general question about pwm and brushless fans

[default]

I have a fan I want to use with continuously variable speed. It is a
brush less type (specs on this one say 7-12 volts) should work. And
it does . . . but won't take PWM directly, it wants to see steady
voltage not chopped. I tried frequencies from 20 kHz to 1 kHz

1. Is this typical?

2. Anyone else with experience, of speed control of generic brush
less fans?

I tried my circuit with two other junk box fans and they also ignore
pwm, but like steady voltage for variable speed.

Right now I have my picaxe starting at 50 ms on, and 150 off, (ersatz
pwm) up to steady on, in four ranges - which gives me the variable
cooling I want, but it sounds like it is panting and is moderately
annoying.

Plan B:

I really want to continue using an N channel logic level MOSFET and
don't want to disturb the gizmo that's working and in-service for a
variety of reasons.

Seems to me that since my mosfet is pulling the fan's minus connection
to ground, it should be possible to just turn it into a switching
supply with steady DC output. Hook an inductor to the drain,
freewheel diode to the mosfet drain and Plus, and put the fan (load
minus) to the inductor's other terminal (and electrolytic cap across
the fan).

sort of a switching supply, but upside down and backwards

3. Does it sound feasible, or am I missing something?
--

 

[Nobody]

I have a fan I want to use with continuously variable speed. It is a
brush less type (specs on this one say 7-12 volts) should work. And
it does . . . but won't take PWM directly, it wants to see steady
voltage not chopped. I tried frequencies from 20 kHz to 1 kHz

1. Is this typical?

Yes. A "brushless DC" motor is a synchronous AC motor which either has a
built-in AC generator or is designed to be driven by an external AC
generator. Fans normally have it built in.

Some fans filter the input voltage (to allow PWM) and adjust the
rotational speed according to the filtered voltage. Others don't. If the
input isn't filtered, it isn't going to like PWM as the built-in
controller will be continually losing power.

2. Anyone else with experience, of speed control of generic brush
less fans?

I tried my circuit with two other junk box fans and they also ignore
pwm, but like steady voltage for variable speed.

Right now I have my picaxe starting at 50 ms on, and 150 off, (ersatz
pwm) up to steady on, in four ranges - which gives me the variable
cooling I want, but it sounds like it is panting and is moderately
annoying.

Plan B:

I really want to continue using an N channel logic level MOSFET and
don't want to disturb the gizmo that's working and in-service for a
variety of reasons.

Seems to me that since my mosfet is pulling the fan's minus connection
to ground, it should be possible to just turn it into a switching
supply with steady DC output. Hook an inductor to the drain,
freewheel diode to the mosfet drain and Plus, and put the fan (load
minus) to the inductor's other terminal (and electrolytic cap across
the fan).

sort of a switching supply, but upside down and backwards

Yep; low-side buck converter; just like a normal buck converter
without the need to drive a high-side switch.

3. Does it sound feasible, or am I missing something?

Should work fine. Just ensure that you maintain the minimum supply voltage
at all times so that the built-in controller isn't getting continuous
power-out/brown-out resets.

 

[Eeyore]

I have a fan I want to use with continuously variable speed. It is a
brush less type (specs on this one say 7-12 volts) should work. And
it does . . . but won't take PWM directly, it wants to see steady
voltage not chopped. I tried frequencies from 20 kHz to 1 kHz

Filter it after chopping it.

Grahama

 

[default]

Filter it after chopping it.

Grahama

Cap filter holds the peak value. LC filter might work.
--

 

[WangoTango]

I have a fan I want to use with continuously variable speed. It is a
brush less type (specs on this one say 7-12 volts) should work. And
it does . . . but won't take PWM directly, it wants to see steady
voltage not chopped. I tried frequencies from 20 kHz to 1 kHz

1. Is this typical?

2. Anyone else with experience, of speed control of generic brush
less fans?

I tried my circuit with two other junk box fans and they also ignore
pwm, but like steady voltage for variable speed.

Right now I have my picaxe starting at 50 ms on, and 150 off, (ersatz
pwm) up to steady on, in four ranges - which gives me the variable
cooling I want, but it sounds like it is panting and is moderately
annoying.

Plan B:

I really want to continue using an N channel logic level MOSFET and
don't want to disturb the gizmo that's working and in-service for a
variety of reasons.

Seems to me that since my mosfet is pulling the fan's minus connection
to ground, it should be possible to just turn it into a switching
supply with steady DC output. Hook an inductor to the drain,
freewheel diode to the mosfet drain and Plus, and put the fan (load
minus) to the inductor's other terminal (and electrolytic cap across
the fan).

sort of a switching supply, but upside down and backwards

3. Does it sound feasible, or am I missing something?

All the other advice you got here is valid, but I have done exactly what
you are trying using a 555 based PWM and to get it to work, with the
brushless fans I had on hand, I had to go WAY up in frequency from where
you are trying. If I remember correctly I'm in the 8KHz area. My
biggest problem wasn't getting it to work, it was getting it to work
QUIETLY. All my fans would squeal at the PWM frequency when set to low
speeds, except when I went above 5KHz or so. Then they would quiet down
and run smoothly.

 

[default]

All the other advice you got here is valid, but I have done exactly what
you are trying using a 555 based PWM and to get it to work, with the
brushless fans I had on hand, I had to go WAY up in frequency from where
you are trying. If I remember correctly I'm in the 8KHz area. My
biggest problem wasn't getting it to work, it was getting it to work
QUIETLY. All my fans would squeal at the PWM frequency when set to low
speeds, except when I went above 5KHz or so. Then they would quiet down
and run smoothly.

You are using direct PWM? (fan terminals have 8 KHZ PWM on the power
terminals?) That's what wouldn't work - fans just acted like they
were a high impedance, or just not connected.

8 Khz is right in the middle of audio band so I'd expect it to be
noisy.

Yeah, at 200 ms I'm just turning the fan on and off and can hear it
start and coast, over and over.


--

 

[WangoTango]

You are using direct PWM? (fan terminals have 8 KHZ PWM on the power
terminals?) That's what wouldn't work - fans just acted like they
were a high impedance, or just not connected.

Fan + lead to +12V - lead is being pulses by a TIP102 transistor.
All 4 of the brushless fan varieties I have here work fine. It is
possible that I just lucked out and happened across of selection of PWM
tolerant fans, but that's not usually my luck.

8 Khz is right in the middle of audio band so I'd expect it to be
noisy.

No noise at most speed settings and just a little when the fan is
stalled and trying to start up.

Yeah, at 200 ms I'm just turning the fan on and off and can hear it
start and coast, over and over.

I can email you a PDF of the schematic, nothing special just a run of
the mill 555 circuit.

 

[default]

Fan + lead to +12V - lead is being pulses by a TIP102 transistor.
All 4 of the brushless fan varieties I have here work fine. It is
possible that I just lucked out and happened across of selection of PWM
tolerant fans, but that's not usually my luck.

No noise at most speed settings and just a little when the fan is
stalled and trying to start up.

I can email you a PDF of the schematic, nothing special just a run of
the mill 555 circuit.

I'm going to be bread boarding my circuit again and will give it
another try. Could be the 555 is swinging more voltage and has a
higher current than the 4.5 volts and 20 ma my pic is able to.
Suggesting my mosfet was not really saturating fully.
--

 

[Nobody]

Cap filter holds the peak value. LC filter might work.

It might work if his PWM was from a totem pole, but the OP is using
open-drain.