Is there any data on Duetwifi noise factors?
-
Ah OK. The only other thing I can suggest then is to try a different stepper motor. I recently changed my extruder motors and although the old motors seemed quite, the new ones, which are smaller and lighter have made the extruders virtually silent. A side effect that I wasn't expecting.
Well I just double checked a few things and I think the stepper on the left side is either bad or not decoupled as well as the right side because it's still loud when I remove the belts and jog that motor. Disconnecting the left stepper and running the right produces a bit less noise so there is something up with the left one.
You said in an earlier post that you have two Nema 23s to drive the Y carriage. Given that I can drive my 6kg gantry using a single (but high torque) Nema 17, then I'd have thought that two of them would do the job in your case with your 10lb gantry. BTW, does it really weigh 10lbs or is that just a guess? Anyway, here is a link to the Nema 17s that I use http://us.omc-stepperonline.com/3d-printer-nema-17-stepper-motor-2a-45ncm64ozin-17hs162004s1-p-16.html. No guarantee they'll fix your issue but I can vouch that they are quiet on my machine (and only USD 8.90 ).
Since my machine is a basic Cartesian, I'm not sure how that would work to have a single stepper on Y given given the length of my X axis. One side only would leave the other side flapping away, otherwise I'd need a long belt or something to drive both sides with one stepper. Hopefully it is the case that it's just something off about that left Y motor, much easier to deal with. If so, then Y shouldn't be much louder than X, if it's any louder at all. I moved the gantry back and forth and there is definitely a negligible amount of force needed to move it along the rails and not audible grinding of bearings so it's definitely not a resistance thing. I did double check an older post where I had a more solid estimate of how much total weight needs to be moved for Y and it's actually 18.5 pounds lol 1 NEMA 23, 1 NEMA 17 and around 1800mm of C Beam aluminum extrusion, among other parts, but those are the bulk of the weight I believe.
That being said, my choice for these steppers was made out of convenience and to make sure I had more than enough torque with a 175oz-in NEMA 23 on each side to keep everything in line. It could very well be overkill but if I'm honest, I don't know how to find the optimal stepper, so I aimed high to be on the safe side
How could I go about calculating how much torque is actually needed and what would be a good safety margin? I'm sure that's standard engineer stuff, but I'm a 3d artist so I'm figuring it out as I go lol If I can get away with much lighter, lower torque steppers, I'd be down to make that change if that got the noise to be essentially zero. I think I'd need new mounting plates and pulleys though so I'll have weight that in also.
-
The post to which this was a reply has subsequently been edited / retracted so I'm doing likewise.
he he, yeah i misread too early in the morning… thought both replies were from the same person which took it out of context
-
The post to which this was a reply has subsequently been edited / retracted so I'm doing likewise.
he he, yeah i misread too early in the morning… thought both replies were from the same person which took it out of context
No worries. I've done pre-caffeine posts myself, then had a rethink.
-
@efficient AF. No, you've misunderstood me. I wasn't proposing that you only use a single motor. My thought processes worked like this. Mine is a coreXY, therefore for moves that have both an X and Y component, both motors are employed. However, with 45 degree diagonal moves (such as infill), only one motor is used. Now one motor is capable of accelerating 3kg at 670mm/sec^2 with no missed steps (possibly more but that is what I have my acceleration set to). So my reasoning was that two of the same motors could accelerate twice the mass - i.e 6kg or roughly 13.2 lbs which is greater than the 10lbs you said your axis weighed. So 2 of the Nema 17s that I am using on you Y axis instead of the two Nema 23s.
As for calculating torque required, well I'll have a stab at it but bear in mind that I am a 63 year old who's current career choice is to cut and screw bits of wood together. However, this is career number 5 (or 6 or 7) and I am an engineer by training and haven't forgotten everything.
I'll have to do it in metric because although it was feet and lbs when I was at school, we've been metricated in the UK for a long time and I've forgotten much of what I learned.
So, Force = Mass x Acceleration. Your 18.5 lbs is about 8.5kg. How fast do you want to accelerate it? Given that it's big and heavy, lets say 500mm/sec^2 (I'm using 660 and that's actually pretty nifty for fast non-print moves). So Force in Newtons is mass in kg x acceleration in m/s^2. which becomes 8.5 x .5 = 4.25 Newtons. Assuming you use a 20 tooth pulley, the diameter is about 12mm so the radius is about 6mm or 0.6 cm. So the torque required = 4.25/1*0.6 = 2.55N.cm. (rounded).
