Double Fan conections and experience...

[r4ix 0]

Hello.

I suppose double fan have some benefits to the prints so the printed part is cooled symmetrically. For this reason i decided to conect another fan to my ultimaker original. I design a fan duc with solidworks, printed it and mount it.

As the fan have low current consumption i decide to connect it in parallel and i supposed the mother board out for the fan could assume this 100mA extra current, and it did it.

Everything was working perfect until i decide to change the speed of the fan.

When i changed the speed a bit, from 255 to 180. The temperature becomes unstable a decrease drastically, to 170/180 and is unable to increase to the temperature i was printing with. This don't happen with one fan only.

I assume the out of the fan doesn't come directly from the arduino board cause arduino out could only give 40mA. But i don't undestand why changing the fan speed down affect the temperature.

Has anyone had any problem like that? This only happen with the double fan.

 

[baasb 6]

Wild guess.. That the fans are blowing to much on the nozzle. I think I've noticed that there is an optimum fan speed for a particular fan duct. For example, from speed 0 to 180 the airspeed coming out the fan duct is increasing but when you go higher then that optimum (180), the airspeed actually decreases because of pressure buildup in the fan duct.

So although you feel like the fan speed is lower, the air output should be lower, the opposite might be true. And so cooling the nozzle.

But again, it's just a guess! I also use two fans in parallel and never had any problems with it. That's why I think the problem could lie in the actual cooling of the nozzle.

 

[r4ix 0]

So you think the optimum fan speed is 180? Maybe my fan duct design is also not the best one because just make all the air directly to the nozzle.

But even this, the termocoupler try to give stable temperature, isn't it? So it should give enough energy to get it stable even there is more cooling. I don't think the air take more energy (heat) to the nozzel than the board is able to give...

I don't know if i explain my self good enough...

 

[baasb 6]

The 180 was just an example

And your explanation is good, and I think you also might be right with that. But there is a limit of how much heat the heater can generate versus the heat that is taken away by the cooling air. If you look underneath the UM when heating, you see a red LED lighting up constantly and it starts flashing when it has reached the set temperature (so turning on and off the heater to maintain the desired temp). Try to heat up to your printing temperature and then turn on the fans. They probably start cooling the nozzle. Then check if at any moment the red LED starts flashing or that it stays on continuesly. If it stays on continuesly, it means the heater is fully on and cannot keep up with the dissipated heat by the fans.

Also, a picture of you ducts would be helpfull

 

[gr5 2,167]

This suggestion - air on the nozzle - makes sense until you realize the nozzle was *more* stable at full speed.

So...

I think the problem is electrical interference. When you set the fan speed to 180, the fan is turned on and off many times per second. This causes strong high frequency spikes in current which causes voltage spikes (noise) especially in wires that are parallel with the fan wiring. Your new fan is probably more inductive or more capacitive than the existing fan.

The fix would be to separate the fan cabling farther from the temperature probe cabling. The temperature wiring is already twisted pair which helps but isn't good enough as they two signals don't go through a differential amplifier like they should but instead one side is grounded and the other goes into the arduino.

Even better shield the power cable to the fan with a grounded shield.

Or just always use 100% fan. I find that 100% fan gives the best results.

 

[owen 19]

You can also try putting on a silicon jacket on your nozzle and heater block like this http://www.thingiverse.com/thing:165750

It works very well.

 

[r4ix 0]

The interferences have sense. Normally the pwm frequency to control motor is over 20khz, 32khz usually... Could interfere. I will do some test and i will post it...

 

[r4ix 0]

Hi!

I've been doing some test for a while.

I have interference with pwm even with one fan.

The temperature became unestable if the fan was not at full speed (so full duty cicle in pwm)

I put a twisted cable out of the printer as gr5 said. It works well and now temperature is totally stable.

The bad thing is now i have one extra cable...

 

[gr5 2,167]

picture please!