Nomad spindle motor....upgrade?

That’s one of the reasons I was asking if anyone has done it. The controller logic should be ok (same phases, same rpm curves), but the drivers might be a little underspec’d for the additional current (4A vs 6A). I haven’t pulled it apart to look at the drivers.

@mikep, C3D has done it. Check posting Nomad 883 Pro - any pictures or specs for this machine yet? where Rob said they didn’t think it was worth it.

Randy

1 Like

Which makes sense for a change to the product. I’m sure they’ve got inventory and contracts for the old motor, would have had to spare both, they would have had to increase the size of the enclosure to keep the 8x8x3 envelope, potentially stretch the uprights 1/2 inch or so, I can see all sorts of reasons they would have called it “not worth it.” Can’t see how another 40% more power out of the spindle isn’t better for what I’m trying to do (aluminum and brass) - the chassis is plenty stiff, I can deal with a few more .001" of tolerance without any issues, I’m willing to give up the 1/2" of working Z, and the inevitable hiccups along the way.

1 Like

70W doesn’t seem like it would be worth the effort of making the change. Until they can pack a 500+W spindle in there I probably won’t be interested in the effort.

Looks like it’s using this drive, which is rated for 6.5A peak, 3.5A continuous per phase (per phase output, so 3.5 * 3 = 10A total)

DRV8312
6.5A Three Phase Brushless DC Motor Driver with Inrush Protection (PWM Ctrl)
http://www.ti.com/product/drv8312

@mikep I too would be interested in this. Are you saying the motor is 1/2" longer so that the top of it would crash into the Nomad Pro’s top panel on retract or do you have a Nomad Classic were the motor orientation is upside down? I would be willing to modify panel for 40% extra power. Also, besides removing pulley and other mechanical stuff, would the motor and driver board be plug and play or would they require some electrical/software mods? I realize you are asking too, but so far you seem to know the most about this.
Thanks,
Jerry

Here’s an update on where I’m at:

I have an 883 Pro. Yes, the motor is about 1/2" longer than the standard one. If I use a slightly different “limit finger” (the black thing screwed to the right side of the z carriage) I can avoid changing the enclosure, at the expense of about 1/2". I’ve done that, and it’s working fine. I made mine out of scrap of 1/8" thick brass I had laying around. The other option is to cut out for clearance from the top of the enclosure. For just trying this all out, I’d rather lose the Z-height than put a hole in the enclosure. I think I might make a new enclosure later.

The wires on the existing motor are quite long. The wires on the replacement motor…aren’t. If I join these with a simple solder splice I can’t 1) switch back to the old motor easily 2) replace the motor when the bearings eventually fail without a bunch more work 3) if I join them past the tie down point, they’ll certainly break at the solder joint. So the plan here is to put in a connector in the section between the motor and the point where the cable is tied down to the carriage so there isn’t any flex in the cable at the connectors and their solder joints. I received the connectors late last week. I’m planning to use a solder cup style DB-9. It’s compact enough to fit here, has enough conductors for the hall sensor assembly and the motor phases, and they’re cheap. Instead of a standard backshell, I’ll pot the backside of the connector with a little hot glue.

I -think- the motor driver should be fine, it seems to have enough current to support the larger motor, so that means that electrically this should be plug and play. I kind of wish they had used a different package for the motor driver IC. There are two packages for this part, one is “right side up” and uses conduction to the underlying board for cooling. The other package is “upside down,” has a nice spot to put a heat sink, and can support more current. Should be fine with the one that is there. I think. We’ll see.

I’ve got everything I need at this point, just need to take some time and get it worked, and time has been a little short the past couple weeks. Should make a little progress this week.

Will post here with successes and failures…

Mixed success:

Here’s whats done so far…

Connectors applied to the new motor, old connection cable.
The larger motor from Wantai doesn’t have a flat on it for the set screw like the original motor. So I added one.
New motor mounted, belt installed.
Spins up under software control, changes speed, shuts down. Run up to 10k rpm, down to 2k.
New stop arm attached (lowers Z-max about 3/8 inch or so) - motor clears top of cabinet at max Z.
All axes have been traversed with the motor on and off, seems to be fine.
Have not cut any material with it yet. Need to solve my next issue for that…

I -am- having trouble with homing for some reason. Almost seems like the machine coordinates are messed up, not sure if it’s directly related or not. Any attempt at homing gives me an alarm about hard/soft limits. Still working that. I had to move the table a few times, and am a little worried I broke something with back EMF from the steppers.

