Fan Speed Control

[ajriding]

Im looking for a fan speed control.
Please dont post your internet search results here. Anyone, including me, can do a search and find mindless stuff. I am looking for specific knowledge, thanks

The Fantastic fan has 3 speed. High, but the Low and Med are regulated with a heating element, so running low save no electricity, and might even use more.

I would like to wire in a speed control that uses less energy at lower speeds (that doesn't just burn off electricity in the form of heat to reduce current to the fan).

I have a PWM controller that worked fine. I let it run all night on a low setting with no issues, and today I left it on a few hours on a higher speed (less than full though), and went back to find it burned out, and hot. The default setting if something happens is to just put out full power, so the fan was running on high.
When it worked it makes a ringing sound at the fan.
I cannot find the exact match, but it was similar to this one on eBay, same case at least.

IDK why it burned out, it is rated at 8 amps, and the FF is about 2.

So, Im trying to figure out what controller will do what I want.

I have another controller, but it has a really big heat sink, so I assume it will waste power making a lot of heat.

[2112]

The one I had did make noise, and I have heard others mention the ringing noise from the pulses, so might not be a bedtime item...

I'll touch on this: All PWM controllers have a fixed operating, or switching frequency. Let's say it's operating frequency is 1000 Hz. At that speed it will turn on and off every millisecond. The time period it is on during that millisecond is called duty cycle. At half speed (50% duty cycle) the fan has current flowing through it (turned on) for.5 milliseconds and no current flowing through it (turned off) for .5 milliseconds. At 3/4 speed (75% duty cycle) the fan has current flowing through it (turned on) for.75 milliseconds and no current flowing through it (turned off) for .25 milliseconds.

The noise you hear often comes from the fan turning on at a too low switching frequency. The human ear can normally hear from about 20hz-15khz. I can't hear anything above 8kHz because I'm old. Some controllers will have a switching frequency from 500Hz to 10kHz. You can hear that. A better controller for a fan or motor will have a switching frequency of 20kHz or higher, above the human hearing range. That's why I said the one you found on EBay SHOULD not be noisy because it operates at 25kHz.

However, we are dealing with a mechanical device, the fan. Its blades may not be balanced, bearings may be loose, may not be aligned, etc...
When you change its speed you may fall into a resonant frequency to where it just starts vibrating to the point of making noise. This is not a result of the controller, it's the fan not operating smoothly at a certain speed. It would be hard to overcome that.

Sorry for carrying on but again, I hope this helps you understand what's under the PWM controllers hood.

[frankwp]

I know you don't want links, but you're going to have to put some effort into this yourself because we don't know what your precise requirements are. You'll find a good selection here:
Aliexpress

Be aware that current ratings are frequently overstated & choosing something that is rated for 2x the fan current would be reasonable.

[frankwp]

Duplicate post

[frankwp]

2112 wrote:

a heating element is a resistor, thats why it works. The FF uses these "things" to waste electricity.

A heating element is designed to produce a certain amount of heat at a certain voltage. They are generally made of materials that can glow red hot without failing. It does have a defined amount of resistance to work properly but generally they are not used in a circuit to serve as a resistor. Think of the heating element in your electric water heater. It has about 10 ohms of resistance but why would you use that as a 10 ohm, 1000 watt resistor? Plus, it would have to be submerged in water or it will cook itself and fail in seconds.

The air core wire wrap resistors on your fan were designed to be used as low resistance, high wattage resistors. You could probably use them as a heating element if you stay within there design parameters. Their datasheet would define their thermal characteristics and other parameters you would have to adhere to. There are other resistors they could have used that look more like resistors. They probably made the ones they used in-house for cost savings.

Short story: I was designing a 3-phase brushless motor controller years ago that required exactly 40A peak current. I needed a precise milli-ohm level current sensing resistor capable of passing 40A +margin in a small footprint. Tolerance stack-up of other components required this resistor to be tuned. I used large gauge enamel coated wire wrapped similar to your resistors. It may have looked like a heater element but its purpose was to create a precise amount of resistance.

I hope this helps

Good explanation. Those coils, in the context of the FF controller are most definitely not heating elements.

