Hammer HNC Table Top Router - VFD Intermediate Spindle Speed Errors <resolved>

[Nigelo]

Following Mike Richards Oscilloscope thread and a prompt from Gary Campbell on another thread for tuning, I got myself a reliable laser tacho and found huge Spindle Speed errors at lower speeds I didn't know existed. My system had otherwise been running with zero problems with several jobs up to 1 hour each.

Machine bought new direct from Felder UK without the standard Eding Controller.

Setup:

• Felder Group's Hammer HNC HF Router with air cooled 2.2kw Teknomotor Spindle https://www.felder-group.com/en-us/prod ... 82-p142995

• Leadshine MX4660 driver, Meanwell RSP-320-48 48v PSU and relays all under-slung the router deck

• Router's External connection interface presented as DB25 (I/O) and DB9 (E-Stop)

• Schneider Altivar 320 VFD (Model ATV320U22M2C) mounted in separate steel case with incoming line filter. The VFD is hard wired to Router with extended 2 metre lead

• Both Router and VFD case have their own mains power plugs with earth connection

• All above installed, pre-wired and configured by Felder prior delivery

• VFD parameters set by Felder to min 100Hz (6,000 rpm) max 400Hz (24,000 rpm) and control uses AI1 set for 0-10Vdc

  VFD Settings HNC HF V1 3 (2).xlsx

  (76.48 KiB) Downloaded 83 times

• Control cabinet therefore just houses the Acorn and its PSU, 8 x Relay board and 2 channel E-Stop plus DB25 BoBs and presenting an external DB25 and DB9 to match the Router's external conectors

  

• Control Cab to router connections: 25 core shielded parallel and 4 core shielded CY cables. Both shields grounded at Control Cab only

Observations:
1) OEM's VFD to spindle power cable shielded but NOT connected to GND at either end
2) OEM's Internal control cable to VFD not shielded (LAPP Olflex Classic 110 7 core x 0.75mm)
3) Felder has directly tied the Router chassis ground to the onboard 0Vdc COM at control cable connection to VFD COM
4) VFD control cable routed parallel to Spindle Power cable over full chassis length and extension to VFD enclosure with no separation.

Test results:

I've attached the VFD installation and Programming manuals. The usual report and the Electrical Layout as provided by Felder in German with my rough translation to follow.

Any ideas?

[Nigelo]

Attatch usual Report and Felder's Electrical Layout

[cncsnw]

If you graph "Observed VFD Hz" vs. "Observed Acorn VDC", is it approximately linear?

There are different ways to enforce a minimum speed (e.g. 100Hz / 6000RPM). One is a cutoff, where the speed command (analog voltage reference) is still scaled from zero, but any command below the minimum speed results in the minimum speed instead. Another is to change the bias and gain of the analog voltage reference, so that a zero-volt reference commands the minimum speed, and it goes up linearly from there.

The standard Acorn PLC program is written to assume the former. It looks like your drive may be set up as the latter.

[Nigelo]

If you graph "Observed VFD Hz" vs. "Observed Acorn VDC", is it approximately linear?

There are different ways to enforce a minimum speed (e.g. 100Hz / 6000RPM). One is a cutoff, where the speed command (analog voltage reference) is still scaled from zero, but any command below the minimum speed results in the minimum speed instead. Another is to change the bias and gain of the analog voltage reference, so that a zero-volt reference commands the minimum speed, and it goes up linearly from there.

The standard Acorn PLC program is written to assume the former. It looks like your drive may be set up as the latter.

Thanks, much appreciated.
Graph as follows:

As Installed

Analog 0-10v & COM Rerouted: Acorn direct to VFD (CY Cable: Shield to GND both ends)

This raises 2 questions:

1) The difference between the 2 graphs appears to show that shielded cabling has improved signalling. The 2nd graph is linear whereas the 1st is not although almost. Do you think the problem is with the VFD config rather than Ground Shielding?

2) The ATV320 is highly configurable. What type of VFD parameter should I look at for correcting this?

Nigel
edit to correct Graph titles: "Observed VFD Vdc" should read "Observed Acorn Vdc"

[cncsnw]

My guess would be:

LSP [Low Speed] Motor frequency at minimum reference, can be set between 0 and [High speed] (HSP).

If that is where the 100Hz minimum is set, you could change that to 0Hz.

Then, if you still want to prevent the spindle trying to run slower than 6000RPM, set a minimum of 6000 in CNC12 Control Configuration.

[Nigelo]

Many thanks Marc, resetting LSP to 0Hz cracked most of the issue as follows:

Original "as installed" but with LSP set to 0

In addition to LSP set to 0, I redirected Acorn's Analog Vout / COM direct to VFD with CY cable (fully grounded shield) to simulate replacing the VFD control cable with a fully shielded CY variant.

Final Comparison

RPM Comparison

RPM Comparison.JPG (39.12 KiB) Viewed 2347 times

Following the excellent results above, it's clear I need to leave the VFD's low speed LSP parameter at zero and replace the VFD control cable. Whilst I can live with the very small speed difference, I will try tuning that out by slightly increasing Acorn's Max Spindle RPM. The minimum speed will of course be controlled by setting 6,000 rpm in Acorn

Again Marc, very many thanks, really appreciated

Nigel

[Nigelo]

Having replaced the VFD control cable with a shielded version grounded at both ends; properly grounded the VFD to spindle power cable; and following Marc's advice left the LSP parameter set to 0, I carried further out RPM tests.

Acorn's analog output appears to be effectively an opamp, and therefore is presumably affected by downstream capacitance and/or resistance. As the preliminary results indicated proper grounding had actually increased the Commanded vs Observed RPM errors in the lower ranges, I tried installing a 0.1 uF capacitor (actually 97 nF or 0.097 uF) as recommended for stability reasons. Incredibly that completely resolved the problem in conjunction with leaving the LSP set at 0.

For completeness and to help other in the same situation, I attach the final RPM comparison

Final Spindle RPM Comparison

Could someone mark this thread as resolved - Many thanks

Nigel