Fanuc Arcmate to 3D printer- Servo Wiring
-
Hello! Im working on converting a 6-axis Fanuc arm to run on duet 3 as a 3D printer. I plan to use a pellet extruder. The servos are 3 phase 90v AC. Now I do have experience with closed loop steppers, but I'm looking for any advice or info on this process. Has anyone done this before? Can I avoid using large expensive drivers and amplifiers by using a 90 volt AC supply and just sending the data line from the Duet board or do I have no choice but to use the drivers designed for this system? If so whats the best driver option for these large servos?
-
@DoodleCube Servos cannot just run on direct power unless it is an integrated-driver style. Otherwise, they must be driven by a drive that is configured for--at the very least--the servo's phase number, voltage, amperage, and encoder.
No one is going to be able to tell you what the best driver option for the servos is without knowing what the servos are. And even then, the answer is likely to be "The servo drives that the company that makes those servos makes to match with those servos". Though, that's not necessarily the only option.
-
@Maestro Would swapping them be a viable option? I looked at drivers, and it will probably cost quite a bit unless I get lucky again. If these drivers are the only way, I can probably do it, just not sure which path would be easier/better
-
I have found 2 possible drivers. Both are 3-Phase, but one is the correct 90v and 2.5 amps but its DC and only for NEMA 23s, 24s, 34s, while the other is AC, for NEMA 42s and 52s but is 176v-253v. Is it possible to rectify the DC signal to AC or would it be better to use the AC driver and reduce the voltage? My servos are 3-Phase 90v AC servos rated for 2.5amps
-
@DoodleCube do the servos have built-in drivers? Many do. If they don't, can you re-use the ones that were in the Fanuc?
Duet WiFi hardware designer and firmware engineer
Please do not ask me for Duet support via PM or email, use the forum
http://www.escher3d.com, https://miscsolutions.wordpress.com -
@dc42 The servos do have encoders built in but not full drivers. I bought the arm of an industrial liquidation company and its just the arm without the controller.
-
@DoodleCube see if you can find the compatible drivers second hand. The other options are buys approproate drivers new (no doubt expensive) or replace the motors with ones that are cheaper to control (also no doubt expensive!)
-
The servos are 3-phase 90v 2.5 amp AC. Motor spec is A06b-0371-B155. I found some 3 phase drivers on stepperonline that might do the trick, if rectified or stepped down. It seems the original configuration included amplifiers, which I am not sure I can bypass. It may genuinely be easier to swap it all with similar size DC closed loop steppers if I can remove the servo's pinions easily enough. I will admit- my experience stops at closed loop nema 23s. I really would like to drive the original motors, but the Fanuc drivers well over 2k each. There has got to be another way to drive these, they are servos after all.
-
@DoodleCube I am not an expert but i expect you are opening up a world of pain trying to rectify/step down the servo driver outputs!
-
@T3P3Tony Its worth a shot, I like this kinda pain
-
@DoodleCube I'm with @T3P3Tony on this one; you are headin' for a hurtin' to use a mis-matched driver. The motors are AC; you can't rectify the driver output, you can only step it down. Trying to appropriately step-down the supply for an inductive & regenerative AC load in what needs to be a finely-tuned closed-loop system, for which you will almost certainly have access to only a limited number of driver parameters with limited adjustment ranges? Good luck. You will run into more than the voltage issue; servo drives are made for certain inertial expectations, torque expectations, vibration algorithms, etc., etc, and I expect you will encounter a point where you simply cannot tune the drivers to your needs.
If the drivers can talk to the encoders... Yes, it can probably be made to work. But by work, I mean move. Will it work well enough to use as a 3D printer? My money is on "no". 3D printing with servos already requires both good encoder accuracy and low settling times to print at speed without artifacts. With a Frankensteined system it's hard to imagine achieving settling times sufficient for printing at anything other than an absolute crawl.
My 2-cents; pick your own pains!
-
Update: Found 90v AC 3- Phase drives. Testing one soon!