Awesome. Thanks. That is most encouraging. One thing that worried me was at one point during a cut in steel my spindle stalled. I couldn’t get it to retract/home or anything else. The job was literally 99 percent complete when this happened. Since i just needed a small plunge cut, I was going to see if I could just do it manually. I moved the table while the power was off and this seemed to generate electricity, because the light on power putton flashed as I was doing this. Nomad has been fine ever since, but there are definately some electrical gremlins taking up residence,

Have it all up and running. I managed to break the Z-limit switch…don’t do that. That’s more of a pain to diagnose than you would think. Fixed now, all working great.

Will post a how to soon.

1 Like

Ok, as promised, here’s the how to. Could probably use a little editing, but it hits all the high spots.

-----How to upgrade the spindle motor for the Carbide3D Nomad 883 Pro to 70 Watts -----

Keep in mind that you are modifying a potentially dangerous
piece of equipment with spinning sharp things. Your mother would be very
upset if you hurt yourself. Stay safe.

This most certainly will void your warranty.

If you are expecting a night and day difference doing this upgrade you
will be disappointed. It makes a difference, it does not make a huge
difference.

You lose some Z travel doing this modification using the standard
Nomad 883 Pro enclosure. If you are using a different enclosure you may not
need to reduce the Z travel. The work envelope doesn’t necessarily change,
but may mean you need to use shorter tools for work near the limit.

This is how I did this process on my 883 Pro. Carbide3D is always making
updates and improvements in their products, and the material here may not be
accurate for newer or older machines.


Tools required
2.5mm hex wrench/driver, at least 3 inches long. T-Handles are not recommended.
1.5mm hex driver
A large flat bladed screwdriver
Hobby knife. Not a utility knife.
A soldering iron

Materials Required
1 Wantai Brushless Motor, model 42BLF03. (Find on ebay for about $30 plus shipping)
This motor is rated for 70 watts vs the 50 of the original. The 42BL motors
match the stock motor mounting holes. This motor is however 16mm longer (79mm)
than the stock motor. The driver IC should be able to handle this with no
issues, and there is plenty of power available from the PSU.
16mm ~ 0.63 inches
You should consider the higher powered motor as a consumable item as it most
certainly does not have the higher quality bearings that Carbide3D specifies
for the original motor.
Solder
Loctite - Medium.
M+F connector of choice. I used a solder cup style DB-9 without backshell. It
isn’t strictly necessary to use a connector, but I did.
A piece of ~1/8" thick brass or aluminum material to make a new limit stop. Use
what you have at hand, but it should be at least as thick as the existing stop.

----------------------------- How to ----------------------------------------

Remove the power plug from the wall and disconnect the USB cable. Both. Really.
The controller board will stay powered if the USB cable is plugged into
a PC, even if the main power is turned off. You can damage the pins on the
controller or the wiring if your soldering iron is grounded (which is
the normal way they are built) and the controller is powered on.

If you are using the standard Nomad Pro enclosure or a replacement without
additional height above the carriage you need to make a new limit stop:
Remove the limit stop from the right side of the Z-carriage with the 2.5mm hex
driver. Use this as a pattern to make a new one that is at least 17mm taller.
Install your new stop. Push the Z axis all the way to the top, and be sure the
new stop engages the limit switch. Do not carelessly slam the Z carriage up into
the switch while fitting and sizing or you will break the switch internally.

Hold the new motor next to the old one and be sure it clears the top of the
enclosure. If it does not note the additional space you need and make a
new stop. This is a pretty easy part to make so just do it by hand.
If you can’t make this part by hand…you probably shouldn’t be
trying this process. You could also just drill and tap new holes for the
existing stop, but it’s very inconvenient to do so accurately without a lot
of disassembly. Seriously: Do not carelessly slam the Z carriage up into the
switch while fitting and sizing or you will break the switch and it is not
obviously broken.