Sizing a stepper that will give you that is a bit tricky. There are a lot of factors involved, many of which I don't fully grasp, but taking a previous thread where I asked a similar question, and working backwards, we can make a few assumptions. Firstly, to be safe we'll probably not run the stepper at it's full rated current, more like 85% so the torque requirement becomes 2.55/0.85 = 3.0. Using micro stepping, we'll probably only get 70% of the holding torque so 3/0.7 now becomes 4.3. The torque drops off at speed so lets use a fudge factor of 50% giving us a stepper motor requirement 8.6N.cm. Now lets say we only get 20% of that for a single micro step which means we now need a motor rated at 43N.cm,
So by those calculation, to accelerate an 18.5lb mass at 500mm/sec^2 you'd need a stepper motor rated 43N.cm. If you want to accelerate it at 1,000mm/sec^2, you need a single stepper rated at 86N.cm. As you are using two motors, each motor need only be half of that (43) so two of the ones I'm using, which are rated at 40N.cm might just about accelerate it at 1,000mm/sec^2 but should easily do 500mm/sec^2.
As you can see, there are a lot of guestimates and fudge factors in the above stepper motor calculations and they could be wildly off. Also there may be things that I haven't taken into account like motor inductance or some such. Hopefully, someone far cleverer than I will step in and correct any errors. Alternatively, you could try emailing a stepper motor manufacturer and ask them which motor they would recommend to accelerate a mass of 8.5kg at 500 mm/sec^2 (or whatever acceleration you'd like to achieve).
HTH
Ian -
Ian, you've explained the torque calculation very well in your post, but I have one comment. Torque only drops off at speed if you don't have enough drive voltage. At https://duet3d.com/wiki/Choosing_stepper_motors#How_to_work_out_the_power_supply_voltage_you_need I show how to work out the drive voltage you need to maintain torque up to a target speed. So you can leave out the fudge factor if you check that your supply voltage is sufficient for the target maximum speed.
-
Ahh, I thought your printer was a CoreXY, but I wasn't sure. That certainly explains a few things!
So what you're trying to say is that my 123.5N.cm steppers are a bit on the overkill side?
I have some 42N.cm NEMA 17s that I could use if need be, hopefully the extra 1N.cm won't make much of a difference. They will require new mounting plates and pulleys w/ an ID of 5mm if I do switch. I'll try that if swapping out the bad one with the same kind to see if that fixes it. Since X is plenty quiet, albeit with a single NEMA 23 I'm optimistic that this will fix it, but if not I have a suitable backup (I think)!In any event, I really appreciate your help Ian! Your "guesstimations" are far more useful than mine given your background, that's for sure
I'll see if I can take your maths and see what the max load on my steppers are just out of curiosity. If I do end up going down to NEMA 17s it'll be helpful to know what a good use for the bigger ones are for sure! -
Thanks David. I'm pleasantly surprised that you only had one comment - thought I'd probably missed more. I'll leave the OP to do the maths to see if he has enough drive voltage or if he still needs to apply a fudge factor.
-
@efficientAF. Let us know how you get on. As I said before, I have heard that Nema 23s can be noisy but it's only hearsay - no direct experience of them. A good use for the big ones might be on the Z axis, especially if it's heavy and uses a course pitch screw with no gearing. At least they only move a small amount and then infrequently compared to X and Y so if they do make a racket, it won't be too intrusive:)
-
Will do! I think my initial choice was something along the lines of, "NEMA 17s are common on smaller lighter weight machines, so I must need something bigger"
At present I have 2 of the same NEMA 23s and 1/2' lead screws for my Z, but if I go for a more radical redesign, I'll be offsetting as much weight from Y onto Z in which case the extra 23s will be handy as I'd likely have 4 on Z at that point. However since there is no correlation between noise and weight, this change would be to speed up print time and improve quality/reduce corner ringing. It'd be a pricey change so I'll keep that one in my back pocket in case it comes down to that.
-
Will do! I think my initial choice was something along the lines of, "NEMA 17s are common on smaller lighter weight machines, so I must need something bigger"
At present I have 2 of the same NEMA 23s and 1/2' lead screws for my Z, but if I go for a more radical redesign, I'll be offsetting as much weight from Y onto Z in which case the extra 23s will be handy as I'd likely have 4 on Z at that point. However since there is no correlation between noise and weight, this change would be to speed up print time and improve quality/reduce corner ringing. It'd be a pricey change so I'll keep that one in my back pocket in case it comes down to that.
It's common for people to "overspec" their builds - my thought processes were the same when I started out. I've seen many, many designs which on paper seem to be the "dogs danglies" - huge motors, big fat multi start ball screws, big wide steel reinforced belts, or the other end of the spectrum with Cartesians using super lightweight heads, carbon fibre parts, etc. It's a bit like building a 4x4 hill climbing, off road capable vehicle or a super light weight high powered racing car for taking the kids to school on a flat road in a 30mph speed limit. Ultimately, we are limited by how fast we can melt filament, which in turn limits the printer speed that we can actually use.
Ref ringing on corners - forget it. With the mass you have, the resonant frequency will be so low that you won't get ringing (one useful side effect of having heavy axes).
