Re: [CAD_CAM_EDM_DRO] BDI with DRO/Encoder Board

Jon:
Thank-you for the reply. The X and Y axiis are working very nicely under
closed loop control. I was somewhat worried that using the Gekos the system
would have "feeback within feedback". But after experimenting I found the
system seems to work fine with X and Y. The following are the relevant
paramters:

P=5
I=0
D=0
FF0=0
FF1=1
FF2=0
Ouput_Scale 40000 0
Input_Scale 4000 0
Dead_Band = xxx

The encoders are 2000 counts/rev (quadrature). My Geckos are set to put out
10 pulses per every input pulse. With a shoptask this gives 0.00025" per
count that EMC can drive the motors with and 0.000025" / count of feedback
information from the encoders. The fact that the feedback sampling is so
much finer from the encoders probably helps overall loop stability. It is
very interesting to hear the motors under the condition of Dead_Band= 0.00.
You can hear (and see on the screen) the motors hunting for true position.
The hunting will stop as Dead_Band is increased and then the motors "sing"
as the normally do. That is the Geckos are "hunting" but the EMC ignores
this "hunting" and applies no correction.

Another interesting thing is that I had to change the polarity of the
Output_Scale for the Y axis. I will check this afternoon but I did not
think we had a wiring polarity error. Without this change EMC would come
back with the "I can't follow" error.

Now for the Z axis. When I try to move under a jog but I get a "I can't
follow" error no matter what I do. Setting Dead_Band to 0.00 should make
EMC "hunt". It does not under Z. My conclusion is that the feedback
information is not getting back in. (I can run motor under one of the other
channels (X or Y channel) and it works fine.). So it could be 1) The DRO
board Z input is not working. 2) Software.

From reading the mail I in this group I see that there is an experimental
BDI EMC source at www.linuxcnc.org. I will install this today and I will
attempt to pinpoint the problem.

By the way we were wondering has anybody got EMC to do CNC threading? Since
EMC can accept encoder information it seems that an encoder on the lathe
shaft and the last input on the DRO board would be a powerful combination.



David Kryskowski

----- Original Message -----
From: "Jon Elson" <elson@...>
To: <CAD_CAM_EDM_DRO@yahoogroups.com>
Sent: Monday, August 27, 2001 1:55 AM
Subject: Re: [CAD_CAM_EDM_DRO] BDI with DRO/Encoder Board


> David Kryskowski wrote:
>
> > To All:
> >
> > We have put together a Shoptask El Dorado with Camtronics Servos, Power
Supply and DRO board. The servos are being driven by Geko 320's. We have
installed the BDI and have been trying to get EMC functioning with smdromod
driving the motors. Our initial attempt failed due to erros in smdromod. I
have downloaded from sourceforge dro.c, dro.h and extsmdromod.c and compiled
them (along with some changes to handle our home and limit switch setup).
We can only get X and Y to work. The Z axis does not function properly.
It seems like the feedback information from the DRO board is not getting
through. When the Z axis is jogged we get and error saying that it can't
follow (I can't remember exactly). Any suggestions?
>
> The first thing is to make the encoders work right. You should be able to
> move the encoders (motors, leadscrew, etc.) manually and see the numbers
> change in the right direction and amount for the movement.
>
> I think you may have a problem with too many cooks in the kitchen!
> The gecko 320 takes encoder feedback, and receives step pulses
> from the CNC, and moves the motor to keep them in sync.
>
> The smdromod.o ALSO takes encoder counts and issues step pulses,
> and will issue additional step pulses when there is a discrepancy.
> Now, in the long term view, all these corrections should cancel out,
> but in the short term, there may be too much hunting around by two
> servo systems that were never designed to work together. maybe
> they are already working, it is not completely clear from your message.
>
> When the axes don't move the way the computer expects, you get a
> following error message. You can increase the permitted following
> error to see if you can visually the what the problem is. but, without
> the feedback working, you WILL get a following error, as EMC is not
> seeing the motion that apparently is ocurring.
>
> Also, with a servo system, all polarities (direction moved, movement
> sensed) must be consistant, or the servos will attempt to correct by
> going the wrong way, making the error greater, until a following error
> ocurrs. If the system jumps in the Z axis, you may have a direction
> polarity wrong.
>
> Jon
>
>
> Addresses:
> FAQ: http://www.ktmarketing.com/faq.html
> FILES: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/files/
>
> Post messages: CAD_CAM_EDM_DRO@yahoogroups.com
> Subscribe: CAD_CAM_EDM_DRO-subscribe@yahoogroups.com
> Unsubscribe: CAD_CAM_EDM_DRO-unsubscribe@yahoogroups.com
> List owner: CAD_CAM_EDM_DRO-owner@yahoogroups.com, wanliker@...
> Moderator: jmelson@... timg@... [Moderator]
> URL to this page: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO
> bill,
> List Manager
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>