FYI, the 8 hour time weighted average noise exposure limit for general industry is 85dba. So 85dba over 8 hours. But peak can be higher. Keeping it under 100db is a good idea.
FYI, the 8 hour time weighted average noise exposure limit for general industry is 85dba. So 85dba over 8 hours. But peak can be higher. Keeping it under 100db is a good idea.
Okay, so I measured my machine while making a finishing pass on a bore with a 2 flue 3/8" end mill at 4500 rpm and 27 ipm in 6061 t6. It was 81dba average with my furnace running. Surprisingly low. A roughing cut would be way louder but suffice it to say that 80dba is achievable while cutting.
I have been rebuilding my G0704 spindle with the goal of reducing noise and heat as well. I am very happy with my results and have gone from a ear piercing scream down to a comfortable 75dBA at 1' using an iPhone app. The things that I did that seemed to help the most were:
-Replacing shot bearings with new AC bearings. Used Kluber IsoFlex NBU15 grease @ 30% fill.
-Set preload to a correct amount. Original preload was excessive
-Installed a plastic tube between the spline and the inside bore of the quill tube. This was a light press fit and stops all rattling
-Changed timing belts from L series to GT3-5mm series with a nylon faced belt. The smaller teeth significantly reduced the running noise.
-Better machining of new timing belts decreased radial runout from .022" to .001" (The first pulleys were hand shimmed to run "true"
-Machined a 1 piece spindle motor mount which seals the pulleys in a nearly air tight shell
The machine is very comfortable to run now, although the bearings still get warmer than desired. I may need to reduce the spindle preload.
Attachment 412806
Attachment 412808
Attachment 412814
Attachment 412816
A pair of new pulleys is on the way. Hope these make the mill a bit quieter.
@CL_MotoTech 81dB is beyond reasonable while cutting. I'll be able to see exactly where I'm at once my new pulleys arrive. My guess is it will be about that NOT cutting but we'll see. My servo motor may just make more out the gate noise than typical AC motors (I think that's what you're using, right?).
@mcardoso Are you using the stock preload nut? I think that's the biggest issue with finding the correct preload. It seems like it's hard to find a middle ground between not having any play in the shaft axially or radially and have it turn smoothly. I may try to make myself a new preload nut and try threadmilling for the first time since the spline is a weird thread IIRC. It also looks like you're using a servo motor. Do you know how loud that gets by itself? E.g. No pulley or belt attached. I noticed something like a 17dB sound difference between ambiance and just that being turned on. It wasn't loud by any means but it seems to be a pretty big jump. I was wondering if maybe AC motors are more quiet than servos but maybe not if you're getting 75dB connected to the spindle. I think I was using XL timing pulleys/belts, whichever Hoss recommended in his plans that you could order direct from SDP/SI. I'm sure those are a large noise maker in the equation.
Side note, it looks like you have or are gearing up for a power drawbar. Hoping I can join that club myself sometime soon!
75dB seems like a lot for just a motor spinning. What does the shaft feel like when you spin it? Can you get a video of it running? Seems to me that the heart of the matter is your servo motor. This is is supported by the fact that its louder with a cast iron pulley - a pulley with no belt shouldn't affect noise at all. Loading up a component that is noisy is just going to make it more noisy, its going to transmit its noise through to the connected components. What type of servo motor is it? BTW - its possible to have bad bearings on new motors - rare but it happens (depends on who made the motor...).
Do you have the original motor for comparison?
It's a DMM servo motor, .75kw. Both the original motor and servo motor feel relatively smooth. The original BLDC feels smoother with less resistance but I would think the servo shaft should naturally have more resistance? Waiting for a new motor driver that's supposed to get here tomorrow then I can get a video of it running. I would double check the noise on the original BLDC but I'd have to rewire it since I had to cut the wires to get it unmounted from the mill. I believe it was pretty quiet though.
I thought it could also be possible the cast iron pulley wasn't rated for that many RPMs so it could be out of balance and cause noise.
Hi Steelcogs. How are you controlling your DMM servo? I just ordered the same one for my machine, and I'm not 100% clear on how to connect it so Mach3 can control it. Any information or links to other threads would be very much appreciated. Thanks! Mike
I assume you're trying to use it for a spindle motor too? You'll want to use 0-10v analog control to control it. I believe that's the only way you can use the speed servo setting (high RPM) with the DMM servos. Depending on your breakout board you'd want to look for 0-10v output then wire it to the corresponding pins on the DYN4 driver wich can be seen on page 11 of this: http://www.dmm-tech.com/Files/DYN4MS-ZM5-A18A.pdf
Steelcogs:
The 75dB is 5000rpm when attached to the entire spindle assembly. The motor by itself only registered like 45-50dBA. Wisper quiet. The only sound you hear is the 4kHz carrier frequency that all VFD's and servo drives make. If you are getting a lot of motor noise, you may need to work on your tuning gains. The sound is different than an AC motor but shouldn't be significantly louder.
