogic]

Are R4 and R8 not 10 Ohm?

 

[QuietMan]

You got me, but why 10Ω around the 555? It is why your simulators aren't working, you are burning out the chip. You could eliminate R4 and R8 and probably not notice. They are there to improve the switching characteristics of the MOSFETs.

 

[driftlogic]

^ , I had actually replaced those 10 Ohms with 10 K Ohms but the simulation in Multisim refused to work. I did get it to work in another program, but then it crashed and now i keep getting "timestep too small" error on the simulation.

Anyway, I think the simulator is not too important right now. I think using your version 1.1 is the way to start going now. I notice you used different value resistors for the Red LEDs vs the Blue LEDs, does that reflect the specs of the LEDs I linked to, in my case the Red and White?

 

[QuietMan]

I've been putting some thought into the current limiting resistors, and basically have come to the conclusion that several LM317 would probably do the job better. It is an odd fact that overall transistors and power ICs are easier to heatsink than resistors.

Looking at your designs in the PMs if you don't mind putting 2 of the LEDs in series it would definitely cool things down.

I'll draw something up and post it up here within 24 hours.

Change of subject, my current PCB generating program is PCB Express. It is free, if limited, but would be more than adequate for this design, or the PIC design you are contemplating. You can download it for free at their website, and when I have something made I can post the files on this thread where you can look at them.

 

[QuietMan]

OK, my latest revision. I'm going to start translating this to a PCB.

If you have any questions feel free to ask. Basically I'm working on reducing the amount of heat in the box. R5 - 11 and R12 - 18 are 7 resistors in parallel, 120Ω ¼W.

 

[dougy83]

you seem to have your OUT & ADJ pins backwards

 

[dougy83]

R5 - 11 and R12 - 18 are 7 resistors in parallel, 120Ω ¼W.

There's only 1.25 volts across that resistor, so @ 72mA, that's 0.091W - you could just use a single 18 ohm 1/4W or 1/8W if you wanted. The heat will mainly be dissipated by the LM317.