PWM radial blower, not starting till 50% duty cycle

Go To Last Post
66 posts / 0 new

Pages

Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi All

 

I experimenting with a blower using PWM. I working with a similar to the following  Blowers, but unfortunately the blower i have been given is not displayed on the web page yet.  The spec sheet shows PWM frequency can be between 10Khz-60Khz.

 

The issue i am having is  that i can not active the blower from duty cycle 1%(OCR0A=5), it seems that i need a minimum of 56%(OCR0A=145) to get is started.

 

Does this seem correct or am i missing understanding something here? 

 

At moment i am restricted to using a GPIO pin for my PWM, there i have used an example of the software PWM by Clawson.

 

	ISR(TIMER0_COMPA_vect) {
		// clear the output pin on OCR0 match
		PORTD |= (1<<2);
		PORTD |= (1<<3);
	}
	ISR(TIMER0_OVF_vect) {
		// set the output pin at timer overflow
		PORTD &= ~(1<<2);
		PORTD &= ~(1<<3);		
	}
	
	void soft_pwm_setup(void)
	{
			// going to use PORTC.0 to PWM the contrast voltage
			DDRD |= (1<<2);
			DDRD |= (1<<3);
			TIMSK0 |= ((1<<OCIE0A) | (1<<TOIE0)); // use both interrupts
			OCR0A = 145; // 10 out of 256 means very short on period (low voltage)
			TCCR0B |=  (1<<CS00); // timer on - nice high PWM frequency
			// Might later consider PWMing the backlight voltage too
			// so it would also be adjustable ...
			sei();		
	}

 

Thanks

 

This topic has a solution.

Thanks

Regards

DJ

Last Edited: Thu. Nov 26, 2020 - 04:54 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The spec sheet shows PWM frequency can be between 10Khz-60Khz.

I find that hard to believe--where do you see that?  You should be able to drive using pwm from  a few hundred Hz (if you don't mind the hum) up to several (or more ) KHz.

 

Before you mess with PWM at all, set your power supply to min & start sloooowly turning up the voltage to 12 (or 24, whatever it needs)...at what voltage does it start revving up?

There an be a lot of friction/inertia to overcome.

 

That will give you some insight into your PWM question.

 

You also need to slowly ramp up the PWM...not just BANG set it to some high PWM level...that will also cause stalling, due to power supply dropout under severe starting load...do you have a beefy supply & mosfet?

If you hit it with a heavier hammer, BANG then it may eventually start...but you have to set it higher than needed to launch, so you get a false impression.

 

Unless you need to change the speed often & super-accurate, any of the timer PWM modes will be perfectly fine for a fan  ...which PWM mode will you use?

 

show your circuit as well.

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Sun. Nov 22, 2020 - 09:49 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Have you asked the manufacturer what duty cycle range you can expect?

 

Running software pwm at 10kHz+ is going to place a significant compute load on the AVR. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The link you provided has three spec sheets with 3-phase interface and one spec sheet with DC supply and speed-control signal. I'm assuming the interface for your blower is the latter. Here's a snippet from the spec sheet for the speed-control line.

Looking at this it's not surprising that the motor doesn't start until PWM is around 40% or higher.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi 

 

I have attached the picture of the paper copy that i have. As you can see the mentioned PWM frequency. 

 

The Fan starts to move about 6.3V, i guess that explains the 56% PWM signal requirement. 

 

At the moment i am using bench top power supply, i eventually aim to use a 12V 3A power adaptor. I got a old pcb and modded a 2n3904 transistor by soldering it on a gpio pin ground and  PWM wire from the blower.  Therefore do not have circuit diagram. My new PCB will use the same transistor, but will use hardware PWM with the appropriate pin. I am using the transistor as PWM wire is connected to 5V, while my AVR is on 3.3V

 

In regards to the slowly ramp up or ramp down, would about 10ms per OCR0A increment or decrement be adequate? 

 

 

 

 

 

 

 

 

 

 

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It was suggested above to ramp up the PWM, as the power supply might brownout under a sudden full load.

 

I think it is better to have a heavy duty power supply that can handle the full stall current load of the motor without difficulty.

Then start up the motor with 75%, (or higher), PWM to overcome the initial stationary inertia and start the rotor spinning.

If you have an RMP sensor then you can rather quickly back off on the PWM once the rotor is spinning, if desired.

I'm not surprised that it take > 50% PMW to nudge the rotor into motion.

 

JC 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks

 

I will add an ramp up , but this blower does not need to a full speed.  How much should the ramp up delay by?

