Pressured air cooling controlled with servo and ball valve



  • Hey all,

    since I got my RC-Servo finally working (Thanks to @achrn)

    I would like to share a little project that i have been working on.
    I have been using pressured air from electric pumps instead of fans for hotend and part cooling à la Berd Air. I found that the motor Pumps are quite noisy, get very hot and do not last very long. Therefore, I switched to this pump It is less noisy and has more than enough airflow.

    I would like to add a ball valve that is opened and closed by an RC-Servo in order to control the airflow for the part cooling.
    This is my fan configuration:

    M950 F1 C"io4.out"       ; create fan 1 on pin out8 and set its frequency
    M106 P1 S0 H-1           ; set fan 1 value. Thermostatic control is turned off
    

    With this I can already somewhat control the fan over the web interface.
    The problem is that the M106 fan speed value (0 to 255 or 0.0 to 1.0) does not correspond to the needed servo values of 0 to 90 for an angle between 0° and 90°.

    For example the servo only responds when the fan control slider in the web interface is set between roughly 5% and 65%.

    Does someone have an idea how to work around this ?

    Cheers,
    Max



  • https://forum.duet3d.com/topic/18110/thermostatic-servo-control

    I haven't actually got this beyond prototype - it will be ventilating my enclosure in due course, but at the moment my 'enclosure' doesn't have any panels on the side, so doesn't need any ventilation.


  • Moderator

    Instead of trying to control the servo with a fan command why now use the actual servo control command?

    https://duet3d.dozuki.com/Wiki/Gcode#Section_M280_Set_servo_position

    Your M950 would use S instead of F to define a servo pin, and then your M280 command would target that pin number.

    https://duet3d.dozuki.com/Wiki/Gcode#Section_M950_Create_heater_fan_or_GPIO_servo_pin



  • @achrn this is a very interesting approach and great idea in general. But as far as I understood, the use of a daemon.g is not yet available for the Duet3+SBC.
    What i basically need is a way of mapping the fan value to the servo position.
    For example M106 P0 S0.5 corresponds to M280 P0 S45 -> When the fan is set to 50% the valve is opened half way.

    @Phaedrux The servo is supposed to adjust the airflow according to the M106 fan command in the G-Code. I Am using Simplify 3D as a slicer, which would give me the option to replace the M106 command with an M280 for servo control. But the problem is that the fan command uses either a percentage value or a PWM range of 0-255.
    Also, it would be nice to adjust the valve position over the web interface.

    Unbenannt.PNG

    Cheers Max



  • @MaxGyver said in Pressured air cooling controlled with servo and ball valve:

    @achrn this is a very interesting approach and great idea in general. But as far as I understood, the use of a daemon.g is not yet available for the Duet3+SBC.
    What i basically need is a way of mapping the fan value to the servo position.
    For example M106 P0 S0.5 corresponds to M280 P0 S45 -> When the fan is set to 50% the valve is opened half way.

    It works on my Duet3 6HC with attached Pi4, DSF Version: 3.1.1, Firmware: RepRapFirmware for Duet 3 MB6HC 3.1., on io7.

    I don't really understand the explanation you're giving (it's not clear to me if you are wanting it specifically to react to a fan rather than to a temperature) but below one temperature the servo is in one position, and above another temperature it's in a differnt position, and it moves linearly from one to the other. Is that not what you're trying to achieve?



  • sounds like he wants to control servo position based on fan speed. what about hooking servo up to a external control ,like what came with the berd air, controled by the pwm fan pins you could use a bec to feed the controler for the servo voltage
    just a thought



  • @chainbreaker2012 said in Pressured air cooling controlled with servo and ball valve:

    what about hooking servo up to a external control

    If it needs to be controlled from a fan signal rather than from a temperature sensor, that's what I'd do:

    A (very) low pass filter to convert the PWM to an analogue level, then have an arduino-ish read the analogue periodically and generate a servo-friendly PWM. That was actually my initial plan for the servo control I'm doing. However, that's largely because everything I'd need for that is in my bits boxes already, it's not necesarily an elegant solution.

    My other initial plan was to try and read the fan PWM directly and skip the analogue stage, which seems to me slightly more elegant and slightly more fraught.



  • ok understandable, I am doing something similar but I am coming off a compressed air line from outside and using a pressure transducer to regulate pressure to the hotend, and print cooling. I am just waiting on the transducers to come in.



  • @chainbreaker2012
    Exactly, I would like to control the servo position based on fan speed.

    Since this already kind of works without any additional electronics on the Duet, it is rather a software issue. My problem is, that the fan S parameter (0-255) does not correspond to the servo angle. The servo can be controlled with an S parameter between 44 and 141 which correspond roughly to 20-80% fan speed.

    Therefore, I need to adjust the given S parameter of the fan output, so that at 0% fan speed the given S parameter is for example at S 44 (0° Servo angle) and at 100% fan speed it is at S 144 (180° Servo angle).

    Cheers Max



  • @MaxGyver can you not use the L and X commands of M106 to do that?

    Lnnn Set the minimum fan speed (0 to 255 or 0.0 to 1.0) when a non-zero fan speed is requested.
    Xnnn Set the maximum fan speed (0 to 255 or 0.0 to 1.0) when a non-zero fan speed is requested. (supported in RRF >= 2.02)
    


  • @ dc42 maybe an option could be added to the M106 command, similar to the R value in M453

    Rbbbor Raaa:bbb RPM values that are achieved at zero PWM and at maximum RPM. Used to convert the S parameter in M3 and M4 commands to a PWM value.
    


