555 timer PWM problem.

[ZeTeX]

Hi, I have built a 555 pwm motor controller and there is a small problem with it:
it adjust fines from 100% to about 20-30% and then then jumps to like 80-90%
I tried adding resistence to both sides of the potentiometers but it didnt work.
here is the schematic: (the only changes I did was changing the npn to n channel mosfet (IRF3415) and insted of 1k base limiting resistor I put 100ohm.

[Audioguru]

Your circuit is missing an important capacitor across the power supply pins of the IC. ALL electronic circuits need it because it cancels the inductance of wires from the battery to the IC. The capacitor should be very close to the IC and have very short wires which is impossible on a wireless breadboard. Try a 0.1uF ceramic and try adding a 10uF electrolytic parallel to the 0.1uF ceramic cap.

The 1k resistor will cause a very high current in one of the very old and fragile 1N34A diodes if the pot is turned to either end. Try 10k instead then reduce the value of the timing capacitor 10 times to match it.

Maybe the very high current burnt a spot on the track of the pot. Replace the pot after reducing the current.

[ZeTeX]

Your circuit is missing an important capacitor across the power supply pins of the IC. ALL electronic circuits need it because it cancels the inductance of wires from the battery to the IC. The capacitor should be very close to the IC and have very short wires which is impossible on a wireless breadboard. Try a 0.1uF ceramic and try adding a 10uF electrolytic parallel to the 0.1uF ceramic cap.

The 1k resistor will cause a very high current in one of the very old and fragile 1N34A diodes if the pot is turned to either end. Try 10k instead then reduce the value of the timing capacitor 10 times to match it.

Maybe the very high current burnt a spot on the track of the pot. Replace the pot after reducing the current.

Sorry, I guess I forgot some information.
1. I have 0.1uF X7R ceramic capacitor DIRECTLY on the 555 pins, they are as close as I can get, the distance of the leads are 1mm, also I'm filtering the +12VDC rail with 100uF 50V panasonic cap.
2. this is soldered on a perfboard.
3. I'm using "fast rectifier" diode 1n4937.
4. the pot is working correctley.

[alsetalokin4017]

Well, something in your circuit is not working correctly, because I have the same circuit assembled on a breadboard right now and it's working properly.

I'm using 1n4148 diodes, and my "100k" trimpot only measures 82k, but the circuit also works fine with a 50 k pot. Also tried it with a 100 nF timing cap instead of 10 nF and it works fine with either pot using that capacitor as well. Duty cycle is smoothly adjustable from under one percent to over 99 percent (measured at Pin 3). I'm driving a IRF3205 mosfet with a 47R gate resistor, with an incandescent lamp load, and also tested it with a computer fan (but the fan drops out at duty cycles less than about 18 percent, not surprisingly).
 
I even tried it with plain old 1n4004 rectifiers and it still works full range, although not quite as well at very short (under 5 percent) duty cycles.

I would also suspect a burned pot in your circuit.

ETA: what version of the 555 are you using? I'm using NE555P.

[alsetalokin4017]

@audioguru: I just measured the average current through my 1n4148 diodes and it is on the order of 55-70 microamps, even at the extreme ends of the pot travel.

ETA: Even if the diodes went directly to ground, and neglecting the Vf of the diode, the maximum current through them and the 1000R resistor would be less than 12 mA. (I=V/R = 12/1000 = 0.012). And even a 1n34a is rated at 50 mA. Am I missing something here?

[ZeTeX]

Well, something in your circuit is not working correctly, because I have the same circuit assembled on a breadboard right now and it's working properly.

I'm using 1n4148 diodes, and my "100k" trimpot only measures 82k, but the circuit also works fine with a 50 k pot. Also tried it with a 100 nF timing cap instead of 10 nF and it works fine with either pot using that capacitor as well. Duty cycle is smoothly adjustable from under one percent to over 99 percent (measured at Pin 3). I'm driving a IRF3205 mosfet with a 47R gate resistor, with an incandescent lamp load, and also tested it with a computer fan (but the fan drops out at duty cycles less than about 18 percent, not surprisingly).
 
I even tried it with plain old 1n4004 rectifiers and it still works full range, although not quite as well at very short (under 5 percent) duty cycles.

I would also suspect a burned pot in your circuit.

ETA: what version of the 555 are you using? I'm using NE555P.

I'm also using a brand new NE555P, bought from my local electronic store.
I accidantly shorted the reset pin to the trigger pin, but I fixed it. could it maybe damaged the 555?

[alsetalokin4017]

Well, something in your circuit is not working correctly, because I have the same circuit assembled on a breadboard right now and it's working properly.

I'm using 1n4148 diodes, and my "100k" trimpot only measures 82k, but the circuit also works fine with a 50 k pot. Also tried it with a 100 nF timing cap instead of 10 nF and it works fine with either pot using that capacitor as well. Duty cycle is smoothly adjustable from under one percent to over 99 percent (measured at Pin 3). I'm driving a IRF3205 mosfet with a 47R gate resistor, with an incandescent lamp load, and also tested it with a computer fan (but the fan drops out at duty cycles less than about 18 percent, not surprisingly).
 
I even tried it with plain old 1n4004 rectifiers and it still works full range, although not quite as well at very short (under 5 percent) duty cycles.

I would also suspect a burned pot in your circuit.

ETA: what version of the 555 are you using? I'm using NE555P.

I'm also using a brand new NE555P, bought from my local electronic store.
I accidantly shorted the reset pin to the trigger pin, but I fixed it. could it maybe damaged the 555?

Pin 4 to Pin 2? I just tried it on my breadboard (briefly)  and it didn't seem to hurt my 555. Of course it stops working when the short is there but when the short is removed it works again.

ETA: OK, I just tried it again with a 1- minute continuous short between Pins 2 and 4. No effect on the 555, I don't think it even got warm. When short is removed it works again, smoothly from <1 percent to >99 percent duty cycle.

[ZeTeX]

Well, something in your circuit is not working correctly, because I have the same circuit assembled on a breadboard right now and it's working properly.

I'm using 1n4148 diodes, and my "100k" trimpot only measures 82k, but the circuit also works fine with a 50 k pot. Also tried it with a 100 nF timing cap instead of 10 nF and it works fine with either pot using that capacitor as well. Duty cycle is smoothly adjustable from under one percent to over 99 percent (measured at Pin 3). I'm driving a IRF3205 mosfet with a 47R gate resistor, with an incandescent lamp load, and also tested it with a computer fan (but the fan drops out at duty cycles less than about 18 percent, not surprisingly).
 
I even tried it with plain old 1n4004 rectifiers and it still works full range, although not quite as well at very short (under 5 percent) duty cycles.

I would also suspect a burned pot in your circuit.

ETA: what version of the 555 are you using? I'm using NE555P.

I'm also using a brand new NE555P, bought from my local electronic store.
I accidantly shorted the reset pin to the trigger pin, but I fixed it. could it maybe damaged the 555?

Pin 4 to Pin 2? I just tried it on my breadboard (briefly)  and it didn't seem to hurt my 555. Of course it stops working when the short is there but when the short is removed it works again.

what could it be? 

[alsetalokin4017]

Try a different potentiometer.

[ZeTeX]

Try a different potentiometer.

I'll try other 100k pot that I have. will edit this comment after.

[alsetalokin4017]

I can't really think of anything else that could be causing your symptoms, except maybe something wrong with the C1 capacitor.

Can you post clear photos of both sides of your assembly? Maybe having a "third eye" take a look at it might reveal something that could be causing the problem.