OK, this will go down to 4.5VDC, and will substitute for the other oscillator. A negative is the square wave isn't anymore.

I found this in the AAC experiments, believe it or not. Substitute a current mirror for the resistor, it becomes linear!

Yep - but Bill, try splitting that R1 into two resistors, and connect pin 7 to the junction of the two resistors, keeping 6 and 2 on the cap.

Then connect a diode, anode towards pin 7, across the lower resistor.

This way, C1 charges via the upper resistor and the diode, and discharges via the lower resistor.

You can get just about any duty cycle you want from that circuit, although the upper resistor can't go lower than 100 Ohms per volt of Vcc. It also has the advantage that it is independent of the load on the output (pin 3). At low Vcc, the current source/sink capability of pin 3 is greatly reduced.

 

Actually it will still do a full PWM signal (as in any duty cycle), the single side will do it all. We had this discussion elsewhere concerning this concept.

http://forum.allaboutcircuits.com/showthread.php?t=12405

I like using minimum parts, and I do mean minimum. If you want straight edges the current mirror works well too.

Upon review, I am positive the original schematic will work.

If the base is connected to ground (the output of the LM339 is taken to ground) Q1 will be fully on, minus BE drop. If the output of the LM339 is off (floating), Q1 will be off, as a common collector design requires some current to turn on. Without the Base to Vcc resistor there is some potential for instability to noise, but not much.

 

Upon review, I am positive the original schematic will work.

If the base is connected to ground (the output of the LM339 is taken to ground) Q1 will be fully on, minus BE drop. If the output of the LM339 is off (floating), Q1 will be off, as a common collector design requires some current to turn on. Without the Base to Vcc resistor there is some potential for instability to noise, but not much.

Funny thing - when I glanced at it last time, somehow I thought Q1 was NPN - go figure.

Have to say though, that without a pull-up resistor on the output of the 339, the PWM out signal will be pretty miserly... if your Vcc is 10v, and you have a 250k Ohm load from the base of Q1 to ground as your PWM output load, you'll get a maximum output of 5v for the PWM signal. The more the load on the PWM output, the more it looks like the 339 output is turning on.

 

OK, I see where you're coming from. I tend to think of the circuit from the LED's point of view, a pull up resistor makes no difference either way. If you're looking at it from the PWM signal point of view, it makes a large difference.

 

There, got #1 made on a breadboard, and added a power transistor - seems to work great. However, at lowest duty cycle, i still get around 2V RMS. Since im running the circuit at 12V, that suggests around 17% duty cycle.

 

Did you use the revised schematic? The one with R8?

 

Did you use the revised schematic? The one with R8?

R8 ??? No idea what youre talking about

 

LEDs, 555s, Flashers, and Light Chasers

The 555 and PWM

I've updated it a bit.

 

LEDs, 555s, Flashers, and Light Chasers

The 555 and PWM

I've updated it a bit.

All i get from those two links, are the top of the page.. would be easier with a direct link til to actual schematic.
I used: http://forum.allaboutcircuits.com/picture.php?albumid=45&pictureid=580

In that one, there is no such thing as R8.

 

There, got #1 made on a breadboard, and added a power transistor - seems to work great. However, at lowest duty cycle, i still get around 2V RMS. Since im running the circuit at 12V, that suggests around 17% duty cycle.

You need to post a schematic of YOUR circuit, showing all values of caps, resistors and part numbers for transistors, since it is not exactly the same as what Bill_Marsden posted.

 

Hi everyone

Im trying to make a general purpose PWM circuit. Ive been using the 555 before, so naturally i went with that one again.

Problem is, the standard circuit from the datasheet, calculates the duty cycle by Ra/(Rb + 2*Ra), which means the lowest duty cycle will be 50%.

Anyone know a good schematic, with full range (~0% to ~100%) duty cycle?


Im planning to use a TIP35A to drive whatever i want to control - should be more than enough (got a few in spare anyway..)


Thanks in advance.

That is correct, the 555 is not the right circuit for this. You want to use a triangle wave, an adder and a comparator to do this right.

Eric

 

You need to post a schematic of YOUR circuit, showing all values of caps, resistors and part numbers for transistors, since it is not exactly the same as what Bill_Marsden posted.

On it - what do you use to draw schematics? Ive worked with CadInt, Eagle and Orcad PSPICE - i dont really like any of them that good

 

Cadsoft Eagle works great for schematic capture and PCBs; but it doesn't have built-in SPICE.

Linear Technology's LTSpice is good free, and fully supported. You can download it for free from their website. Google is your buddy on this. There are also newsgroups for LTSpice support and obtaining more SPICE models for it, like 4000 series CMOS, TTL, etc. It takes a bit of getting used to, as all such programs do.