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You could probably give it 50% duty cycle to get started and then go back down to 10%.

As mentioned, there is probably some stiction to overcome to get it going.  Note that the curves don't go down to zero speed.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Just take 200 ms or 1000 ms , to ramp it up 

 

If the final pwm value (duty) is 384 counts, start at 0 & add a few counts every 10ms , or what ever you feel, just like you were turning up your power supply knob, or pressing down on your gas pedal.

It is much smoother and easier on everything, than a sudden jar/kick/blast/thump/surge

 

2n3904 ?? ...you certainly won't be using a little logic signal transistor to POWER the fan?  What mosfet will you be using?  Not even sure why you are mentioning a 2N3904.

 

Get yourself some of these Mosfets for experimenting...they are dirt cheap & you will probably be unable to ever accidentally blow them out---they can handle all the current your experimenting will probably ever need (and you can parallel fets to increase)

 

https://www.digikey.com/en/produ...

 

 

 

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Sun. Nov 22, 2020 - 11:26 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

MattRW wrote:

You could probably give it 50% duty cycle to get started and then go back down to 10%.

As mentioned, there is probably some stiction to overcome to get it going.  Note that the curves don't go down to zero speed.

It switches off at 10%,it seems like pwm 50% is datasheets 10%

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

avrcandies wrote:

Just take 200 ms or 1000 ms , to ramp it up 

 

If the final pwm value (duty) is 384 counts, start at 0 & add a few counts every 10ms , or what ever you feel, just like you were turning up your power supply knob, or pressing down on your gas pedal.

It is much smoother and easier on everything, than a sudden jar/kick/blast/thump/surge

 

2n3904 ?? ...you certainly won't be using a little logic signal transistor to POWER the fan?  What mosfet will you be using?  Not even sure why you are mentioning a 2N3904.

 

Get yourself some of these Mosfets for experimenting...they are dirt cheap & you will probably be unable to ever accidentally blow them out---they can handle all the current your experimenting will probably ever need (and you can parallel fets to increase)

 

https://www.digikey.com/en/products/detail/infineon-technologies/IRLZ24NPBF/811787

 

 

 

 

The fans connect to 12v directly.

I am using 2n3904 for own signal only.

Why would I need a MOSFET?

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

 

You need to apply the PWM to a power transistor to drive the motor, correct?  That is your power fet.

 

The 2N3904 will not drive a motor that requires any current, like a large blower that requires amps.   It might drive an led, but not a motor, unless it was some microscopic size motor, like a pager buzzer that draws 0.05 Amps.  Show your circuit.  Do you already have some sort of driver?

 

Here is an example PWM for driving a motor with a fet

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Mon. Nov 23, 2020 - 01:36 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

I expect the fan in question has its own drive electronics - if this is the case, no power fet required.

 

Have you put a 'scope on the pwm signal?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I expect the fan in question has its own drive electronics - if this is the case, no power fet required.

I see, now---thought it was strange the fan datasheet even mentioned PWM.   So they must be providing some type of driver that is not disclosed.  

Of course the OP sort of skips mentioning that the driver is already included  (I experimenting with a blower using PWM)  !!

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

A  number of years ago I had a similar problem with PC fans.

They would not start at low PWM duty cycles.

I could put them on a low PWM , they would not start, until I moved the fan blades a bit by hand.

In my case I concluded it was a start up power problem.

The motor would not get enough power to start rotating the blades.

What I did is start at 100% pwm for 1 second and then drop back to the minimum allowed ( was a user setting what minimum duty cycle was).

I also checked that the Tacho readback would never become 0 as that would mean the fans had stopped. I then gave a short burst of 100% to get them going again.

I do not recall if I then also incremented the minimum set PWM to ensure the fans would now keep on turning. It was many moons ago.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kartman wrote:

I expect the fan in question has its own drive electronics - if this is the case, no power fet required.

 

Have you put a 'scope on the pwm signal?

 

Its a 4 wire version. 

 

The PWM signal is pulled up to 5V internally, therefore i used the transistors.

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

avrcandies wrote:

 

You need to apply the PWM to a power transistor to drive the motor, correct?  That is your power fet.

 

The 2N3904 will not drive a motor that requires any current, like a large blower that requires amps.   It might drive an led, but not a motor, unless it was some microscopic size motor, like a pager buzzer that draws 0.05 Amps.  Show your circuit.  Do you already have some sort of driver?

