Encoder for axis tacho feed back

More
14 Feb 2021 13:13 #198804 by tommylight
Velocity servo system with tacho should always move slowly when enabled and no voltage applied to input, not stand still.
The following user(s) said Thank You: johnmc1

Please Log in or Create an account to join the conversation.

More
15 Feb 2021 02:33 #198895 by johnmc1
Good day ,
Thanks for the replies,
At the moment it looks like the fanuc pulse coders only supply a negative
going pulse .
The linuxcnc machine config page encoder velocity, shows minus - (pulse count )when jogging minus.
when trying to jog positive the jog count is trying to give a negative pulse code.
I believe the fanuc pulse counter must be only counting negative pulse.

So I will try a non fanuc encoder, do not know how long that will take .:)
Thanks again for the replies
cheers john

Please Log in or Create an account to join the conversation.

More
15 Feb 2021 04:36 #198897 by johnmc1
I need a reality check
After lunch reality returned
With the linuxcnc axis cmd disconnected but linuxcnc enabled .
On the x axis servo amp amplifier gain it is possible to drive the motor cw or ccw,
this the showed the x axis pulse coder to have both + and - going pulses.
(other wise no machine position reference):(
cheers john

Please Log in or Create an account to join the conversation.

More
15 Feb 2021 11:56 #198921 by aleksamc
Your system supports only negative direction of encoder because you have made wrong wirering.
-Change wires of analog feedback of encoder
After this your motor will be operated in positive direction.
Or you may live as is, it doesn't matter, but make in your scaling some math operation tha will always give one polarity.
The following user(s) said Thank You: johnmc1

Please Log in or Create an account to join the conversation.

More
15 Feb 2021 13:31 #198931 by johnmc1
Good Day Aleksmac,
Thanks for your reply,
I converted to linuxcnc a few years ago my mazak m3 lathe which has encoder and tacho feed back.
and I had no trouble tuning the x z axis's

The problem I feel, is with the simulated tacho feed back .

setp hm2_5i24.0.7i83.0.4.analogout1-maxlim 278
setp hm2_5i24.0.7i83.0.4.analogout1-minlim -278
setp hm2_5i24.0.7i83.0.4.analogout1-maxscale -278 "I must always the minus sign -278 here"

I believe that the polarity should change with the sign of the encoder signal

Maybe when I connected it up wrong I have damaged the 7i83 card.
I have ordered a new set of mesa cards

Many thanks john

Please Log in or Create an account to join the conversation.

More
15 Feb 2021 20:35 #198975 by aleksamc
I'm not specialist in 7i83 so I can't tell you anything. Disconnect your wires and measure if analog output is present on pins.
I don't know what Servodrives you are using. But that servos I used, one input was connected to signal and other was connected to GND (it could bad news for you I think if it so). Look in schematic for your servos.
As for limits as you wrote, I think thay shoul be from 0 to 278 (or -278) as you counted. But really I didn't use such method of changing tacho generator, so I can't tell you for sure.
But tacho definetly has only one polarity.

Please Log in or Create an account to join the conversation.

More
29 Mar 2021 11:50 #204017 by johnmc1
Good Day , tacho feedback update tales of frustration

The tacho feed back works correctly in the positive direction,jogs smoothly to 8000mm/min.
The tacho feed back to velocity command ratio of 7volt cmd to 6volt tacho input correct as per manual.

######################Problem########################
When a negative cmd is applied, the tacho feedback is unstable varying between positve and negative output
and the motor reflects this by shuddering, this problems applies to all axis motors.
When the linuxcnc axis is homed the axis jumps negative for a instant then settles down.
Since early feb, I have done the following,
Replaced the fanuc encoder with a new encoder, even though it looked ok, on the oscilloscope
Replaced the fanuc servo drive , with a new semis servo drive.
Replaced the mesa cards, 7i83 the servo card 7i48.

All of the above has made no difference to the problem.
My last hope is if I replace the Mesa 5i24 pci card with a mesa 7i80hd_16 card .
Here lies my next problem.
When trying to mesaflash the 7i80hd-16 card with the following.
mesaflash --device 7i80 --addr 192.168.1.121 --write 7i80hd_16_svss_8.bit
,the terminal reports that 7i80 can not be found.
I have tried reading the manual for a week but no luck.

Any help or direction will be most appreciated.
Cheers john

Please Log in or Create an account to join the conversation.

More
29 Mar 2021 13:26 #204025 by tommylight
Points to bad brushes on the tacho.
The following user(s) said Thank You: johnmc1

Please Log in or Create an account to join the conversation.

More
29 Mar 2021 14:08 #204031 by PCW
I would disable the drives then verify that you
get correct +and - analog tach simulator outputs
when you rotate the shaft by hand. If you do, the
trouble is not related to the encoder feedback -->
analog simulated tach output path.
The following user(s) said Thank You: johnmc1

Please Log in or Create an account to join the conversation.

More
29 Mar 2021 14:09 #204033 by aleksamc
Possibly that motor doesn't have brushes, only simulated analog output.

I think he should make some test:
1) disconect motor from mechanic
2) run motor in both directions (possibly it will move only on maximum speed)
3) measure output voltage from tacho with multimeter or osciloscope.
4) compare signal from tacho and his simulated output from board.

But as for me, If I've lost so many time for such question, I would beter find working used tacho and connected it to the motor.
The following user(s) said Thank You: johnmc1

Please Log in or Create an account to join the conversation.

Time to create page: 0.192 seconds
Powered by Kunena Forum