Two Axial Fans, two different currents; what is the difference?

[mike buba]

Hi, I am using two different axial fans (1) and (2) and when I record the currents, I get two different current waveforms.

The first one is this (Ch1 is current):


And the second one is this (Ch1 is current):


The second fan current is now what I wanted or expected.
Ideally, I would like to have simple v/f control, where, as I increase the voltage, the fan frequency (i.e. speed) increases, the current is sinusoidal and only the amplitude changes. Similar to the first fan behaviour.
Edit: speed control is not a primary requirement. However, it would be good to have a sinusoidal current as the fan is supplied from the converter output and a non-linear current might affect closed-loop converter control

Did I miss something in the datasheets when I bought the second fan?
Can this be somehow "fixed" to get the sinusoidal current?
If not, what should I look for if I want to get another 80 x 80 x 25mm 230 Vac fan, but with the current waveform similar to the first fan?

Thanks

[wraper]

There is no voltage waveform on the first picture. The difference between these fans it that first one (NMB) uses shaded-pole motor which is extremely inefficient and heats like hell.
If you want to control speed it's better to not use AC fans at all. Use PWM controlled DC motor. It won't need power electronics to control speed, will consume less power, heat less and will be more reliable.

[mike buba]

There is no voltage waveform on the first picture.

Didn't record it
It is the second fan behaviour that I didn't understand so I wanted t see if there is any effect on the voltage; no effect!

If you want to control speed it's better to not use AC fans at all. Use PWM controlled DC motor. 

Not interested in fan speed control.
The fan is used for cooling the DC-to-AC converter and is supplied from the output AC voltage. My initial idea was to control the speed by controlling the voltage, but in the final version, the converter output will be constant at 230 Vac, so I'm okay with the nominal fan speed.
My concern is that this non-linear current might affect the closed-loop control. Hence, I would be more "happy" with the sinusoidal fan's current.

[wraper]

My concern is that this non-linear current might affect the closed-loop control.

What closed loop? Nobody does closed loop current control with AC fans. Normally it's just a triac dimmer. Second fan likely will have a problem with that though, not because of current waveform but because of motor type. Sunon calls it "Alveolate motor" but as I understand it's basically a an induction motor with run capacitor.

[mike buba]

Still, the question is why am I getting this current waveform, and not sinusoidal? And how to make it sinusoidal?

[wraper]

Still, the question is why am I getting this current waveform, and not sinusoidal?

First of all current would be shifted in time from voltage waveform if you actually made voltage measurement with the first fan. So there is nothing good about it being similar to sinusoidal which it actually is not as it's a triangle.

And how to make it sinusoidal?

Very complicated, won't achieve anything and would be totally stupid. Basically you would need active PFC + convert DC back to AC.

[Siwastaja]

Why do you think you need to make the current sinusoidal, and why do you think the current waveform matters at all? (I don't understand the closed loop control comment, please elaborate.)

Why have you ruled out using a DC fan supplied from the DC input of the DC/AC converter? BLDC fans are way more efficient and you could easily control the speed with PWM + MOSFET, while keeping good efficiency, if you ever decide you want speed control after all.

[wraper]

BLDC fans are way more efficient and you could easily control the speed with PWM + MOSFET, while keeping good efficiency, if you ever decide you want speed control after all.

It's more complicated. To do it properly, you basically need an adjustable buck converter when using a brushless DC fan which does not have PWM input. I have no idea why some DIYers and engineers (even in commercial product) assume you can simply PWM power of brushless DC fan which is not just a motor but a circuit with controller. Which already makes it strange to cycle power of some circuit hundreds of times per second end expect it to operate properly. It kills the fan and often produces annoying noise (often once electrolytic capacitor within the fan dies after such abuse). Some fans are more tolerant than others but I've not seen it allowed in the datasheets.
Here is a Sunon datasheet explicitly prohibiting it https://www.tme.eu/Document/69f9cfd6dcfdd925f7d57f14b587f425/EE40101S1-999-A.pd

DO NOT use power or ground PWM to control the fan speed, If the fan needs to be adjusted, please contact SUNON to customize the product design for your application

[shapirus]

It's more complicated. To do it properly, you basically need an adjustable buck converter when using a brushless DC fan which does not have PWM input.

Or use a 12V 4-wire computer fan which is easy to both power and control.

[wraper]

It's more complicated. To do it properly, you basically need an adjustable buck converter when using a brushless DC fan which does not have PWM input.

Or use a 12V 4-wire computer fan which is easy to both power and control.

Well, I suggested to use PWM controlled fan in my first post in this thread. Many are tailored to PC market, but there are plenty which are not.

[Siwastaja]

It's more complicated. To do it properly, you basically need an adjustable buck converter when using a brushless DC fan which does not have PWM input.

... or get a fan which can be PWM'd. Sure enough, it seems that in $current_year, BLDC fans with more complicated inverter circuitry are common.

You are of course right, my advice is outdated. In the past, all / most BLDC fans could be PWM'd because they had a simple array of transistors controlled by hall sensors and not much if any decoupling capacitance on DC. It was widely known that "the only problem" PWMing the GND side is losing the tacho output signal. Gazillion of products and DIY projects just PWM'd these fans without a problem.

I second the advice of getting a fan with PWM input designed for the job.

[wraper]

Gazillion of products and DIY projects just PWM'd these fans without a problem.

I have GW instek PSP-405 PSU which had very annoying noises, fan not starting at low load and just singing. Tried changing the fan to something else but it did not solve the problem. They effing PWMed the fan, even though in older revision of this model there was linear regulation. IIRC I added like 2-3 resistors, a cap or two, reused existing 2 transistors and made it convert PWM signal into DC within acceptable range. Problem gone.