Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....Still there....
Posted by
Jon Elson
on 2002-09-14 22:14:38 UTC
Les Watts wrote:
quirk, I surmised it likely to be INSIDE the LS7166 chip. The fact that the
STG-II still has the same bug makes me think even more strongly that it
is a defect in the way the 7166 works, or in the way the signals are timed in
the STG cards to read out the successive bytes of the counters.
It may also be a complete failure of the encoder latch to actually latch the
contents of the 24-bit position value when the software gives the command!
noise on the encoder A and B signals. The counters are REPORTING incorrect
numbers to the CPU! There's no way the counter could jump plus 65536 counts
and then back 65536 counts if it was just noise on the encoder lines.
It has to be that external noise is being brought INSIDE the card, by passing
shield currents through the traces on the card. They may also be passing ground
loop currents through the board, if the encoder ground connection is tied to the
motor frame, for instance. Also, the return signal from the servo amp velocity
command probably is grounded at the servo amp terminals, another source of
noise being injected back into the STG card's traces. That's why I used
differential inputs on my servo amp, as I assumed that I'd have ground loop
potentials due to the DC power supply current supplying the motor.
But, the jumping bits were a problem on my system, even with the servo amp off.
My servo amp didn't make it any worse, but it was bad enough as is to require a
software fix.
Jon
> The count rolling over at 65k (16 bit) is perplexing considering the stgBut, this is a classic problem in the STG-I card. Due to the nature of the
> has 24 bit counters.
quirk, I surmised it likely to be INSIDE the LS7166 chip. The fact that the
STG-II still has the same bug makes me think even more strongly that it
is a defect in the way the 7166 works, or in the way the signals are timed in
the STG cards to read out the successive bytes of the counters.
It may also be a complete failure of the encoder latch to actually latch the
contents of the 24-bit position value when the software gives the command!
>I don't think a filter will help. The counters are not getting erroneous counts from
> I am beginning to think filters... or a malfunctioning stg card.
> Possibly filtering the encoders (Babar's bit rate seems to be pretty low)
> Possibly easier than filtering amp outputs. RC = a microsecond or so
> at the stg inputs?
> What do you think?
noise on the encoder A and B signals. The counters are REPORTING incorrect
numbers to the CPU! There's no way the counter could jump plus 65536 counts
and then back 65536 counts if it was just noise on the encoder lines.
It has to be that external noise is being brought INSIDE the card, by passing
shield currents through the traces on the card. They may also be passing ground
loop currents through the board, if the encoder ground connection is tied to the
motor frame, for instance. Also, the return signal from the servo amp velocity
command probably is grounded at the servo amp terminals, another source of
noise being injected back into the STG card's traces. That's why I used
differential inputs on my servo amp, as I assumed that I'd have ground loop
potentials due to the DC power supply current supplying the motor.
But, the jumping bits were a problem on my system, even with the servo amp off.
My servo amp didn't make it any worse, but it was bad enough as is to require a
software fix.
Jon
Discussion Thread
lti_cnc
2002-08-28 06:50:42 UTC
Position counter stop updating....
Jon Elson
2002-08-28 10:09:47 UTC
Re: [CAD_CAM_EDM_DRO] Position counter stop updating....
riet91
2002-08-29 21:30:22 UTC
Re: Position counter stop updating....
Jon Elson
2002-08-30 10:54:54 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....
Babar
2002-09-02 04:28:15 UTC
Re: Position counter stop updating....
Jon Elson
2002-09-02 15:11:56 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....
Babar
2002-09-04 08:39:00 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....
Country Bubba
2002-09-04 13:32:20 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....
Babar
2002-09-13 04:21:42 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....Still there....
Les Watts
2002-09-13 07:17:04 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....Still there....
Jon Elson
2002-09-13 16:40:38 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....Still there....
Babar
2002-09-14 05:05:59 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....Still there....
Les Watts
2002-09-14 16:17:30 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....Still there....
Jon Elson
2002-09-14 21:47:47 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....Still there....
Jon Elson
2002-09-14 22:14:38 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....Still there....
Jon Elson
2002-09-17 10:18:34 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....Still there....
Babar
2002-09-19 02:24:52 UTC
Re: [CAD_CAM_EDM_DRO] Re: Position counter stop updating....Still there....