Use the 2.5mm hex driver to loosen the 4 cap screws that hold the motor to
the carriage. The back two can be easily accessed through two holes
in the underside of the Z carriage. The motor should be easy to move a few mm
closer to the spindle which makes the belt easy to remove.

Remove the belt from the pulleys and carriage.

Use the 1.5mm hex driver to back off the set screw from the large pulley. It’s
way down in there and may not be visible. You don’t need to remove the set
screw, you just need to back it off 5 or 6 turns. If you remove it you will
drop it. Its easy to lose and really tiny.

Use the large flat bladed screwdriver to lever off the big pulley from the motor.
If you’re smart you’ll put a (metal) putty knife between the shaft of the
screwdriver and the underside of the motor mount on the carriage so you
don’t leave a mark like I did. This should take a little effort but if it’s
LOTS of effort check the set screw again, think about what you are doing,
and don’t break or bend anything.

Remove the 4 cap screws you loosened earlier that hold the motor to the
mount with the 2.5mm driver.

VERY CAREFULLY and SLOWLY, with multiple shallow cuts, cut through the heat
shrink at the end of the cable cover braid lengthwise. Do not use a utility
knife for this task. The shrink tube is adhered to the braid. Be VERY careful
not to cut the wires or insulation. Peel the heat shrink off the end of the
chafe braid.

Peel back the chafe braid. It will just unwind from around the wire bundle.
You can now cut the wires about 1.5" or so from the motor. This leaves enough
to put a connector onto the old motor so you can put it back in if you need to.

Prepare the new motor. My motor did not have a flat on the shaft to accommodate
the set screw, so I added one with a file. The shaft material is pretty soft.
The original motor has a flat. If you make the flat too deep the set screw
will be overdriven and get jammed between the inside of the pulley and the shaft
which is “a problem.” You don’t need a deep flat, just enough for the set screw
to catch. The flat on the original motor shaft is much deeper than necessary.

Wire the motor, matching the wires. Be sure to match the group of 5 wires to
the group of 5 wires and the group of 3 to the group of 3. The group of
5 is the hall sensor assembly, and has colors that are the same as those in the
phases. These wires are not interchangeable. If you want to use a connector,
do that. A simple 9 pin DB-9 works fine, but many different connectors will
work. The wire is 26 gauge, so be sure the connector/pin you
choose can support wire this small. I -strongly- recommend you do not use
automotive type butt connectors. If you use a DB-9, pot the back of the
connector with hot glue to help with vibration/fatigue. If you need to do so,
hot glue is easier to remove than silicone. There really isn’t room for a
normal DB-9 backshell.

Be careful that you leave enough slack so that the cable doesn’t bind or pull
hard against the geometry of the carriage anywhere. Any solder connections
should be between the cable-tie clip and the motor. Do not solder anywhere that
will flex later.

Install the motor and tension the belt

  1. Place the motor into the carriage without any screws.
  2. Put the pulley back on, aligning the set screw with the flat in the shaft.
  3. Align the pulley vertically with the small pulley on the spindle
  4. Loosely install the motor with the cap screws. This is a little fussy.
    Don’t apply loctite at this time.
  5. Put the belt back on, and use it as a reference to adjust any pulley
    misalignment.
  6. Reinstall the set screw on the big pulley. It should be firmly tightened,
    but not “Gorilla Tight”
  7. Push gently on the motor away from the spindle cartridge to apply tension
    to the belt. Tighten the motor mount screws. Light belt tension is
    sufficient.
  8. Remove one motor mount screw at a time, apply Loctite to each screw and
    replace them.
  9. Check your wiring carefully one more time. Check for “whiskers” between the
    connections, shorts, melted wire insulation, etc.

Plug it in, fire it up, start the spindle. If it makes noise, the tension is
too high on the belt, so back off the screws and try again. If the motor
doesn’t rotate, check the wiring, check that the wires are firmly connected.
Run the spindle at high and low speeds (S10000 M3 - start spindle at 10k rpm,
M5 stop spindle)

I recommend wrapping the connector area with silicone tape (“Rescue
Tape”) to seal it from swarf and other material that might be kicked up.
Electrical tape is not a good substitute in this case. Then cover as much as
you can with the chafe shield you unwrapped at the beginning of the process.
If you cut off the zip-tie earlier reinstall a new one now. Do not operate
without that zip-tie in place, or the wires will fatigue and break off at
the connectors or internally at the motor over time. Be sure to leave enough
slack so that Z travel doesn’t pull the cables hard against an edge.