I found that the L and XL series timing belts tended to suck air into the teeth and make a loud pop when it exited. This got much worse the faster you went. This also excited vibration on the span of the belt adding to the noise. I spend a few hours on the phone with SDP-SI and was left with the recommendation to use the curvilinear GT2, GT3, and GT5 pulleys with fabric faced belts for high speed applications. Also to keep the tension to the minimum acceptable for the application. I have been extremely happy with this so far.
Fusion_crank:
The DMM drive can be run 0-10VDC using an appropriate converter board OR you can run it as a step/dir servo through Mach 3 (just like a stepper). This is what I did on my original design running a DYN3 drive and it worked perfectly.
@mcardoso I think you're right about gains. Are you also using a DMM servo? The DMM ones like I'm using have an auto-tuning feature and I definitely noticed a difference between redoing an autotune on it yesterday (I did one when I first got it a few months back). It was only a slight difference but definitely different. Is there a resource on tuning gains out there? I wouldn't have the first clue where to start.
Side note, I got the pulleys that @maxspongebob made. They turned out great! Still waiting on proper belts to show up but they both fit good and look great. For sure need to look into motor noise though.
As a comparison I just ran a test on my SIeg X3 with GT3 toothed belts and pulleys from SDP/SI and a DMM 1.8kW servo for the motor. I read 76db max a one foot from the spindle.
@rcheli Did you autotune your servo or manually tune it? This is what I'm looking into next since everybody seems to have pretty quiet motors.
@steelcogs, in this machine's previous life, it ran a 0.75kW DMM servo with their DyN3 drive. No autotune sadly! Still it ran pretty well. Servos will pick up on the vibrations from the belt and either damp or amplify them based on tuning. Run your motor unloaded with factory gains. It should be fairly quiet. All my motors are AB MPL servos and Ultra 3000 drives.
As far as advice on manually tuning, that is a very deep subject. A very VERY high level overview is set all gains to 0, raise the P gain until unstable then back off 10-20%. Then raise the I gain to compensate for steady state error, finally add a very small amount of D gain to damp the system. If your drive has filters or skip bands then things get more complicated still. It isn't hard, but you have to be ok spending an afternoon making some terrible sounds and learning what each gain does to the system. This needs to be learned hands on. A rough starting place for the I gain should be ~ 1/10 of your P gain, and the D gain 1/100 of your P gain.
I did run the auto-tune but I'm not really sure what it did. When I started running it in Velocity mode it would start and stop real abruptly. I contacted DMM tech support for advice on settings and this is what they said. I also installed the resistor an line reactor as recommended. I have a screen shot of the settings somewhere. If I can find it I will post it.
--------
For use as spindle motor, set Integration Gain to "1" and keep it at "1". Then increase Main Gain and Speed Gain until the motion is rigid and suitable. Keep Main Gain and Speed Gain at the same value.
Make sure to install a regenerative resistor on the drive to help absorb deceleration energy from the motor. Also install a transformer or AC line reactor on the AC power input to smooth and filter the power coming into the drive.
@mcardoso Oh interesting, didn't realize it was PID tuning. I actually did a little bit of PID tuning with racing drones a couple of years ago. I remember how a lot of those settings work but only to how they apply to drones. The DMM software has "Main, Speed, and Integration" gains. I assume Main would be P, Integration would be I, and Speed would be D? Does that sound correct or no?
@rcheli Ok I'll give that (what they said) a go tomorrow. Not sure how much of that exactly transfers from a 1.8kw motor to a .75kw but I'd imagine it should be the same. I still need to add the resistor/input reactor but I'm just trying to do things one at a time (plus I don't want to mess with 240). I can't imagine either of those would affect audible motor noise.
Ahh yeah, I remember DMM called it some weird things. I think that is right? It has been a few years. Honestly just write down your starting values and start playing with them. Tug on the spindle when it is stopped to feel how well it holds position. The run it at speed and listen to the sound. Very loud noises are oscillations, slow surging can be compensated with integral gain, bogging down under load can be compensated with Proportional and Integral.
A shunt (braking resistor) is important when decelerating a load. The larger the motor the more the need for the resistor. Not strictly necessary, but you can fault the drive when decelerating without one. A line filter or reactor is nice, but not required, especially on residential wiring. It helps smooth incoming line which can be needed when other equipment start and stop nearby on the line.
Yes, as mcardoso said the braking resistor is important particularly if you are going to be setting acceleration so that the spindle starts and stops quickly. IIRC the acceleration setting controls both. Until you get the resistor installed don't get too aggressive. I'm not using position mode but when they talk about stiffness in velocity mode I think that meant closely it tries to follow the commanded speed. I remember adjusting (The gain I think) it from a point where if I grabbed the spindle while it was running and tried to slow it down it would get harder to slow as I increased the setting.
They said the reactor would also help with stopping noise from the servo from feeding back into the power line. I didn't have any issues (that I knew of) before I installed the reactor but I was trying to go decked out on my build where I could. I also used all of the line filters they specified in the drawing in the manual.
@rcheli Is velocity mode also called “speed servo?” That’s the closest I could think of to a velocity mode.
I have no intentions for needing a fast accel/decel so I think I MAY be alright there without a resistor. Was planning to sort of play it by ear and if I started having alarms then add those extras to the circuit.
Yes, That is what DMM calls it. They have Position, Torque, and Speed.