 

Here is an example PWM for driving a motor with a fet

 

 

The fan is 4 wire fan, with a PWM signal wire

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Spend a bit more time when formulating a question. You rely on us to read between the lines.

 

Have you used an oscilloscope to view your pwm pulse? Don't assume it is correct. In fact, this should've been one of the first things you did.

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes i am using the oscilloscope to observe the PWM signal

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

An observation based over a *lot* of experimentation with small motors and PWM: the torque available for a motor is not constant with PWM frequency... at low PWM frequencies (200-1kHz), the torque is significantly greater than at higher frequencies (20kHz and more). I have a lot of graphs and measurements, but they're probably classified as company info so I'm afraid you're out of luck.

 

To use the high frequencies (to avoid audio hum) I had to jump through hoops including starting at flat out for a few milliseconds to get things moving before switching to a closed loop control (measuring speed by measuring the generated voltage as the motor free-runs briefly) and with different parameters depending whether the motor was over or under speed.

 

I speculate, but don't know, that at the higher frequencies the inductance of the motor becomes significant, and is busily turning the supply into heat instead of into rotation...

 

Neil

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'll just note that my Dell PCs start up with the cooling fans running at high speed, and then slow down to a slow, and quiet, RMP after several seconds.

 

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

Note: op is NOT driving the motor directly with the PWM signal, OP said it has a motor control board that uses the PWM duty cycle to set the speed!

But I agree with the others, most motors need to be started at higher PWM duty, to overcome static friction, once spinning, the OP can lower duty cycle to whatever they choose.

Jim

 

 

(Possum Lodge oath) Quando omni flunkus, moritati.

"I thought growing old would take longer"

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Why did you not show us this? In future, take some time to formulate your question better. Tell us what you’ve tried etc. it means you will get a more precise answer, faster.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0



But I agree with the others, most motors need to be started at higher PWM duty, to overcome static friction, once spinning, the OP can lower duty cycle to whatever they choose.

It depends on the load you are trying to move...if you have a load that has a lot of dead inertia, trying to suddenly move it at full tilt easily cause such a surge that your supply and/or your processor conks out.  You can greatly reduce the peak demand by taking it easy on  the ramp up.   So, instead of needing an 6 amp power brick, you might only need a 3 amp wall wart.  The more stiction you have to break free starts to move the equation in the other direction, so you really have to look at the mechanical drive assembly as well as the proposed load to be driven.   We had some assemblies that would spin quite easily (very low friction), but bringing them up to 6000 rpm using different methods were able to trade off surge vs startup time. 

 

The DP Series also offers options for a variety of combinations of Internal PWM Soft Start/Ramp Up, PWM Soft Stop/Ramp Down and Dynamic Brake functions. Available with a selection of set ramp times, the soft start and stop functions provide a convenient means to eliminate or reduce the mechanical shocks associated with starting and stopping DC electro-mechanical loads.

 

Many years ago we had a high mechanical load that was brought to an instant dead stop by throwing the driver in full reverse  (called "chogging?")...you could almost hear the electrical cables groan.  I bet it was 100+ amps peak on a 5 amp normal circuit.  

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks , i will be implementing a gradual duty cycle change, when the every a change happens.

 

At the moment i am using software PWM, due to my PCB, but will have hardware PWM on my new PCB.

 

I have noticed that with my code, if i was to have frequency of 31 Khz and i gradually increase the duty cycle to 90% about 250, i am not really get the 90% from the fan. I am expecting similar results as if the FAN is directly connected to maybe 11V, but i am experiencing the same level as i would do with 7-8V.

 

Where else if was to use a 256 as pre scaler in my timer, i am getting similar results then to 11V, but i do hear a humming noise.

 

Is this expected? Would this be due PWM being software based?

 

Would it be better to use Fan control IC?

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Would it be better to use Fan control IC?

Prob not.  I gave you a circuit in #17 that can drive almost any fan, small or gigantic, that will blow the paint off your walls.  Why not find a fan, a FET, and have fantastic fun trying it?  It really doesn't get much simpler.

 

https://www.allelectronics.com/i...

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

At 31kHz, are you really getting 99-100% duty cycle? You have an oscilloscope - it will tell you in seconds.

 

If you get 100% duty cycle, then the pwm frequency is irrelevant - it should be DC. Thus the fan should operate no differently.

 

You say it is operating differently, so you need to find out why that is so.

 

If you use a prescaler in your timer, then that affects the pwm frequency yes? Is the pwm frequency below what the manufacturer suggests?

Might it be in the audible frequency range? Humming? quite likely.

