RF31 clone conversion and parts selection

[kregan]

I think I am going to go for the conversion of my RF-31 clone, a Jet JMD -15 round column mill/drill.

I will be using the plans purchased from Jeff Davis at: Learning Home CNC with some modifications to the parts listed on his plans.

I am very interested in the DMM Technology Corp. brushless AC servo's and controls at: Servo products
Mr Li was very helpful on the phone.

with 300W AC servo motor (57DHT04E) Standard NEMA23 - 48V, Max 3000rpm, rated torque: 1.14Nm(160 oz-in), peak torque: 3.0Nm(423 oz-in)

OR

250W AC servo motor (57DML2530L8LK) - 48V, rated 3000rpm, rated torque: 0.8 Nm(114 oz-in), peak torque: 2.38 Nm(340 oz-in)

The complete kit with all wires, power supplies, e switch, limit switches, drivers, 300 watt motors, encoders... Approx. $1100.00

The plans call for Geko's 320's, DC servo motors, and power supplies all sourced from different vendors. Being new to CNC I anticipate a few mistakes in sourcing parts form different places so I feel more comfortable going with a complete package like above.

Keling has a DC setup $1000.00 but the motors are 350 oz
Page Title

Ball Screws and nuts per the plans from Jeff are approx $230.00

Misc materials I will put at $200 - $300.

Software: MACH3, Autocad, Cam - ??? CamBam, Lazycam or something else.

Opinions and thoughts are welcome

[Bubba]

kregan,
Either one of the servo setups looks good for this machine. As far as the 350 rating on the keling motors, that is the peak rating! Do your design work based on the rated torque. The peak torque is basically a thou shalt not exceed rating or you let the magic smoke out.
As for the cad software, why not look at either progecad or draftsight (both are free) and they are look alikes for acad. For a cam, my personal preference is SheetCam (SheetCam homepage)
How are you planning on doing the Z axis? If you do the down feed mechanism, be aware there is quite a bit of backlash in the assembly. There are ways to minimize it, but I recently ran across a ballscrew setup done by Robin Hewlett (sp?) in England. Well thought out and uses a ball screw to operate the quill directly. The biggest advantage to his setup that I have seen over others is the ball screw is held quite close to the quill and minimizes the potential for flex in the mount to the quill.
HTH

[kregan]

The DC servo's have 50 oz rating and Jeff said they would work. His plans call for 350 or 430 peak servo's but he no longer stocks the 430 ones only the 350's. 430's seem hard to find in NEMA 23.

The AC servo's are rated much higher in normal operation.

I am really torn.... I really like the DMM setup but am concerned about replacement parts in the future. With the popularity of the DC setup with Gecko's its a hard decision to make.

Another advantage of the DMM AC setup is that the motors are powerful enough to do 1:1 on the x and y axis if I choose. I would only need to change the pulley from the 2:1 that the plans call for.

The Z axis is as you describe. A ball screw design that is inserted up into the head of the machine right behind the quill.

I already have Autocad for my architectural drawings otherwise I would be looking!

kregan,
Either one of the servo setups looks good for this machine. As far as the 350 rating on the keling motors, that is the peak rating! Do your design work based on the rated torque. The peak torque is basically a thou shalt not exceed rating or you let the magic smoke out.
As for the cad software, why not look at either progecad or draftsight (both are free) and they are look alikes for acad. For a cam, my personal preference is SheetCam (SheetCam homepage)
How are you planning on doing the Z axis? If you do the down feed mechanism, be aware there is quite a bit of backlash in the assembly. There are ways to minimize it, but I recently ran across a ballscrew setup done by Robin Hewlett (sp?) in England. Well thought out and uses a ball screw to operate the quill directly. The biggest advantage to his setup that I have seen over others is the ball screw is held quite close to the quill and minimizes the potential for flex in the mount to the quill.
HTH

[pete from TN]

If I were to go with a 1-1 drive on a machine of this size I would look to a stepper based system. Then you would not need the brackets for the belt drive just a shaft coupler and some simple brackets to support the motor in position. If you decide to go servos then I would try to use some sort of reduction for speed of the motor. Just because the motor has enough torque to go 1-1 the nature of a servo is that it is better at high speed and has more torque due to that. If you set it up with a servo and belt drive I would go at least 2-1 as long as your ratios are right with your chosen max rapid speed and ballscrew pitch and other factors. If you want to go AC servos I seriously doubt you will have to worry too much about buying replacement motors for a long time. Those things are built to last and even a DC servo unless you do something stupid is gonna last a long time especially if you are not planning on any sort of serious production. I would say go with what you are comfortable with. I used servos and geckos and geckos are pretty nice and they have EXCELLENT company backing and tech information and they are not too expensive in case you manage to fry one or two like I did setting this thing up. The AC servos theoretically will be more powerful and faster, your positioning accuracy is more a factor of your encoders and drives and ballscrews than anything else. It really comes down to how much you wanna spend. Another thing to consider is that you can buy complete controllers already assembled from several vendors including Bob Campbell of Campbell designs as well as several others. If you are really wanting to get rolling with as little issues as possible and can afford to not DIY it, there are several options here. Good luck... peace


Pete

[Bubba]

Something else to think about if you go 1:1; are you sure you want rapids of 600IPM (.2*3000)!
I have done 180IPM on my RF31 and had to change my drawers:}) I would worry more about having the grunt to accelerate quickly rather than insane rapids. I am not aware of what you plan on doing with this machine (although might guess from you avitar:}) but you really need to look at what you NEED the machine to do and base your design on that.

