RRF 3.5beta2/closed-loop encoder config-problems
-
Hi,
I have problems getting my encoder to work.
The relevant config.g line looks like this:
M569.1 P50.0 T2 C400.0 S200.0 R100.0 I0.0 D0.0
M569 P50.0 D4 S0This doesn't seemed to be supported by the firmware, because I get this error:
"Error in start-up file macro line 26: Encoder counts/rev must be at least four times steps/rev"
The encoder is an RLS LM10IA Linear encoder that outputs an quadrature signal. The encoder and the steps/revolution is fixed. Is there someway I can ignore the error meassage?
The purpose of the encoder is to counteract a bad/inconsistent Linear Motion System.
I know that linear encoders should be combined with a magnetic shaft encoder but why shouldn't it work standalone?
Thanks!
-
@CMP As you say, the documentation says that linear encoders should be combined with a magnetic shaft encoder on the motor. The last sentence explains the reasoning behind this.
A linear composite encoder comprises a linear quadrature encoder that tracks position on a linear axis (for example the Renishaw LM10IA or LM10IB) and a Duet3D magnetic shaft encoder. The shaft encoder handles motor commutation, the linear encoder is used to close the loop.
I have asked @dc42 to comment on how it is expected to be set up.
Ian
TronXY X5S with Duet 3 MB6HC and Roto toolboard : Cartesian bed-slinger with Duet 3 Mini 5+ WiFi and 1LC : RRP Fisher Delta v1 with Duet 2 Maestro : Polargraph with Duet 2 WiFi
-
@CMP to use a linear axis encoder you also need a magnetic encoder on the stepper motor. The magnetic encoder is connected to the EXP1HCL via the ribbon cable connector and he linear quadrature cable is connected to the Molex KK connector. Our magnetic encoders will be in production shortly.
Note, although we demonstrated this configuration at Formnext last November, it is still experimental. We have not tested it in a real printer or CNC machine yet.
-
@dc42 thanks for the quick reply!
I will test it as soon as it becomes available.
Is there a reason the ratio 'Encoder counts/rev' vs 'steps/rev' has to be that high? With my current encoder I'll only get a ratio of 2:1.
-
@CMP said in RRF 3.5beta2/closed-loop encoder config-problems:
Is there a reason the ratio 'Encoder counts/rev' vs 'steps/rev' has to be that high? With my current encoder I'll only get a ratio of 2:1.
For the shaft encoders we need a resolution high enough to perform commutation. When using both a linear axis encoder and a shaft encoder, the resolution of the linear encoder should not be as important.
Btw I've updated the EXP1HCL wiki page and the M569.1 entry on the GCodes wiki page.
-
@droftarts Is there scope for doing linear encoding (with the Renishaw LM10, for example) without also having the magnetic shaft encoder? I'm very space constrained so adding a magnetic encoder is very challenging, and feels superfluous as I would have thought we can do closed loop with just the linear encoder.
-
@jjem said in RRF 3.5beta2/closed-loop encoder config-problems:
Is there scope for doing linear encoding (with the Renishaw LM10, for example) without also having the magnetic shaft encoder?
No. We need the magnetic encoder so that we can set the coil currents the desired angle ahead or behind the current step phase position to reasonable accuracy. In effect we are treating the stepper motor like a BLDC motor with a very high number of poles. BLDC motors that can operate over a wide range of speeds in either direction have built-in sensors for commutation.
OTOH it may in future be possible to run the stepper motor in open loop mode and use the linear encoder to monitor deviation between commanded and actual position.
Our magnetic encoder board takes up about 15mm of depth on the back of the stepper motor, including the connector.
