PWM! That's awesome. The postman brought me a 120mm axial computer fan today so I've got that on it now, it starts slowly at 35.5c and if the temperature keeps going up it speeds up until it's going at full speed at 40c.
Now you mention it I can hear the PWM in the motor windings when it's going slow. It's quite a low frequency as PWM goes. If it's not going to overheat anything I'll leave it as it is. I've got the TIP32 transistor on a tiny heatsink and it's barely warm, same as the 1w base resistor.

 

If you have the time, try the two suggestions at the bottom of post #40 (one at a time). I'm curious as to why it is happening, and if either of those suggestions would stop it?

 

I tried the capacitor on pin 5 and it made no difference. R4 is 3.3k I'll change it to 4.3k...

 

No, it makes no noticeable difference. What does snap it back into full speed mode though, is if I disconnect the negative from the PSU and reconnect it while the fan is running. Also I've always had a 4001 diode across the motor.

 

I've got a lamp pointing on the sensor so it slowly warms up, it's in a plastic tube with my digital thermometer thermocouple against it. The difference between cut in and cut out temperatures is almost non existent, like 0.1c so it's switching on and off every 30 seconds or so. R1 and R2 are 56k at the minute.

 

Thanks for the feed back. In circuits I have build with the 555, I have never found the capacitor on pin 5 to do anything, but it is shown on the data sheet.

Again, if you are curious as I am, and have the time, make R4 even higher, like 5.6K or 6.8K or even 10K.

 

To stop it cycling as fast I could insulate the sensor I suppose.

 

6.6k in there and the differential is about 3c. The fan sometimes starts up full speed and sometimes in PWM mode. Somtimes it starts slow and then jumps to full speed mode. Cut in is about 35c cut out 32c.

 

10K will likely stop the chattering, but increase the cutin to cutout differential.

Another thought: you do have the 10uF bypass directly across the 555 supply pins?

 

Yes it's all on veroboard and the cap is across pins 1 and 4, not where the supply goes onto the board, if that makes any difference.
It's mostly running fast now. Will it be ok in chattering mode? I quite like how quiet it is and it still moves the air about. multimeter was showing 7-8v when going slow.

 

Close to the 555 pins as possible is best. If it chatters, and nothing is getting hot, then you are good to go...

 

I've been experimenting putting bigger and bigger values in for R4 and it's not getting any better, R4 is now 9.5k and it's a 50/50 chance wether it starts fast or in PWM mode. When it does start slow it soon goes off again.

 

Gone back to 6.2k and most of the time it's starting fast. Seems like there's another factor at play here. Maybe about 5k will be the sweet spot?

 

One more idea about taming that beast: The leads going to the sensor could be acting as an "antenna" and picking the fields from the motor. The network consisting of R2, R3, and R4 is high impedance, so sensitive. Try putting a 100nF to 10uF capacitor between 555 pin6 and pin 1.

I'm really interested in the feedback if it works.

 

With a 330nf between pun 1 and 6 there is only 1 slight difference. Just as the transistor starts pulsing, the motor won't start for about 3 seconds. I can hear the windings 'rattling' and the blade twitches but it doesn't start until the temperature rises more...
Alas. Play time is over until Monday now. We have to visit in-laws in Glasgow... Yawn.
Have a nice weekend.

 

Hi again, I finished the circuit and the fan was quite noisy at full speed so I ordered another and wired them in series. This stopped the PWM in it's tracks, it always switches on and off cleanly. It's been working great in the van ever since.

Now the reason I'm posting is I've built a new circuit for a friend and the slow running 'pwm' problem returned. I've been faffing about for 2 days until I accidentally discovered (by putting a resistor in wrong) that if I make R2 10k lower than R3 it works flawlessly with 2 fans in parallel. So R2=37k, R3=47k.
Also I've got a 100nf capacitor on pin 2 to ground, that also made it run smoother before I changed R2.