Getting rid of resonances
Is there more ways to deaden resonances that cause horrible racket that just microstepping and cutting down the speed? (if not getting into frame construction)
I have my workshop in a residential block and would like to keep the relationship nice with my neighbour.
The heavy Y axis still makes noise after 32 microstepping (tried 62 and 128 too, no differences in the sound) and as everybody who has tried to make their printer silent knows, just dropping the speed can raise another resonance on different speeds.
burtoogle last edited by
Mount the printer on a heavy base and sit the whole thing on top of an inner tube?
What sort of "horrible racket"?
You can dampen resonance by adding mass. Maybe not the thing to do on a Delta but for other printers it's fine. I know it sounds crazy as the entire 3D printing community (apart from me) has a fixation about building the lightest possible hot end assembly but trust me, my 2kg hot end (4kg including the separate extruder gantry) does not resonate and is still capable of printing at far higher speeds than I can melt filament. My default speed setting for non-print moves is 350mm/sec and the noisiest thing by far is the hot end fan.
But what is making the noise on your printer? You say it's the Y axis so I'm guessing it's the guides or guide bearings? I'd say that if 128x micro-stepping doesn't help, then you need to look at the mechanics.
The only machine I've ever had which had nasty noises was my corexy, and the way motors were mounted made a big difference. Belts shouldnt make a lot of noise if you run smooth sides over any small idler pulleys, (or use toothed pulleys) try cork motor damper mounts, but I'd avoid those mounts with loads of rubber bushings in them they must introduce backlash in some way.
The resonances come directly from the nema 23:s that are bolted to the frame. It´s sort of a big build so it has mass already. I would love to make the enclosure of heavier and better dampening stuff than acrylic but all the enclosure is made to be opened easily (every wall is a door to make it easy to service, fix and develop), so there are limits with adding mass. I will add glass to the doors next to the acrylics to get a air gap for better insulation, so it will surely help a lot.
I Was just poundering if I could do something for the source of the problem instead of trying to cover it afterwards. Maybe I put thin sheets of rubber between the frame and the motor mounts.
…........................... Maybe I put thin sheets of rubber between the frame and the motor mounts.
Sorbothane sheet is an excellent choice for that. You can get it from good HiFi shops (used to dampen speaker enclosures and the like). It's not cheap though.
number40fan last edited by
Have you tried lowering the voltage on the steppers?
No I haven´t tried lowering the voltage. Wouldn´t that make lost steps more probable? Have to try it today…
I don't know about lowering the voltage but lowering the current will make motors quieter, and may be quite sufficient for your needs, it may even be possible you can lower the current to levels where a nema 17 would work just as well, and that will reduce noise even further, you say its a big printer but there are countless examples of quite large machines running from nema 17 motors. Are you using the interpolated 1/16th mode? This does seem to offer an excellent middle ground between having very high stepping in higher modes and the load on the processor and noise reduction.
Im running without interpolation. Clueless about how that affects. What line sets it into something worth trying?
I have the Y motors in series in lack of drivers right now but around christmas will set up the Duex5 so then both motors run on their own driver. Somehow I feel that will help also. Those masses in movement do effect on each other probably in surprising ways…
To run interpolated you set I1 at the end on your M350 command - I0 turns it off. AFAIK, the interpolation is done by the driver chip itself.
I have heard that Nema23s are noisier than Nema17s but have no direct experience of that because a single Nema 17 is more than adequate to throw my 2 kg hot end around. Admittedly mine is a CoreXY and some would say that I'm using two motors but for 45 degree infill, only one motor is employed. My default speed for non print moves is 350mm/sec at 1,000 mm/sec^2 acceleration. My total mass in Y is around 4Kg but that's due to the 5 Titan extruders which are on a separate gantry so for 4Kg, I do indeed use two Nema 17s.
Well I could not wait for the holidays to connect the Duex5 and it looks to have been a lot simpler task than I expected. last time doing custom mapping I almost lost my mind XD
Soon to set on a test print, but just by trying the Y homing it looks a lot quieter already!
Sent M350 I1 and toggled a couple of time with the on off, but no difference in sound…
You need to specify which axes you want to use interpolation on, e.g.
M350 X16 Y16 Z16 I1
However, x16 microstepping with interpolation is the default, so unless you already had a M360 command in config.g then it would already have been enabled. Interpolation only works with x16 microstepping.
All right, I'll give it a try tomorrow!
X is set up to 16 microsteps and is suuuper silent. Maybe it could work on Y since now the 2 motors are connected to their own drivers. Before (motors in series) 16 ms was a bit louder than 32, allthough might be placebo effect and the same sound in reality.
stellator last edited by
Hi, interpolation really only works with 1/16 microstepping? So I1 is just beeing ignored when i set for example 1/64 microstepping?
Do have a link where the effect interpolation is explained maybe?
thx and good night !
The TMC2660 datasheet doesn't make it clear whether interpolation works with microstep resolution other than x16, so I asked Trinamic. They told me that it works with x16 only, and that the interpolation bit is ignored when using other microstep settings.
I've tried this, I switched my extruder to 1/4 and it was not interpolating, and it meant my other axes were not either, noisier but not as much noisier as I expected. It reduced my missed steps for extruder which I was getting but then with 4200 steps/mm on 1/16th, and a 30:1 gear reduction, a few missed ones is not going to affect print quality so I'm back on interpolation for all axes.