MPG jog speed different when in jogging in positive

16 Mar 2022 13:17 - 16 Mar 2022 13:50 #237467 by Wischi
Hello guys,

I just finished my control panel and as I was setting up the mpg for jogging I noticed that the speed at which the axis moves is different depending on the direction. For example: If I move the Y axis with a 0.1mm scale in the negative direction and turn the mpg encoder as fast as I can I get a velocity of about 400-500 mm/min. If I now turn in the positive direction I get 1400-1600 mm/min. The Z axis is the same. On the X axis it is reversed, so negative = fast, positive = slow. I am using a 7i84D and one of these 100ppr ebay mpg encoders(which has been fitted to an original Deckel mpg). So only A and B phase (also no shielded cable).
I checked the raw counts of the encoder. Sometimes it missed one or two counts, but is that causing the speed difference? The difference in speed is extremly repeatable, even after reboot.

Here is my HAL for the mpg (the top half is only for selecting the axis and scale):

loadrt mux4 count=1
addf mux4.0 servo-thread

setp mux4.0.in0 0
setp mux4.0.in1 0.00001
setp mux4.0.in2 0.0001
setp mux4.0.in3 0

loadrt handrad
addf handrad.0 servo-thread

loadrt ilowpass
addf ilowpass.0 servo-thread

setp ilowpass.0.scale 1000
setp ilowpass.0.gain 0.01

net mpgx_in handrad.0.inx <= hm2_7i92.0.7i84.0.2.input-04
net mpgy_in handrad.0.iny <= hm2_7i92.0.7i84.0.2.input-05
net mpgz_in handrad.0.inz <= hm2_7i92.0.7i84.0.2.input-06
net mpga_in handrad.0.ina <= hm2_7i92.0.7i84.0.2.input-07
net mpg10_in handrad.0.in10 <= hm2_7i92.0.7i84.0.2.input-09
net mpg100_in handrad.0.in100 <= hm2_7i92.0.7i84.0.2.input-08

net mpgx_led handrad.0.outx => hm2_7i92.0.7i84.0.2.output-01 axis.x.jog-enable
net mpgy_led handrad.0.outy => hm2_7i92.0.7i84.0.2.output-02 axis.y.jog-enable
net mpgz_led handrad.0.outz => hm2_7i92.0.7i84.0.2.output-03 axis.z.jog-enable
net mpga_led handrad.0.outa => hm2_7i92.0.7i84.0.2.output-04 axis.a.jog-enable
net mpg10_led handrad.0.out10 => hm2_7i92.0.7i84.0.2.output-06 mux4.0.sel1
net mpg100_led handrad.0.out100 => hm2_7i92.0.7i84.0.2.output-05 mux4.0.sel0

net mpg-scale <= mux4.0.out
net mpg-scale => axis.x.jog-scale
net mpg-scale => axis.y.jog-scale
net mpg-scale => axis.z.jog-scale
net mpg-scale => axis.a.jog-scale

#did not see any change when 1 or 0
#setp axis.x.jog-vel-mode 1
#setp axis.y.jog-vel-mode 1
#setp axis.z.jog-vel-mode 1

net jog-counts-raw <= hm2_7i92.0.7i84.0.2.enc0.count

net jog-counts ilowpass.0.out => axis.x.jog-counts
net jog-counts ilowpass.0.out => axis.y.jog-counts
net jog-counts ilowpass.0.out => axis.z.jog-counts
net jog-counts ilowpass.0.out => axis.a.jog-counts

Last edit: 16 Mar 2022 13:50 by Wischi.

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

16 Mar 2022 15:36 #237472 by PCW
Losing counts is a sign that something is wrong
I would:

1. Make sure the MPG GND is 7I78D field power GND
2. Make sure that you have close to a 5V swing on the MPG encoder
A, B signals (some types may need pullup resistors to 5V to get a full swing)

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

16 Mar 2022 17:06 #237478 by Wischi
1. The encoder gets 5V from a different power supply than the 7i84. The 7i84 VFIELDA and B both get 24V. Do I have to power VFIELDB with the same 5V power supply as the encoder? Otherwise the encoder GND is not connected to field power GND

2. I measured 4.97V at the encoder.

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

16 Mar 2022 18:37 #237483 by PCW
The two GNDS must be common so your 5V MPG supply
GND must connect to field power GND

Note: you cannot run VFIELDB on 5V on a 7I84D
as VFIELDB is connected to VIN on current cards
The following user(s) said Thank You: Wischi

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

17 Mar 2022 06:53 #237520 by Wischi
I just tested it with the GNDs of the two power supplies connected. That didn‘t work so well. The axis wouldn‘t enable after turning on the machine. But nothing got destroyed.

I read in a different thread that it is possible to use a step down converter on the 24V line. Then the 5V and 24V would share a GND. And it worked! No more missed counts and same speed in both directions!

Learned something new, thank you very much!
The following user(s) said Thank You: tommylight

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

Time to create page: 0.097 seconds
Powered by Kunena Forum