This is how I did this modification. It might or might not work for you
and you might have better ways to accomplish the same end. This worked for
me and I’m happy with the result.


2 Likes

For anyone that is interested, After extended running (about an hour) the spindle cartridge is running 102F, the body of the motor is 136F, and the top cover (where the hall sensors are) is 99F. The back aluminum panel that serves as the heat sink for the boards is 92F. The stepper bodies I can reach are 104F.

I’ve just installed the upgraded spindle motor as per @mikep 's excellent tutorial with the following revisions, omissions and additions. I have the late model Nomad Pro.
I didn’t want to modify limit switch or lose Z travel so I am designing a new enclosure top section and window solution. One thing I noticed was that my new motor is spec’d at 78 watts, not the 70 I keep hearing about.

  1. I used D-Sub connectors with the backing shells (I liked the industrial look.) It is true that there normally wouldn’t be room for these due to how they interfere with the Z carriage and frame uprights. I worked around this by pre-forming bends at stress points using Spikeez, which is a fast curing silicone marketed to tween age girls for jewelry making and sold at Michael’s craft stores. You could use a similar product Sugru or even make your own using clear caulk and cornstarch. I then wrapped ends with electrical tape. I hot-glued inside of shells to lock everything in place as an added precaution. The connector on the end which exits the drag chain lightly touches right side upright during homing. I will keep an eye on it, but don’t think it will be a problem.


  2. I used o-rings between motor and motor mount where the screws go through to dampen vibration and allow a little airflow underneath motor. I ran motor for over an hour at 8000 rpm with and without o-rings to see what effect carriage had on conducting heat. I didn’t see any difference. This surprised me as I was expecting carriage to become warm. The o-rings didn’t necessarily quiet anything, but did change the frequency of sound in a pleasing way. It did however cut down on vibration. Motor does get pretty warm to touch after 30 minutes, but not alarmingly so. I didn’t test it very scientifically.

  3. I temporarily rigged top of enclosure and window to make clearance for taller motor. Since I was planning on dampening sound and vibration throughout machine I purchased a pack of inter-locking anti-fatigue mats from Harbor Freight. These have thin end caps to terminate after installation. I hot-glued two of these together to form an “L” shape which I am using as a stop for the bottom of enclosure door and make up the ~half inch height difference. I plan on adding magnets to door and stop after I finalize design.

I will test throughout the coming days for heat and performance and report back.
Jerry

You are correct, it’s not 70, it’s 78. The original is also rated for 52, not 50 watts. If it’s like any other parts I get from China, that’s well within the margin of error/marketing.

I’ve been running this way for a few weeks now, have done a few projects in this and that (wood, aluminum). You would think that 50% more power would make a difference. It really doesn’t. I don’t regret doing this (it was kind of fun) but if you’re looking for a big difference, this isn’t it.

Yeah. I’m still testing, but I haven’t noticed any performance gains. My hope is that there will be enough extra torque to power through some of the cuts that were previously stalling the spindle. I’m gonna try some mild steel this week with a Fusion360 file that was giving me problems. This will really point out the (hopefully big) differences. I’m also looking into the possibility of somehow fitting either a Nema 23 motor or maybe using two motors connected to a common pulley. I know this has been done in other applications, just not sure how.
Jerry

Was just thinking myself about how I would get two running in parallel…I think there’s room with a bracket on the carriage and a new/different pulley on the spindle…
I don’t think I’ll go further than just thinking about it though. If I really feel the need for more power, at this point I would just add a shapeoko3 to the stable.

Just out of curiosity, do you have the new version spindle motor driver (the one with 6 chips) and brass pulley?

Two IC’s, rev 4.3, The spindle and motor pulleys appear in the photos above.

Well I’ll be… Carbide 3D uses Imagineering for PCB’s! Me too.