| Electronics > Projects, Designs, and Technical Stuff |
| Troubles powering PWM cooling fan from solar panel |
| << < (3/4) > >> |
| David Hess:
It is not uncommon for a controller like those found inside of a DC brushless fan to fail to start if the supply voltage is raised slowly enough. Startup circuits in linear ICs sometimes have the same problem. The best solution is to use a different fan which does not suffer from this problem. If I could not do that, then I would probably use an astable oscillator to periodically short the input supply voltage if the tachometer output is not pulsing. |
| Prehistoricman:
--- Quote from: David Hess on March 23, 2020, 07:28:53 pm --- If I could not do that, then I would probably use an astable oscillator to periodically short the input supply voltage if the tachometer output is not pulsing. --- End quote --- That's actually a really nice improvement of my idea. Characterise the fan speed across voltage and if the fan doesn't spin at the right speed for long enough, reset it. |
| valley001:
As usual, this forum delivers. Cheers everyone. The fan seems to have no trouble starting so long as the voltage is there, im hoping these little buck converters will do the trick, the other ideas are slightly above my skill set..oscillators and such. I do have a non PWM fan on standby in case fixing the voltage dropout does not solve things. Maybe ill put my scope on those other two leads and see what they are doing. |
| valley001:
--- Quote from: duak on March 23, 2020, 05:23:43 pm ---Valley, does the fan do the same thing if you reduce the voltage while operating if from a variable supply? If it's a four wire fan, there might be some sort of digital logic inside that gets upset if the the voltage dips. If applying other voltages to the control pin doesn't help does pulsing it make a difference? That fan should produce a pulse output for the MB to determine RPM. Does it make a difference if the RPM out goes into the speed input? You might need a transistor to invert the signal. --- End quote --- I dont have a low voltage supply to test this, but varying the voltage from the panel by controlling the amount of light seems to accomplish the same idea. If voltage drops to 6v the fan will lock into idle mode. |
| David Hess:
--- Quote from: Prehistoricman on March 23, 2020, 08:51:24 pm --- --- Quote from: David Hess on March 23, 2020, 07:28:53 pm --- If I could not do that, then I would probably use an astable oscillator to periodically short the input supply voltage if the tachometer output is not pulsing. --- End quote --- That's actually a really nice improvement of my idea. Characterise the fan speed across voltage and if the fan doesn't spin at the right speed for long enough, reset it. --- End quote --- The practical implementation is to implement a low frequency 2 transistor astable multivibrator which drives another transistor to short the input. Then the tachometer output is wired to reset the multivibrator so it never switches while the fan is operating. Or the multivibrator could be replaced with a relaxation oscillator of some sort. Essentially the idea is to make a watchdog circuit for the fan and actually you could use processor supervisor ASIC to do it. |
| Navigation |
| Message Index |
| Next page |
| Previous page |