missing steps?

I am not sure if this is the right sub-forum or not, but here goes...

After fiddling with StepConf, I was able to get my (book built) machine to move the appropriate distance when told to. Not sure if this is normal, but when I edited the ini file, the program told me there was a "jog error(?)" so I hafd to do it using StepConf.

Now I still have the problem I was originally trying to resolve. I thought the problem might be with Mach3, but I guess not. When I go up and down a set distance (say g0 z.5, g0 z0) a few cycles, the machine drops down with each cycle. This seems to happen on all three axis.

There is nothing loose. I can assure you have have triple checked. Nothing is slipping. I have swapped axis on the controller. I have swapped the stepper motors around.

I am attaching my ini file in case that's needed. I know that the X- and Y-axis do not match the Z-axis. I only corrected the scale on the Z-axis so I can work with one at a time. If any more info is needed, rather than retyping it all, I posted most of it in this topic: www.linuxcnc.org/index.php/english/forum...configuring-linuxcnc

Can someone please help me figure out what is going on?

have you done the latency test?
can you post your main hal file also?
I think you can focus on driver timing settings and scale setting to start with
Your accel doesnt look too high at 7, I would hope not atleast.

I read your other thread.
your scale should be 16000. you have 1.8* step motors, 1/8 microstepping, and 10tpi screws, right?
I dont know what mach3 step pulse and dir pulse relate to, but I see you have 5 as the recommendation.

The electronics I use came with a mach3 xml It also had 5 as step pulse and dir pulse.
I don't know if that actually translates to anything, but incase you want to try, my settings are:

When I go up and down a set distance (say g0 z.5, g0 z0) a few cycles, the machine drops down with each cycle. This seems to happen on all three axis.

If you set your feed rate to a low speed and make the same moves with g1 instead of g0 do you still have the problem?
Same with go if you dramatically reduce your max velocity and acceleration?

Rick G

I'll be honest with you guys. I don't know a whole lot about CNC or gcode. What I do know is only what I have learned while building this machine and from reading the posts on several forums. I didn't even know what g1 was. I just looked it and and it seems to be a slower feed rate than g0? I will try that this afternoon and report back with the results.

I will also try the latency test and attach the HAL file.

When I used the "correct" scale value as per the "1.8* step motors, 1/8 microstepping, and 10tpi screws" logic, when issuing a g0 z5 command, for instance, the machine would only move about 0.5" +- 0.02". Only by changing the scale value to what it currently is set to under Axis 2 (the Z-axis) was I able to get it to move (darn close to) the right distance.

I'll be honest with you guys. I don't know a whole lot about CNC or gcode

No problem.

Might want to look here...
linuxcnc.org/docs/devel/html/gcode/gcode...ck_reference_table_a

G90 - absolute distance mode
G0 say to make a move at the fastest possible speed the machine can.
G1 say to move at a set speed.
F sets the travel speed.

So if you are using inches try..

G90
F5
g1 z0
g1 z4
g1 z0

Z should go to 0
Z should go to 4 at a speed of 5
The last statement should bring Z back to 0

Rick G

OK... so I tried g90 f5 g1 z1 g1 z0 g1 z1 g1 z0 g1 z1 g1 z0 and each time it got progressively lower on both cycles.

Here is the HAL file that was requested.

Now perhaps try to dramatically reduce your acceleration and try the same test again.
Also double check your and as pointed out double check your step and direction timing settings.

Rick G

Some things I didnt see in your earlier post, how do you have your motors hooked up to the driver?
It will either be bipolar series, bipolar parallel.
What is the voltage of the power supply you use?
The steppers you have are fairly high inductance 6.8 or 27.2mH depending on wiring , which means a high voltage power supply is preferred.
49volts or 166volts depending on inductance. (estimations from gecko calculation)
Both are higher than the 36v max input of the tb6560.
Also have you set the current dip switches to match the rated motor current?

Verify your dip switches for motor current (2.8amp parallel) , and microstepping.
If you aren't wired bipolar parallel, I'd recommend that also.

In the pdf for your driver board, I didnt see any mention of driver timing.

The motors are wired in Bipolar Parallel. The power supply is 24v 14.6 amp. The DIP switches are set as follows:

Current = 100% (out of 100%, 75%, 50%, 25%)
Decay Mode = FAST
MicroStep = 1/8

Not sure what the Decay Mode is or where that should be set.

If I had to guess, decay mode would be idle current reduction.

I would recommend setting the scale to 16000 and playing with the driver timing settings to get the movement right.
You can try those values I posted earlier, by editing the hal file.
I'm, not exactly sure how the numbers in the hal relate to the numbers in stepconf. My brain doesnt work in nanoseconds.
Does your motors sound smooth and fairly quiet when jogging?
If they are jumpy are make strange noises, that is an indication the driver timing is off.

You may try to search on google for linuxcnc tb6560 timing to see if you find anything.

these are number I found on a quick search, maybe you should try these first.