Is there any data on Duetwifi noise factors?



  • Sorry if the subject is a little clunky, I was trying to keep it short 😕 Anyways, what I'm hoping someone has is a data set that graphs out or somehow describes the correlation, if any, between load (or maybe current) and how much it affects stepper noise output. Also if there is something similar that describes the noise produced in relation to the temperature of the drivers, if this varies much from the other.

    I know it sounds like an odd request but I'm pretty sure the weight of my gantry (10+ lbs) is making the corresponding steppers a bit louder than the other ones carrying much less weight and I'm less sure about this, but I feel like my printer is much quieter at the start of a print compared to the rest of a print over the course of a few hours using NEMA 23 steppers at 2.2a so I have a hunch my little heatsink and fan setup it capping out at some point.

    I know the solution is basically less moving mass and more cooling but I'm trying to maintain some semblance of a budget while still sufficiently improving my printer. So if I don't need a carbon fiber gantry and liquid cooled drivers, I'd prefer not to go quite that far lol

    Thanks a bunch for any info guys!



  • PTFE lube… I know it doesnt answer your question but it helps



  • @Whitewolf:

    PTFE lube… I know it doesnt answer your question but it helps

    It is normally my position that lube is always an acceptable answer, if not THE answer to many of lifes questions, but I have to admit, I'm not sure how this relates 😛



  • You stated the weight of the gantry makes your steppers louder. PTFE lube has a really low friction coefficient which will reduce the effort that the steppers need to move it around… or maybe im reading something from your post wrong



  • Ahh, that's cool I appreciate the input, I hadn't thought to check that. Without any belts attached when the gantry is just rolling freely, there isn't much of any resistance that I can tell using Openbuildparts delrin V wheels and linear rails, but I should test it again since I haven't check the health of those particular bearings for a while.



  • I too use Open builds Delrin wheels and can testify to their low friction characteristics. As a little exercise I wondered just what it would take to move my extruder carriage consisting of 3 off Titan extruders plus steppers mounted on printed PETG upper and lower carriages with 4 off Delrin Wheels running on 2020 Vslot. I rigged up a simple pulley with a cord attached to the carriage then added weights. The carriage assembly weighs in at a tad over 950gms but to get it rolling and keep it rolling needed only 65gms. There would be some friction for the cord running over the pulley and from the pulley itself so the actual force to overcome friction would be a bit less than the 65 gms.

    TBH though, the force required to overcome friction is tiny compared to the force required to accelerate the mass, (Force = mass x acceleration) so although minimising friction will help, the killer is the mass (and how fast you want to accelerate it).

    As for the original question, I have heard that Nema23s are noisy compared to Nema 17s but have no actual data to back this up (just here say). Micro stepping does help with noise so maybe you could try increasing it?



  • Damn, sounds like you scienced the eff out of it lol Yeah I haven't heard any noise to indicate a bearing has gone out so it hadn't occurred to me to check rolling resistance which is why my focus is elsewhere for the time being. I'm currently running at 32x microstepping, so there should be plenty of headroom to see how much that helps for sure!

    Right now my X carriage is using a single NEMA 23 stepper to drive maybe a few pounds, while the Y gantry is moving that and a lot more using 2 steppers. The X movements are WAY quieter than Y. I even tried disconnecting one of the Y steppers to see if that would halve the noise, but it had a minimal effect which is what lead me to conclude that the weight was a pretty significant contributing factor.

    The driver temp aspect is a bit less certain and I definitely do some more tests to figure what exactly is going on there. For my last print I pointed a large blower fan toward where my normal 120mm PC fan and that "seems" to have helped, but I can't really measure a noise difference because of the extra fan, so I'll need to monitor the resulting temps instead. I'll probably get to that this weekend for sure!



  • I know budget is a key factor for you so clearpath sdsk is out of the question…. I wonder if Mechaduino (budget servo) can make the acceleration better on steppers.