  • @jay_s_uk said in Pressured air cooling controlled with servo and ball valve:

    @ dc42 maybe an option could be added to the M106 command, similar to the R value in M453

    Rbbbor Raaa:bbb RPM values that are achieved at zero PWM and at maximum RPM. Used to convert the S parameter in M3 and M4 commands to a PWM value.
    

    Yes, this would solve my problem. 🙂

    Cheers
    Max



  • As an alternative to my Ball Valve+RC Servo design, I have ordered a PWM capable proportional control valve. The advantage of the Proportional valve is that it is basically plug & play and has less moving parts. But just like the servo, it has a specific PWM range from 300-400 Hz for 0-100% flow. So my Problem remains the same...

    IMG_20201019_140004.jpg

    @dc42 would it be possible to add an R value to the M106/M950 command to control the PWM frequencies that are achieved at zero PWM and at maximum RPM similar to the R value in M453?

    Cheers
    Max



  • The data sheet at https://www.burkert.com/en/Media/plm/DTS/DS/DS2863-Standard-EU-EN.pdf doesn't seem to match your decription.

    That just refers to PWM and the figure on page 8 shows a pretty linear repsonse from about 0.1 to about 0.9 input - probably more linear than any fan, so I don't believe you do have the problem you think you do - surely you just use this as a 24V fan, with a 400Hz PWM? You could put 0.1 low limit on it so it opens immediately you have any non-zero demand.

    I do note that I don't understand the X axis scale on that graph, because it's a 24V 420mA coil. Also I'm assuming that response to PWM will be similar to the response to whatever that scale is indicating.

    Where are you seeing something that suggets you need to vary PWM frequency? M950 already lets you specify teh frequnecy and then the standard fan control will vary duty.



  • @achrn

    Thank you for your feedback. I think you are partly correct.

    I have tried setting different PWM frequencies in the M915 command. The valve behavior does not change much. Although on higher frequencies (Q>4000) the valve stopped whining, so I am going to stick to those.

    This is my recent Fan/Valve config:

    M950 F1 C"1.out1" Q-1
    M106 P1 S0 H-1
    

    I have good control over the airflow with a fan speed from 78% (valve closed) to roughly 90% (valve fully open).

    Cheers
    Max


  • Moderator

    @MaxGyver said in Pressured air cooling controlled with servo and ball valve:

    78% (valve closed) to roughly 90% (valve fully open).

    Look at the M106 L and X parameters to set the min and max range to match your usable range. That should scale the S parameter to make S0 to S255 within your usable range.



  • @Phaedrux

    I have already tried this. The L and X factor of M106 will not scale the range but rather set the minimum and maximum fan speed.

    With this configuration the valve will stay in the same slightly open position at a fans peed of 5-80%. So I still only have real control over the flow from 80-90% fan speed.

    M106 P1 S0 H-1 L200 X255
    

    Cheers
    Max


  • Moderator

    @MaxGyver My mistake.



  • @MaxGyver said in Pressured air cooling controlled with servo and ball valve:

    Although on higher frequencies (Q>4000) the valve stopped whining, so I am going to stick to those.

    This will for sure change the valve response characteristics. Perhaps this is why the valv response is not as stated in spec.

    400Hz is the way to go.



  • @BoA The valve response characteristics do not change with the PWM-Setting. I have the same response with 300, 400 or higher Hz. Only with high Hz the valve is noiseless.

    Cheers
    Max



  • @MaxGyver how is this valve working out for you? I'm looking at a similar setup but using a compressed air source instead of an aquarium pump.

    Have you put a flowmeter after your solenoid to determine how much air you are delivering and able to control the airflow?



  • Hey @lael,

    yes the setup is working well,

    I am using this pump by the way.

    It has more than enough airflow, good pressure and is not too noisy.
    I am switching the 230V power for the pump with a relay.

    The valve is also working good so far. Although I have not figured out how to map it correctly to get a range of 0-100% airflow like mentioned in my previous posts.
    Right now I am printing with PLA where I have it fully open all the time anyway.

    I have tested the airflow with a flow meter. At the hotend I have an airflow of around 20-25l/min. This reduction is mostly due to the 4 mm internal diameter of my tubing. But so far the airflow is sufficient.

    -Max



  • @MaxGyver Good to hear! That is quite a lot more airflow than Berd Air will supply. What is the max and min airflows you can control it to?

    So for the wiring configuration you have a relay in line on the PWM fan output that triggers the Relay for 240V to the pump? That way when the fan would be on the pump turns on?



  • In terms of part cooling at the hotend - this guy: https://twitter.com/brendonbuilds/status/1313914876922912768/photo/1 has an interesting setup. Looks like a resin printed part, but a 6mm aluminium tube should be able to be bent into the same design.



  • @lael I can control the flow from 0% to 100%. The problem is that this range corresponds to the area of 80-100% fan speed on the duet. My workaround is to set the minimum fanspeed in my slicer to just over 80%.

    Since I am also cooling my hotend with the air pump, I have set up the relay to trigger when the hotend reaches a certain temperature.


Log in to reply