Texas Instruments has Tina-TI available for a free download. Many people like it. I personally find it rather awkward, but in all fairness I have not used it very much. It's hard enough just trying to keep up with what else I use.

Several on here simply use MS Paint to draw schematics. Bill_Marsden has uploaded a zip file that he calls "PaintCad"; it consists of a bunch of symbols that you can use to quickly copy from the library and paste into your schematic. Obviously, there is no spice simulation, but it's much more convenient than trying to draw a schematic from scratch.

I frequently use an obsolete and unsupported program called Circuitmaker Student for "quick and dirty" simulations here. It will not run on Windows Vista, or anything past Windows XP. The component library is limited, and you cannot add to it. Schematics are limited to 50 components. I just use it because I find the interface intuitive, but it is a dead end product.

 

So, there it is. I couldnt find an exact model for the TIP35A, so i used whatever there was.
I found a 100n instead of the 2n - seems a tiny bit better. But i gotta go find my oscilloscope to get a better view.

 

TIP35A transistors are good for controlling high currents, but they have very low gain.

Let's figure out your TIP35A's base current.
Your Vcc is 12v. Assume that the 555's maximum Vout at pin 3 under light load will be Vcc-1.3v, or about 10.7v.

Assume for the moment that Vbe of the TIP35A will be about 0.7v.
10.7v - 0.7v = 10v to drop across R2.

R2 is 1k Ohms.
10v / 1k Ohms = 10mA current flow from the output of the 555 through the base of the TIP35.
The standard formula for transistor saturation is Ic=Ib*10, or Ib=Ic/10.
So, 10mA base current will give you the capability to drive a 100mA load with your TIP35A's, and have the saturation voltage (Vce) low.

If you try to sink more current, the transistor will come out of saturation, and Vce will be higher.

555 timers can source or sink up to 200mA from pin 3 when supplied with 15v. However, using pin 3 as the charging source/sink for the timing capacitor means that you really need to keep the 555's output load much less than what it is rated for.

If you want to virtually eliminate the problem of the high base current requirement, you should consider using enhanced-mode power MOSFETs.

Enhanced power MOSFETs are basically voltage controlled switches. If you want to turn them on, you charge the gate terminal to be 10v higher than the source terminal (N-channel standard MOSFET, 5v for logic level MOSFET). If you want to turn it off, you discharge the gate so that it is equal to the source terminal.

Enhanced MOSFETs are very different from bipolar junction transistors.
However, if you want a very rough analogy:
Mosfet Transistor
Drain = Collector
Gate = Base
Source = Emitter

 

SgtWookie --> What i needed, was to find out to get a lower duty cycle
Atm i just needed something to drive a board with LEDs.

But.. i found my good old oscilloscope, got it powered it, connected it, and.. it killed my 555. Swapped with 555 with a fresh one, and it ran fine. The connector to the oscilloscope fell off - i put it back on, and.. it killed the 555.

I hate old oscilloscopes, i really do.

 

What is the part number of the 555 timers that you are trying to use?

If they are CMOS, you will have a hard time trying to use them with Bill's circuit with that heavy of a load on it.

Old 'scopes are great - you can actually fix them yourself if they break.
The new scopes, you have to pay a great deal of money to have them fixed.

 

JMD, you need a current limiting resistor in series with the LED, or it will probably be destroyed. I can't recommend a value without knowing what type LED you are using.

 

You should have better results with this version:

Note that R1 has a minimum value of 100 Ohms per volt of Vcc.

Otherwise, when R2 is set to a low value on the high side, R1 will dissipate too much power, and the 555 timer pin 7 will have to sink too much current.

D1 provides for rapid charging of C1, but the discharge path is still through R2 to pin 7. This is what gives it a wide range of adjustment.

C1 has been increased to 100nF, or 0.1uF. This will result in an average 150Hz frequency, which should be fast enough. If you want it to be a little faster, you could decrease C1 to perhaps 0.047uF, but I wouldn't go much lower when trying to drive that big slow transistor.

C2 helps a great deal with the large transients caused by the 555 switching output states. This is a value suggested by the datasheets.

C3 is for transient suppression right at the 555 timer. It should be as close to the 555 as possible. You might do without it, but it is good practice to include it.

R3 has been reduced to 100 Ohms, which will give you up to 100mA base current in Q1, with a resulting 1A collector current while saturated. R1 should be rated for 2W.

 

SgtWookie --> Its an 7555CM. Agree, old scopes can be good - actually fixed this scope once (the high-voltage circuit sort of died).

Ron H -> Im not THAT dumb, dont worry - i only placed the LED there, to illustrate where the load were

SgtWookie again --> Looks good, gonna give it a shot when i get some fresh 555's. Transient supression shouldnt be a problem here, after all, this isnt used for something super serious. Keeping overall cost down, is more needed