Just trying to help.

[Maglin]

I have only been able to find one person using the DMM setup on youtube. It looked really nice and I almost purchased it but I was looking for something very specific. AC servo's are all brushless design and DC brushless servo's operate almost the same way. DCBL will output less power than the same servo ran AC cause their is 3 fields in constant flux in AC and DC only operates 2 fields at a time and times which fields to run. What I'm saying is a far as DCBL to AC servo's are concerned the AC servo's should provide more power for the same servo. The one thing that worried me about the DMM is the encoders are 14bit corrected on a 16bit encoder. I'm not to sure if you can swap out the encoders with say the CUI encoders and have the drives still operate properly. I might have been wrong in my thinking but it was the main reason I didn't go with that system. But the router that I saw using them had some large travels and accuracy or lack their off was not mentioned.

If you went with the DMM ask if they would offer a refund if their products don't give you the accuracy that you need. Being +/- .001 on the X. The rest of the travels are so small that I'm sure it will work great. Also you get everything you need for the electronics including limit's for the DMM setup. It's a step/dir setup out of the box and the datasheet for the drives looks like they will accept a number of different control schemes.

[kregan]

600 ipm... LOL That would be scary!

Per Jeff Davis's advice I contacted Camtronics looking for the 430 oz peak servo's and was told they are no longer available.

That leaves me with a simple choice of the 350 oz peak DC brushed servo's from Jeff or the DMM setup.

This is really for hobby parts and just to learn on. I don't have dreams of opening a machine shop or selling parts. I just want the ability to make things I normally could not do by hand cranking.

Thanks so much for all your help so far, I have learned a lot in the short time I have been a member here.

[ninefinger]

Kregan,

The DMM setup looks to be a great buy. The encoders are 12 bit accuracy which is 4096 counts per rev, which will provide more than enough accuracy, even if you went 1:1 drive. I would stick with your 2:1 reduction plan just to be able to realize better accelerations due to increased torque.

The 350 oz-in peak DC brush servos will work too, thats what I have on my machine. I'm running them at a 4:1 reduction to get the torque up there and have good acceleration.

The brushed servos are ok but I did have a small problem with one of my servos (my fault), and the brushes can wear out. The brushless desigh looks to be nearly indestructable with lots of good failsafes built in (over current, overheat, over voltage and lost position protection).

Also, don't overlook the value of all the parts being from the same "manufacturer" as they all are chosen to work together nicely, that break out board looks like its plug and play with the servo drives - thats a **huge** time saver right there.

Mike

[kregan]

I am waiting on a final quote including shipping to Maryland from DMM. Once I have that in hand I think I pull the trigger.

I also contacted linermotion2008 on Ebay to get a quote for the ballscrews cut per the drawings and ball nuts.

I have an Enco lathe but would pay them to machine the ends of the ball screws for the right price.

Kregan,

The DMM setup looks to be a great buy. The encoders are 12 bit accuracy which is 4096 counts per rev, which will provide more than enough accuracy, even if you went 1:1 drive. I would stick with your 2:1 reduction plan just to be able to realize better accelerations due to increased torque.

The 350 oz-in peak DC brush servos will work too, thats what I have on my machine. I'm running them at a 4:1 reduction to get the torque up there and have good acceleration.

The brushed servos are ok but I did have a small problem with one of my servos (my fault), and the brushes can wear out. The brushless desigh looks to be nearly indestructable with lots of good failsafes built in (over current, overheat, over voltage and lost position protection).

Also, don't overlook the value of all the parts being from the same "manufacturer" as they all are chosen to work together nicely, that break out board looks like its plug and play with the servo drives - thats a **huge** time saver right there.

Mike

[kregan]

Update,

I pulled the machine apart to check some measurements and found that my tables are a little shorter then the machine Jeff D. made the plans for. Not a big deal but I am glad I checked before ordering the ball screws.

As of tonight most of the parts are ordered... Ball screws, ball nuts, bearings, pulley's, belts, and a bit of aluminum that I did not have on hand.

I purchased the ball screws from linearmotion2008 on ebay and paid for the end machining for a whopping 24.00.

I am up to about 650.00 including shipping for all the parts except the servo's and controllers etc.

[kregan]

OK I ordered the DMM servo and driver package.

[pete from TN]

Let us know how it goes with them... should be cool!! Good luck with your build. holler if you need any help!! peace

Pete