[2112]

a heating element is a resistor, thats why it works. The FF uses these "things" to waste electricity.

A heating element is designed to produce a certain amount of heat at a certain voltage. They are generally made of materials that can glow red hot without failing. It does have a defined amount of resistance to work properly but generally they are not used in a circuit to serve as a resistor. Think of the heating element in your electric water heater. It has about 10 ohms of resistance but why would you use that as a 10 ohm, 1000 watt resistor? Plus, it would have to be submerged in water or it will cook itself and fail in seconds.

The air core wire wrap resistors on your fan were designed to be used as low resistance, high wattage resistors. You could probably use them as a heating element if you stay within there design parameters. Their datasheet would define their thermal characteristics and other parameters you would have to adhere to. There are other resistors they could have used that look more like resistors. They probably made the ones they used in-house for cost savings.

Short story: I was designing a 3-phase brushless motor controller years ago that required exactly 40A peak current. I needed a precise milli-ohm level current sensing resistor capable of passing 40A +margin in a small footprint. Tolerance stack-up of other components required this resistor to be tuned. I used large gauge enamel coated wire wrapped similar to your resistors. It may have looked like a heater element but its purpose was to create a precise amount of resistance.

I hope this helps

[wa8yxm]

Ok, how do PWM converters work and how do DC-DC converters work?
PWM's are ON/OFF there is no "Half on" the ratio of ON to OFF is what decides speed. as a result they dissipate very little power themselves. So they don't get all that hot.

DC/DC converter come in two types. buck only and buck boost.. I suspect the buck only are PWM with a filter cap on the output... Again very very very low power dissipation Don't mean they should not feel a breeze.. just means they do not get all that hot.

[ajriding]

2112 wrote:

The PWM things have big heat sinks, so are producing heat.

Most everything electrical dissipates some wattage in the form of heat, even if it's 1/10th of a watt. There is no free lunch here. It can be such a small amount that the properly designed component in a properly designed circuit doesn't heat up enough to notice, if at all.

The MOSFET transistors have heat sinks because they are capable of passing 15A at 95% duty cycle. That's probably about 8 watts. The controller is designed to handle worse case. But you are only drawing 3 amps or less. The transistors might dissipate a watt or less with the fan running full speed. When you slow it down that wattage dissipation goes down. That's the beauty of a PWM controller. If your knob is turned down half way you are using 50% power, creating less heat.

Thanks for the answer... This is more the discussion I am interested in. So, the heat sink on a PWM is to keep the electronics cool, not to burn off heat in any way to reduce fan speed...

Sounds like the pwm is the way to go for efficiency at least.
The one I had did make noise, and I have heard others mention the ringing noise from the pulses, so might not be a bedtime item... At night I can just use a clip fan.


Also, a heating element is a resistor, thats why it works. The FF uses these "things" to waste electricity. I don't have an amp meter to test how much though.

[2112]

The PWM things have big heat sinks, so are producing heat.

Most everything electrical dissipates some wattage in the form of heat, even if it's 1/10th of a watt. There is no free lunch here. It can be such a small amount that the properly designed component in a properly designed circuit doesn't heat up enough to notice, if at all.

The MOSFET transistors have heat sinks because they are capable of passing 15A at 95% duty cycle. That's probably about 8 watts. The controller is designed to handle worse case. But you are only drawing 3 amps or less. The transistors might dissipate a watt or less with the fan running full speed. When you slow it down that wattage dissipation goes down. That's the beauty of a PWM controller. If your knob is turned down half way you are using 50% power, creating less heat.

[frankwp]

ajriding wrote:
mindless is when people just look up "speed controller" and post a bunch of links, none that are relevant usually.

That Guide to put a controller in is nice, but not anything to do with what I want - to have a controller that uses less energy at lower speeds.
I have my controller feeding the FF, so have control on a wall, not at the fan. I can always put controller at max and run the FF normally. I set FF on High and control the speed by using the controller.

