Warner/Swasey QCM-6 lathe with Kflop/Kanalog

Fairly rare W/S QCM-6 cnc from a college. Low hours, but very early model GE2000 control. Issues with old boards. I have no copy of source code, so upgrades to control are not an available option. Looking at retrofit using Kflop/Kanalog. Machine has Inland/ Kollmorgan DC brush servos, spindle motor, and drives. Has +-10v analog drive inputs, tach connections on drives. Has 8 wire external encoder on spindle motor, connected to AXS board in control. Has 2, 6 wire connectors on RLV board in control. I assume they go to the x and z axis motors. I ran machine a few years ago. Been sitting a while. Will not power up now, board errors. Drives will move axis' and spindle with 1.5v battery. Does this sound like a good candidate for Kflop/Kanalog? What if the axis motors have resolvers? Machine has 8 station air operated tool turret, hydraulic pump, coolant pump, lots of IO.

Hi Charles,

It seems like KFLOP/Kanalog may be a good candidate. But you would need differential encoder feedback. Not sure what your AXS or RLV boards might be. 8 encoder wires wires would be expected for +5V, GND, A+ A- B+ B- Z+ Z-. If it has Resolvers instead of encoders I'd suggest changing to encoders.

You can add up to 192 additional optically isolated 24V IO by adding one or more Konnect Boards.

Thanks for replying Tom. I've read a lot of your responses to others. I'm sure everyone appreciates your assistance. The AXS and RLV boards are in the GE2000 control rack, and based on limited experience with GE2000 controls a long time ago, they are commonly known as axis and resolver boards. That's what made me suspect the x and z may have resolvers. The z axis motor model # is TTR-4202-3006-F-03-11. Drive # is TPA/2-15640-603-4202F2-4202I2. It is a combo drive, running x and z. Can't see the x motor model number without taking off covers, which I will soon do. Just trying to see what my options are. Is it common to change from resolver to encoder, or do you leave the resolver and add the encoder externally? I saw something in the Kflop manual about using the resolver output to create an encoder signal for KMotion. I'm guessing that won't do what I need. My Kollmorgan drives have tach inputs. Any advice on encoders that may work for me?

Hi Charles,

Thanks.

Quote:

---

Is it common to change from resolver to encoder, or do you leave the resolver and add the encoder externally?

---

If adding encoders the resolvers shouldn't be needed so whichever is easier.

Quote:

---

I saw something in the Kflop manual about using the resolver output to create an encoder signal for KMotion.

---

In some cases it is possible to make use of the resolvers. But it is complex, may involve a lot of experimentation and scoping, and the result may not be be good due to distortions and nonlinearities.

Quote:

---

My Kollmorgan drives have tach inputs.

---

You may be able to keep the tachs on the motors and keep them connected to the amplifiers in the original manner. But you may not need those if the amplifiers can be operated in current/torque mode instead of velocity mode. Sometimes just disconnecting the tach and reducing the input gain may result in this. The servo loop can basically derive velocity from the derivative of the encoder position. Basically D Gain. It just makes tuning a bit different or slightly more complex. You might read this.

HTH

Thanks Tom. I'll get the covers and maybe the motors off soon and see what I have. I've been tracing wiring, trying to develop a plan. I don't have the correct drawings for the machine, but I have a set for a larger Warner/Swasey with the same control, but different drives and motors. A lot of the wiring appears to be similar. I'll have a GE2000 with several spare boards available if anyone is interested.

Removed motor end plate. Resolver is gear driven, looks like 2/1, larger gear on motor shaft. Resolver is Harowe 11BRW-300-F-1/3. Need to replace resolver with a differential encoder. Any suggestions?

Hi Charles2,

USDigital is a good source of encoders. Select differential output type.

Thanks Tom. Would that be incremental or absolute?

Incremental

Thanks again. USDigital has a lot of options. Looking at the E6 hollow center, to eliminate the gear drives. What would you select as cycles per revolution?

Hi Charles,

Quote:

---

What would you select as cycles per revolution?

---

Choose the highest resolution that will not exceed KFLOP's 1 million quadraure counts/second limit. Note that 1 quadrature cycle = 4 quadrature counts.

So for example 4000 cycles/rev at 3000RPM would be:


4000 x 4 x 3000/60 = 800,000 quadrature counts/second and would be ok.

