direct drive of fan with solar panel through buck

Thread Starter

ludovic_pommier

Joined Jan 10, 2012
10
Hi,

I've been googling forever and can't seem to progress. More of a software than hardware guy...

I'm working on a projet where I want to run a 12v fan (PWM) according to temp in room. I'm controlling everything through an ESP32.

The fan (12v/0.5A) is connected to a buck converter (set at 12v) which is connected to a solar panel (32W, open voltage 21v). This works as long as the panel delivers enough energy. When the supply is low (low light), the voltage of the panel drops and the fan run slow. This is expected, but I'd like to stop the fan from running when it can't reach 11.5v.

I've studied comparators and mosfet, but I don't think that what I need. I could also use software (reading fan voltage and adjusting pwm down to zero if unable to get to 12v) but that would mean my ESP32 is always running and I need it to get into deep sleep. If I use a comparator to disconnect load when voltage under 11.5, then voltage shoot up immediately. Even with hysteresis, it seems hard (when no load, the voltage is fairly constant).

Really desperate here. Any idea?

Much appreciated. Thanks in advance. Ludo
 

Papabravo

Joined Feb 24, 2006
21,304
I think you can implement the UVLO (Under Voltage Lock Out) in two different ways:
  1. Have the ESP32 "wakeup" for a couple of milliseconds every minute, check the voltage and go back to sleep if it is still insufficient.
  2. Have the UVLO circuit put the PWM signal in the inactive state and stop it from changing.
 
Last edited:

BobTPH

Joined Jun 5, 2013
9,259
If I use a comparator to disconnect load when voltage under 11.5, then voltage shoot up immediately.
This will be a problem no matter how you sense the voltage.

The only way I can see to get around it is to switch in an equivalent dummy load when the fan is iff. This would keep the voltage from changing.
 

Externet

Joined Nov 29, 2005
2,241
Solar panels are current sources. Their voltage is nearly constant at any light exposure, from cloudy to bright when unloaded.
When loaded with the motor, voltage collapses as the solar panel in uncapable of sourcing the current demanded at weak insolation. Try with several panels in parallel or a bigger, more current capable panel to feed the fan motor in cloudy exposure.
What you call 'buck' manipulates voltage; not current. You should aim to a more current capable solar source as a larger panel to maintain a voltage under load.
Key is choosing a watts capable panel when cloudy, not a volts capable.
 

Thread Starter

ludovic_pommier

Joined Jan 10, 2012
10
I think you can implement the UVLO (Under voltage Lock Out) in two different ways:
  1. Have the ESP32 "wakeup" for a couple of milliseconds every minute, check the voltage and go back to sleep if it is still insufficient.
  2. Have the UVLO circuit put the PWM signal in the inactive state and stop it from changing.
Thanks. I could do that with software and a few components:
My panel is given for 32W max and I'm losing about 20% with the 12v converter. Anyhow, my load is max 0.5A for the fan (12v) and 0.65A for the esp32 (from the 12v through a 5v converter fed into vcc of esp32 to benefit from the on board battery charger that takes most of the amps). So I guess I can approximate my load to 1A at 12v to be on the safe side.

I could have a 12v 20w resistor (if available). A MOSFET to swich it on/off and a resistor divider to read the voltage. If I get 12v, I run the fan, if not I wait. Would that work?

Which mosfet would you use with 3.3v logic?

Cheers
 

Thread Starter

ludovic_pommier

Joined Jan 10, 2012
10
This will be a problem no matter how you sense the voltage.

The only way I can see to get around it is to switch in an equivalent dummy load when the fan is iff. This would keep the voltage from changing.
Thanks. I could do that with software and a few components:
My panel is given for 32W max and I'm losing about 20% with the 12v converter. Anyhow, my load is max 0.5A for the fan (12v) and 0.65A for the esp32 (from the 12v through a 5v converter fed into vcc of esp32 to benefit from the on board battery charger that takes most of the amps). So I guess I can approximate my load to 1A at 12v to be on the safe side.

I could have a 12v 20w resistor (if available). A MOSFET to swich it on/off and a resistor divider to read the voltage. If I get 12v, I run the fan, if not I wait. Would that work?

Which mosfet would you use with 3.3v logic?

Cheers
 

BobTPH

Joined Jun 5, 2013
9,259
Yes, you have got the idea. You need logic level MOSFETs. You will need two, one for the fan and on for the dummy load.
 

ThePanMan

Joined Mar 13, 2020
803
My panel is given for 32W max and I'm losing about 20% with the 12v converter.
I'm wondering how the fan would perform on a direct connect to the solar panel. 12V @ 500mA suggests a load of 24Ω. So 32W @ 21V(no load) suggests a 1500mA capable panel. IF you're drawing 500mA then the voltage will be a bit higher; off hand I don't know what that would be. But the fan may run at a slightly higher RPM and move more air during peak output of the PVP. (Photo Voltaic Panel). I don't see a need to regulate a fan that closely. Just moving air means just moving - um - Air. Little more, little less; and as the day progresses, a lot less. Is that a problem?

I get it. You want to shut the fan OFF when power becomes so low. Low power won't hurt the fan, and in all that extra effort to make a system to either switch the fan hard ON or hard OFF seems like more work than you need to pursue. If it were me I'd opt for direct powering of the fan and not worry about anything else. Motors are VERY forgiving when it comes to voltage.
 

Externet

Joined Nov 29, 2005
2,241
Agree. Direct connecting the solar panel will make the motor work sunrise to sunset with zero complications of adding circuitry. From as soon as daybreak generated current is enough to start until dark.
A nominal 12V panel should have ~24 cells in series. A 12V battery charger panel should have ~26 to 30 cells in series
 

Thread Starter

ludovic_pommier

Joined Jan 10, 2012
10
Hi, thanks all. Overvoltage is taken care of by a buck converter. The fan is a 4 pin pwm and operating voltage 12.5 to 11.5. Just wondering is under voltage would not damage it (apart from it stalling sometimes). Would a 25v capacitor help restarting when stalled. Fan is 0.5 amp. If so, which size?
 

ThePanMan

Joined Mar 13, 2020
803
There's one other option you might consider: a Buck/Boost converter. It'll keep the voltage down to the voltage you want. When the panel is not producing enough it can boost the voltage. At some point it'll not be able to keep up the boost and shut down until the next day. Your fan should run all day and into the early evening or late afternoon, whichever may be the case.

Here's one now:
https://www.amazon.com/Adjustable-Converter-Automatic-Voltage-Regulator/dp/B07NTXSJHB/ref=sr_1_3
 

wayneh

Joined Sep 9, 2010
17,499
You might consider using another small panel or sensor, even a CdS cell, to control the on/off status of the fan. That's how many of those landscape LED lights work; they shut off the LED if the CdS cell sees even a dim light. And of course you can adjust the sensitivity.
 
Top