    If you can afford it Clearpath SDSK are powerful, i dont really care what the mass is, these babies move and are dead quiet



  • @Whitewolf:

    If you can afford it Clearpath SDSK are powerful, I don't really care what the mass is, these babies move and are dead quiet

    I could feel my wallet shudder just looking at the price, especially x5 lol

    I know budget is a key factor for you so clearpath sdsk is out of the question…. I wonder if Mechaduino (budget servo) can make the acceleration better on steppers.

    I'll have to look into these when I get home.

    One thing I'm considering is just redesigning things a bit to relocate the bulk of the moving mass from the Y gantry onto the Z axis steppers and adding 2 more steppers for Z in addition to the 2 already there to cope with the extra weight. This would GREATLY reduce the moving mass on Y and will help with overall speed and reduce ringing quite a bit, but I'm not looking forward to spending another $500+ to get there. But, if upping the microstepping and applying a modest weight reduction doesn't do it, I'll look elsewhere. But alas, without any data to base my decisions on, I'm just kinda guessing based on what ought to make a difference 🙂



  • @efficientAF. I don't know if you saw this that I posted on another thread. https://www.youtube.com/watch?v=5RptHCX7z0o&feature=youtu.be. You can judge the noise level by listening to the extruders retracting.

    Admittedly my total mass is a bit less than yours at around 3kg (about 6.6lbs) but stepper noise just isn't an issue. The hot end fan is by far the loudest thing I can hear. This is all stock Duet. I have thermistors stuck to some of the stepper driver chips which switch fans on is they get over 45deg C but that had no effect on noise level. That is to say, the steppers are quiet anyway. I also do the same with the motors themselves to keep the temperature below 45deg C but that's to prevent the PLA mounts from distorting - again no difference to noise level noticed after I fitted them.

    Maybe it's just your Nema 23s?



  • That's an interesting printer design! Is there any artifacting caused by the upper part being tugged along once the slack in the ties is taken up?

    I probably should have mentioned this but I'm using Openbuilds NEMA 23s, I'm not sure where they rank in terms of quality but I got a decent discount on them so that's what I went with, but it might be worth revisiting at some point.

    It's probably worth noting that after I switched to a Duetwifi, it got significantly more quiet compared to running Smoothieboard and off the shelf external drivers, so I'm really just trying to get it as reasonably quiet as possible. The only reason I'm trying to get it even quieter than it already runs at is because my apartment has wood floors and acoustically, it made the most sense to put the printer in my room. TBH, I don't even know if it's "too loud" for my upstairs neighbor, I'm just trying to keep it from being a problem at all since prints could easily run night and day for a week or more depending on print settings. So far I've been printing at reasonable hours and for a few hours at a time. It's perfectly quiet up to a point when there is outside ambient noise drowning out what little sound it does produce, but in the dead of the night that ambient noise goes away and you could hear a mouse fart, especially if it was doing so all night long lol

    I'll try to get some meaningful data this weekend to share and get feedback on if upping microstepping doesn't do much.



  • @efficientAF:

    I could feel my wallet shudder just looking at the price, especially x5 lol

    Yeah I just bought 4 but I have a 60" welded steel gauntry to move around… You cant hear them and they are very powerful

    For my setup i used two for y axis, 1 for x and 1 for z



  • Jesus! How much does that weigh? My gantry is a smidge longer, but the bulk of the weight is from a 40x80mm aluminum C Beam spanning the distance. I have no doubt that those badboys will work well for me so I'll save that as a last step if need be, but I'm mostly just curious about how much that amount of steel weighs lol



  • Couldnt tell ya, the whole machine weighs about 500lbs it comes pre-assembled on crate without electronics….

    Here is my buildlog with photos:
    http://www.cnczone.com/forums/fineline-automation/336056-cnc-software.html



  • Ahh, that makes more sense that it's a CNC machine. I thought you had that on a 3d printer. That looks like a properly sturdy machine for sure!