HTH

Thanks Tom, and thanks for the explanation. Will be placing an order for the Kflop and Kanalog. Still tracing wiring. My diagrams are for a larger W/S, and are helpful, but not exact.

Some progress. Traced and labeled wiring. Made lots of notes. Removed GE2000 control. Replaced axis motor resolvers with USDigital E6 differential encoders and cables. Mounted the Kflop/Kanalog boards and the 5V power supply. Connected the Z-axis encoder cable A+,A-,B+,B- to JP1, index + & - to JP2, ground to JP6, and +5V to JP8. Connected Z-axis servo drive output and ground to JP11. I have axis limit switches that are 24V NC and home switches that are 24V NO. I assume they go to the opto inputs at JP15, with the negative side of the opto inputs going to ground? Will the onboard series resistors handle the load? Is there a tutorial available that will guide me through this?

Hi Charles,

Quote:

---

Replaced axis motor resolvers with USDigital E6 differential encoders and cables.

---

Great.

Quote:

---

Connected the Z-axis encoder cable A+,A-,B+,B- to JP1, index + & - to JP2, ground to JP6, and +5V to JP8. Connected Z-axis servo drive output and ground to JP11

---

Good.

Quote:

---

I have axis limit switches that are 24V NC and home switches that are 24V NO. I assume they go to the opto inputs at JP15, with the negative side of the opto inputs going to ground?

---

Its best to have all the Limit and Home wiring isolated from the KFLOP/Kanalog GND. So an isolated 24V supply should be used for those. The - side of the opto should then go to the negative side of the +24V Supply not Kanalog GND.

Quote:

---

Will the onboard series resistors handle the load?

---

Yes.On Board resistors are 10K Ohms.

Quote:

---

Is there a tutorial available that will guide me through this?

---

You might follow the steps here then here.

Thanks for your reply, Tom. I'm sure I will need a lot more help as I go. I've been reading other conversion threads. Great info there. Reading everything I can find from Dynomotion. Just haven't found a step 1,2,3. The limit and home switches are powered by the machine 24V power supply. Would it be ok to ground the optos to the 24V system ground? That is, use the available 24V power?

Hi Charles,

Quote:

---

The limit and home switches are powered by the machine 24V power supply. Would it be ok to ground the optos to the 24V system ground? That is, use the available 24V power?

---

That should be ok if it is isolated from the KFLOP/Kanalog/PC GND.

There is a relay panel on a circuit board with 30 relays of which I'll only use about 10-15. It has its own 24V power supply. Relays are switched by +24V. I assume I should use the opto isolated outputs to trigger them, again isolating the 24V and ground from the Kflop/Kanalog?

Hi Charles,

Quote:

---

I assume I should use the opto isolated outputs to trigger them, again isolating the 24V and ground from the Kflop/Kanalog?

---

The 8 Kanalog opto outputs are only rated for 25ma each. So please determine the relay's current requirements. Also add a reverse diode across any coils. Kanalog has 8 relay driver outputs rated for 1A but they are not isolated from KFLOP/Kanalog.

Consider adding a Konnect to add 16 more opto isolated Outputs rated for 0.25A and 32 more opto inputs.

I was thinking the Konnect may be needed.

I tested the z axis encoder as was suggested on multiple threads. It appears to count up and down correctly, repeating position pretty close on + and - manual rotations of the the motor (one turn yields .200" on an indicator). When I tried to power up the drives, x and z axis both took off in the minus direction wide open! Only z axis is connected to Kanalog. When I replaced the resolvers with encoders, I cut the tach wires from the motor connectors and reconnected them color to color, eliminating the connector. After the run-away, I checked the motor connectors, and the tach wires had originally been cross-connected. I switched the tach wires at the drives. Still runs away. Even worse, (the end covers are still off the motors), it appears the brushes are throwing sparks! Big sparks! This has nothing to do with Kanalog. A few months back I tested the axis and spindle drives using the 1.5V battery method, and they worked perfectly. Since then I have removed the original GE2000 control and replaced the resolvers with encoders. The magnetics cabinet is unchanged. Before I started to test the drives with Kanalog, I removed the fuses from the spindle drive. After the axis motors ran away, I removed the fuses from them and replaced the spindle drive fuses. When I activated the drive contactor, the circuit breaker in the shop panel tripped! What has happened here? See third post for drive details.