[rainman6272]

I managed to get a Gerber Samplemaker with a 5'x10' table from a government auction. The machine was originally at a technical college in San Jose, CA, in their fashion design program. This is a tangential knife type cutter, but I want to remove the original tool holder and install a 12" z-axis to run a plasma torch. If I can, I want to be able to equip the z-axis with a spindle/router and get a dual use table if you guys think the servos can make the torque.

My plan is to use the power supply, gantry, servos, and the tracks (with their gear racks and linear bearings). When I got this thing, I only knew that I wanted DC servos with quadrature encoders for an easy MACH 3 conversion. The gantry has two Parvex RS320H (65V max, 13A max) moving it along the track (X-axis) and another RS320H moving the tool head along the gantry (Y-axis). The cutter heads were steered by a Parvex RS130E (33V max, 9A max) that I plan to use to run the Z-axis. All four are equipped with Renco RM15 differential encoders (5V, 125mA).

My plan to use the original servo drivers was scrapped when Gerber's technical support couldn't even acknowledge that they had manufactured and sold it. No one in a support role was working there when the machine was made in 1996 and they didn't have any technical documentation to provide either. So much for bypassing the original control system and feeding x/y data directly to the drivers.

That leaves me with the option of using either Gecko G320X (which only supply 50mA for the encoders when they need 125mA) or CNCDrive's DG4S-08020. Right now I'm leaning towards the CNCDrive drivers since they can handle the current the encoders require and accept the differential output signal of the RM15 encoders. I understand the Gecko servo drivers accept a differential signal as well, you just don't use the negative wires or the index pair, but I don't know if there's a disadvantage to this. It seems that differential encoder signals are less susceptible to noise and better for longer runs.

Right now, it'd be great if I could get some advice on which servo amplifiers would be the best for what I'm trying to achieve. The CNCDrive's seem to be better matched to what I need, but the Gecko drivers seem to be more popular and easier to use. I also need to know if you guys think I'll need braking circuits for the X and Y servos.

Here's the stats for the RS320H servos that run the X and Y axis:

Low speed torque 0.54 N.m Mo
Permanent current at low speed 4.5 A Io
Supply voltage with loaded motor 49 V U
Definition speed 3000 rpm N
Maximum supply voltage 65 V Umax
Maximum speed 4800 rpm Nmax
Peak current 13 A Imax
Back emf constant at 1000 rpm (25°C)* 13.3 V Ke
Torque constant 0.127 N.m/A Kt
Static friction torque 2.4 N.cm Tf
Viscous damping for 1000 rpm 0.53 N.cm Kd
Winding resistance(25°C) 1.52 Ω Rb
Winding inductance 2.2 mH L
Rotor inertia 0.000083 kg.m² J
Thermal time constant 7 min Tth
Motor mass 1.34 kg M

Pictures of the gantry as I got it:
Attachment 258672Attachment 258674
After I took off some superfluous parts:
Attachment 258676
The track extrusions with the gear rack and linear bearing rail
Attachment 258678

The power suppy (on the left) and the control computer/driver housing (on the right)
Attachment 258680
The inside of the power supply
Attachment 258682
The RM-15 encoder
Attachment 258684Attachment 258686

Pictures of the inside of the control box in case someone sees a way for me to reuse the original drivers
Attachment 258692Attachment 258694Attachment 258696


Pictures of the amplifier board assembly. You can see where the power supply wires in the molex connectors from the power supply and go to the motors. In the upper left of the main board, there's a 14-pin header that handles the encoder signals and direction signals. Obviously that isn't enough to give a conductor to each differential encoder for each driver or for control signals to each driver, so I don't know how that's handled.
Attachment 258700Attachment 258706Attachment 258704


And here's some pictures of one of the individual driver boards close up.
Attachment 258702Attachment 258698Attachment 258708Attachment 258710Attachment 258712Attachment 258714Attachment 258716Attachment 258718Attachment 258720Attachment 258722

Lastly, here's some pictures of the silly looking control panel that I, regrettably, can't use.
Attachment 258688Attachment 258690

So in summary: Is the CNCdrive or the Geckodriver better given the encoders I already have? Do I need braking circuits? Can I reuse the old servo amplifier board and connect a breakout board to it? Am I just totally in over my head? Thanks in advance guys. I'm really excited about tackling this project.

[rainman6272]

Aaaaand here it is up on cinder blocks running a diagnostic test just so I can make sure the whole thing works. The vacuum table it was on was permanently attached to the ground, so I had to rip the track off, which was epoxied to the table.