  • Oh I am pretty sure I see something similar in my future with clearpath motors for a multiple gauntry independant xyz axis machine If Duet3d is willing to make some improvements to the board

    Check these out:

    http://www.titan3drobotics.com/the-cronus/

    http://projectescher.com



  • Thats a pretty crazy setup! I'm wondering what use cases they have in mind, I might have to look further into that concept. Right now mine is just a typical side by side dual so nothing super fancy like that.

    So I bumped it up to 64x microstepping and it didn't seem to do much. I got a short video jogging each axis a few times and you can hear the Y axis is noticeably louder, but you might need headphones because they're still pretty quiet lol. Ill run some tests tomorrow to get a sense of what hardware changes I can make to see if that helps.
    https://youtu.be/6EDsGkZgETM



  • @efficientAF:

    That's an interesting printer design! Is there any artifacting caused by the upper part being tugged along once the slack in the ties is taken up? …........................

    Not so far but I've only been using that particular layout for a week or so. Prior to that I had the extruders suspended above the centre of the bed, counter balanced by a container of lead shot attached via cord and pulleys. I didn't have any artefacts with that layout so don't expect any with this. The slack in the cords is deliberate to give some degree of decoupling. So with short zigzag infill moves and small circles, the flex in the bowden tubes takes up most of the carriage movement and the extruders are more or less stationary. I did consider having the extruder carriage on mirrored coreXY axes with their own motors but discounted it on the grounds of complexity and cost, as well as the fact that there would be no decoupling and all moves, no matter how small would involve the entire mass. Although it is generally best practice to keep the moving mass to a minimum, one advantage of high mass is that you don't get ringing - it's simply too heavy to resonate.



  • @efficientAF:

    Thats a pretty crazy setup! I'm wondering what use cases they have in mind, I might have to look further into that concept. Right now mine is just a typical side by side dual so nothing super fancy like that.

    So I bumped it up to 64x microstepping and it didn't seem to do much. I got a short video jogging each axis a few times and you can hear the Y axis is noticeably louder, but you might need headphones because they're still pretty quiet lol. Ill run some tests tomorrow to get a sense of what hardware changes I can make to see if that helps.
    https://youtu.be/6EDsGkZgETM

    So the concept is that most large format printers have to sacrifice quality for speed using Volcano hotends etc. Having 5 independant heads you can print in good resolution in less time as well as multi material.

    In the future the intent of the design is to be able to incorporate subtractive manufacturing as well as others.

    Think automated tool changes, lasers, dremmel, pick and place. It opens up the ability to do some really cool stuff.

    As far as your noise goes try this:

    With the machine off loosen any screws\bolts that secure and align the x and y axis. Then manually move the gantry all around. Then snug them up a little but not so much that it cant be adjusted still.

    Then power your machine on and jog the x and y to the minima maxima several times using the controls , then home all and slowly and evenly tighten each screw.

    I am not familiar with your setup but this process smoothed out my Davinci.



  • @Whitewolf:

    I am not familiar with your setup but this process smoothed out my Davinci.

    So is the idea to try and relieve any stress from tweaked or twisted parts so everything settles? I've basically done the first part already when I assembled everything, but I did what I could to make sure everything was squared and settled before tightening everything down. I didn't do it with the controls before locking it down, but I think I achieved more or less the same thing. Is this something that should be done every now and then?

    So are load/current and driver heat just not that much of a factor then? Is there no data for this? I know deckingman wasn't experiencing any of this with his setup, but I'm wondering if instead of a linear or flat trend line, if it plots more like a ramping curve and perhaps he's fortunately operating in the more straight/small change area and with the added weight, I'm on the up swing? Swapping in new motors would rule out bad steppers as a reason why the Y axis is louder than X, but I was under the impression that this is unlikely as steppers are supposed to be pretty robust.


Locked
 

Looks like your connection to Duet3D was lost, please wait while we try to reconnect.