 

Due to software pwm? More likely due to YOUR implementation of software pwm. The fan neither knows or cares how the pwm signal it sees is generated. I did mention earlier in regards to the compute load of generating high frequency pwm by software. This is why I suggested measuring with an oscilloscope.

 

Would a fan control ic be better? Firstly you need to ensure you comply with the fan manufacturer's specs and not try to work against physics.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

At 31Khz, i am getting the duty cycle as set by the AVR, this is shown in oscilloscope .  I was expecting as the duty cycle increases from 0-100%, that fan would also increase from 0 to 100 in fan speed.   But it seems that as duty cycle get close to 100%, the fan speed is about 40-50% or a little less .

 

After a little investigation, it seems that this FAN can also be controlled by applying an voltage on the same wire. When i measured the voltage when duty cycle is 90-100% on the PWM wire, it was 1.6V.  Based on the datasheet, 1.6V on PWM wire would produce about 40%.  This also means as duty cycle increased from 0-90%, the fan speed must be lower then what it should be.

 

Does this make sense that a FAN can be controlled but both PWM and Voltage on same wire? It seems the PWM signal needs to replicate the voltage.

 

I have ordered a standalone PWM generate so that i can experiment with various frequencies and duty cycles.

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

But it seems that as duty cycle get close to 100%, the fan speed is about 40-50% or a little less .

How are you measuring fan speed--how can you say this? 

 

When i measured the voltage when duty cycle is 90-100% on the PWM wire, it was 1.6V.

...what voltage?   PWM  requires an Resistor- capacitor to generate a DC avg voltage.

 

For additional help, you need to show a schematic of your circuit, including the fan.   

 

If you only need to supply logic signal to the fan, get rid of your transistor.

 

 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Wed. Nov 25, 2020 - 03:43 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Well i am comparing when connecting fan directly to 12V, when PWM is not connected to when duty cycle 90-100%, and there is a lot of difference. 

 

At the moment, i have moded a PCB by soldering a transistor to a GPIO pin and GND, and wire to PWM line. Its very basic, as i want to see it working in principle before i start making a new PCB,

 

 

In regards to the resistor/Capacitor that is what i believed, would be the case, but the datasheet and conversation with company, they mentioned the circuit is already included in the fan, as shown in image below.

 

 

 

 

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Are you intending to burn out the transistor by forgetting your base resistor?  

 

Get rid of the transistor--it dos nothing for you...you can see you just need a logic signal PWM

 

They provide the RC, or you can provide you own and use a low freq PWM, say 5 KHZ, use 2 kohm & 1uF to filter it.

 

If you apply Vapp (DC) to the input... Vctrl=0.4545 +0.9091*Vapp 

 

as i want to see it working in principle   

You transistor circuit needs scrapped--serves no purpose, plus it inverts.  Why did you have it?  Only reason   is if you needed a wider logic level swing. 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Wed. Nov 25, 2020 - 09:14 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

My AVR is working on 3.3V. ,as the fan has an internal pull up to 5V, would this not cause an issue?

 

This is the reason for using a transistor. 

 

 

 

 

 

 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

My AVR is working on 3.3V. ,as the fan has an internal pull up to 5V, would this not cause an issue?

Ok, NOW you have a reason.

 

Add a base resistor of 1k to 10k between the AVR & base  (throw away the blown out/damaged transistor).   You would be better using a fet, since it will pull down to 0V, The collector won't quite hit  down to 0., but you take that into account.

 

Note the transistorized PWM will be upside down...when you give 30%, it will appear as 70% to the fan, your software will consider that 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

Last Edited: Wed. Nov 25, 2020 - 09:25 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You seem to do a lot of guessing. Yes, you need the transistor or a mosfet. With the transistor you'll need a base resistor. Remember your transistor basics!

 

 

It's taken us 34 posts to get an idea of what is required. Especially when it was all in the datasheet.

Last Edited: Wed. Nov 25, 2020 - 09:25 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Whoever designed this fan seems to be a dum-dum...they should have used 1uF to get a reasonable time constant (instead of 0.909 us !!!), so then the PWM freq could be kept low (or just add your own RC filter) at the fan input pin.

 

There is NO reason you'd need to control the fan down to the submillisecond!  Ask the nearest fanboy for a return.

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The transistor seems to be still working, but i got another PCB where i got a FET controlled by GPIO. The FET is FDC637BNZ. I am still get the same issue, where as duty cycle increase from 0 to 100%,  i am not getting a gradual speed increase till the fans max speed.