The PWM things have big heat sinks, so are producing heat. IDK beyond that, so am asking any electrical geeks out there who might know

You seem to be contradicting yourself here. The controller in the linked guide is a PWM, as is your controller that burned out. It's about the most efficient speed controller for your fan that you will find. The reason for the big heat sink is so it won't burn out like your old one. You can still mount it remotely from the fan & bypass the local fan switch, or just leave it on high.

And those coils are not "heating elements", they are resistors.

[ajriding]

2112 wrote:

What happens when battery power is too low to move the fan and the PWM is sending out power still? Harm to fan? Harm to controller?

That could be a problem. A quality controller would use a low voltage dropout regulator to protect itself from this. The one you pointed to can regulate down to 10 volts. It could heat up below that. You will have some line loss depending on what wire you use. A battery voltage below 11.5V (just my guess) might create 10V at the controller.

Why not try the one you pointed to on eBay? It's $13 delivered. 25KHz switching frequency so it shouldn't be noisy, rated for 10A and has a power switch. I assume you already have the wires located where you want it on the wall. Did you run 16awg wire? Wires being too small/long could be a problem.

The eBay link one is the closest to what I have, same plastic case (no vent holes), so it will fit right in.
I will have to go investigate what wire. Most of the wire run is big wire, and I do not own any smaller than 18g, but if it is 18 then it is just a small section... However, running fan on a low output speed my guess is 18g wire is not coming into play here.

What Im worried is that I forgot and left the fan on overnight, on a low speed, just fast enough to keep running, and maybe the batteries, 7 yrs old now, got a little lower over night and the fan stopped running, but was getting power sent to it, and sent through the controller still.
I know from audio that when this sort of thing happens to a speaker you can melt the voice coils, low-power can hurt a speaker more than too much. (weak signal to a two-way speaker and the smaller speaker takes all the power, not the big speaker... I wont get into details).
So, I am not sure if it was my fault and I melted something with a non-running fan, or it the controller just died naturally. I dont want to blame the vax if it was truly a natural death sort of thing...

[wa8yxm]

There are two ways to control the fan speed on a DC fan
One is to use a resistor. this has been the way it has been done traditionally since at least the 1950's or before. (I only go back to the 50s)

The other is a "Buck" converter. These are DC-DC converters. they swap amps and volts just like a transformer.. I've never tried them but ALL ELECTRONICS dot com has a selection of them at a decent price. What size fuse is in the fan? Get one at lest that many amps.

[2112]

What happens when battery power is too low to move the fan and the PWM is sending out power still? Harm to fan? Harm to controller?

That could be a problem. A quality controller would use a low voltage dropout regulator to protect itself from this. The one you pointed to can regulate down to 10 volts. It could heat up below that. You will have some line loss depending on what wire you use. A battery voltage below 11.5V (just my guess) might create 10V at the controller.

Why not try the one you pointed to on eBay? It's $13 delivered. 25KHz switching frequency so it shouldn't be noisy, rated for 10A and has a power switch. I assume you already have the wires located where you want it on the wall. Did you run 16awg wire? Wires being too small/long could be a problem.

[theoldwizard1]

QCMan wrote:
You will find that just about every fan speed control uses resistors, ...

What the OP wants is a pulse width modulated (PWM) speed control. It is the most "optimal". Very little energy is wasted (turned into heat).

[ajriding]

Hmm, ok. I guess how hot the coil resistor in the FF gets determines how much waste there is. IDK.

The PWM I have has no holes for ventilation to move the heat out.
What happens when battery power is too low to move the fan and the PWM is sending out power still? Harm to fan? Harm to controller?

I guess I will just get a PWM but one with vent holes this time.

[DrewE]

A PWM circuit will use less power at lower speeds.

The original resistor will, as well, but is not as efficient as the PWM units. (Adding the resistance into the circuit decreases the current, per Ohm's law, which means less power consumed.) Ideally the switching devices in a PWM circuit will dissipate no power, since they are either fully on or fully off with no time spent in between where they act as a resistive divider with the load. For practical circuits, that's not quite achievable since semiconductor switching times are not truly instantaneous and since the on state has a bit of impedance, but it gets quite close.

The unit you had that failed was most likely some sort of a PWM controller. It may well have failed due to the lack of a big enough heat sink, allowing the switching transistor to overheat.