 

I guess i will experiment more tomorrow.

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Once the fan is actually moving, does the PWM correspond to the fan speed?  Plot a graph.

Add your own RC, so you can use a 5 KHz PWM, you have no need for some sort of high freq control.

  

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Why use the FDC637BNZ mosfet? Did you look at its gate capacitance? You would use this mosfet for switching higher currents, like is a switchmode supply. 

You want something like a BSS138.

You randomly do things then are mystified when they don't work. Check each stage - carefully. Working this way you are more likely to have success.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

avrcandies wrote:

Once the fan is actually moving, does the PWM correspond to the fan speed?  Plot a graph.

Add your own RC, so you can use a 5 KHz PWM, you have no need for some sort of high freq control.

  

The pwm duty cycle does not correspond to fan speed but the measured voltage does. It seem voltage gets capped to about 1.6V at 90%, I would expect about 4.5V, as fan has internal pull up of 5V.

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Plot a graph.

 

It seem voltage gets capped to about 1.6V at 90%

 

...what exactly are you talking about?   What voltage?  WHAT CIRCUIT?  Why are you unable to show circuits or scope traces?

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

avrcandies wrote:

Plot a graph.

 

It seem voltage gets capped to about 1.6V at 90%

 

...what exactly are you talking about?   What voltage?  WHAT CIRCUIT?  Why are you unable to show circuits or scope traces?

Will try to produce some plots tommorow.

The 1.6v I am referring to is the voltage measurement when using a voltmeter on own pwm wire. This is at 90% duty cycle.

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Post 30 shows the internal circuit of the fan.

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You may need to drive the input to the fan actively high. Do the calcs - you know what the circuit is and the component values. The numbers will guide you.

 

At a pinch, put a 1k pullup to 5V. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Post 30 shows the internal circuit of the fan.

  where is YOUR CIRCUIT & COMPLETE HOOKUP?

 

why do you seem to be intent on wasting everyone's time?     Where are your plots & scope traces? 

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

avrcandies wrote:

Post 30 shows the internal circuit of the fan.

  where is YOUR CIRCUIT & COMPLETE HOOKUP?

 

why do you seem to be intent on wasting everyone's time?     Where are your plots & scope traces? 

Will produce them tommorow. At the moment I do not have my own circuit, I am just using an old pcb with avr to get it working correctly.
I have modded a transistor onto a board to see if the fan can be controlled.

Thanks

Regards

DJ

Last Edited: Thu. Nov 26, 2020 - 01:04 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

djoshi wrote:
I am just using an old pcb with avr to get it working correctly.

 

That is the very reason why you are not having success!

 

Solve the problem on paper first. If it doesn't work on paper, then it is unlikely to work in reality.

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kartman wrote:

djoshi wrote:
I am just using an old pcb with avr to get it working correctly.

 

That is the very reason why you are not having success!

 

Solve the problem on paper first. If it doesn't work on paper, then it is unlikely to work in reality.

 

 

True. I will make a circuit tommorow on paper . It late at the moment

Thanks

Thanks

Regards

DJ

Last Edited: Thu. Nov 26, 2020 - 01:09 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

At the moment I do not have my own circuit, I am just using an old pcb with avr to get it working correctly 

Well that IS the circuit you are using, why paly around with words?  ....you have not got it working, but skip the details.   Please be sure to provide them, when you can.

When in the dark remember-the future looks brighter than ever.   I look forward to being able to predict the future!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Where have you been Atta hing your scope probe to? The pwm pin on the avr side of the transistor or the fan side?

As Avrcandies mentioned in post #33 your pwm signal will be inverted at the fan compared to the avr. So when you say you have it at 90% is that the avr is outputting at 90% or at 10%?

Last Edited: Thu. Nov 26, 2020 - 02:25 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi 

 

I have changed the PCB to the one with Mosfet and this my design no, therefore this my design. At moment I am ignoring the TACT. 

 

 

 

Once the design is correct, I can then post any plot or scope outputs if it still does not work. 

 

I have been attaching the probe on AVR side. I agree that the signal will be inverted,  I have compensated for this in my firmware, as I am using software PWM. But even though, 90% would be equal 10%, then 10% would be equal to 90%, but still, the fan is not at full speed. 
 

Attachment(s): 

Thanks

Regards

DJ

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

As I mentioned earlier, the mosfet is not a suitable type and you most likely need a pullup to 5V for the pwm signal. Design involves a bit more than just using Altium.

 

 

Pages