XX-Axis Gantry negative direction drift

Hi, I have a double X - Axis 1200 x 2400mm gantry router running on G540 and 387 oz motors (3.5) A

The configuration is XXYZ the second X being the A drive slaved to X.

I recently upgraded the gearing on the machine from 1.75mm per rev to 42mm per rev on the X and Y drives and 10.5mm per rev on the Z drive.

I then experienced negative drifting see ( Y axis negative direction drift. Help! 05 Dec 2012 03:00 ) loosing steps ( gaining steps? ) on the second X that is the A drive.

As a result of reading the afore mentioned excellent thread I did a number of tests all in air cuts.

Running 8000 lines of code in air cuts after 15 minutes I got about 2.5mm neg drift on the left hand X(A drive) axis running any longer starts an exponential drift (heat related?)

Swapping drives so that A drives the right hand side X motor.
Running 8000 lines of code in air cuts after 15 minutes I got about 2.5mm neg drift on the right hand X(A drive) axis (not the motor)

I then ran about 17000 lines of code (as suggested in "Y axis negative direction drift. Help!") with Y drive disconnected , negative drift became marginal after about half an hour

I then ran with about 34000 lines of code with Y and Z drive disconnected for an hour and no drift at all.

I then did some stationary tests (by feel turning the motor by hand) :-

with the machine on and stationary, X, Y and Z hold well but A has about half the holding torque.

with the machine on and stationary and Z disconnected, X and Y hold well and A has improved holding torque.

with the machine on and stationary and Z and Y disconnected, X and A have equally good holding torque.

I tried another test I swapped the A driver board with the Z driver board position on the G540 and find a lower torque on the Z drive motor.

My local Gecko agent assures me that the 48V 7.3A power supply is specked to drive the Gecko G540 with four 387 Oz/in motors. the conclusion being that either the A driver board is faulty or the power supply is faulty and only presenting on the A driver.

I will keep you posted on any developments and any comments will be greatly appreciated. I now need to go figure how to test my power supply.

Cheers

Sounds like you have done a fair amount of testing.
If possible monitor the voltage and amps from the power supply under power. Also to the stepper motors.
Are you using a heat sink/fan on the drives?
Checked the amp settings of all drives, (amp setting resistors)
For testing adjust acceleration and velocity way down to see if that helps.

I recently upgraded the gearing on the machine from 1.75mm per rev to 42mm per rev

That is a seems to be a very large change and a much greater demand on the motors.

Rick G

captain wrote:

Hi, I have a double X - Axis 1200 x 2400mm gantry router running on G540 and 387 oz motors (3.5) A

The configuration is XXYZ the second X being the A drive slaved to X.

I recently upgraded the gearing on the machine from 1.75mm per rev to 42mm per rev on the X and Y drives and 10.5mm per rev on the Z drive.

I then experienced negative drifting see ( Y axis negative direction drift. Help! 05 Dec 2012 03:00 ) loosing steps ( gaining steps? ) on the second X that is the A drive.

As a result of reading the afore mentioned excellent thread I did a number of tests all in air cuts.

Running 8000 lines of code in air cuts after 15 minutes I got about 2.5mm neg drift on the left hand X(A drive) axis running any longer starts an exponential drift (heat related?)

Swapping drives so that A drives the right hand side X motor.
Running 8000 lines of code in air cuts after 15 minutes I got about 2.5mm neg drift on the right hand X(A drive) axis (not the motor)

I then ran about 17000 lines of code (as suggested in "Y axis negative direction drift. Help!") with Y drive disconnected , negative drift became marginal after about half an hour

I then ran with about 34000 lines of code with Y and Z drive disconnected for an hour and no drift at all.

I then did some stationary tests (by feel turning the motor by hand) :-

with the machine on and stationary, X, Y and Z hold well but A has about half the holding torque.

with the machine on and stationary and Z disconnected, X and Y hold well and A has improved holding torque.

with the machine on and stationary and Z and Y disconnected, X and A have equally good holding torque.

I tried another test I swapped the A driver board with the Z driver board position on the G540 and find a lower torque on the Z drive motor.

My local Gecko agent assures me that the 48V 7.3A power supply is specked to drive the Gecko G540 with four 387 Oz/in motors. the conclusion being that either the A driver board is faulty or the power supply is faulty and only presenting on the A driver.

I will keep you posted on any developments and any comments will be greatly appreciated. I now need to go figure how to test my power supply.

Cheers

no way a 7.5a power supply can hold 4 drives stationary .. with 3.5A per phase winding , you have just about proved that by removing one stepper and the the other improves it's hold
i'd suggest 2 power supplys and split the load appropratey , also check that you have the correct current limiting resistors installed ..

check your psu voltage whilst under load with all 4 motors powered , if the voltage has dropped , then the assumption can be made that the power supply is under powered
i doubt very much that you have faulty board or psu , also make sure your power supply cables are heavey especialy if they have to run some distance as voltage can drop , and would give the same symptoms
either from your power supply to the drivers and to the motors ..

it's best to wire from the psu to the drivers as individual cables rather than run just one cable and link across drivers .. i.e star style ..

steppers cant gain steps they can only loose them

also try reducing your accelleration by say 10% ,

Electronics isn't my specialty, but I had the same configuration for a couple of years with no noticable problems. According to the gecko website:

The easiest factor in choosing a power supply is its current rating, which is based on your motor ratings. A motor control will always draw less than 2/3 of the motor’s rated current when it is parallel (or half-winding) connected and 1/3 of the motor’s rated current when it is series (or full-winding) connected. That is to say, a 6 amp per phase motor will require a 4 amp power supply when wired in parallel and a 2 amp power supply when wired in series. If multiple motors and drives are used, add the current requirements of each to arrive at the total power supply current rating.

What this means to me is a 3.5amp per phase bipolar motor will draw 1/3 of 7 amps. 2.333 amps * 4 motors = 9.3333 amps.
Now when I bought my setup, the retailer supplied the 7.3 amp supply with either 3 or 4 motor config.
I see now that they recommend the 7.3 amp with 3 motors, and a 12.5 amp supply with 4 motors.

I said in the other thread that your supply rating was ok, because I had no problems with the same setup, but based on the gecko suggestions, It wouldnt hurt to try a larger power supply.

I think the best bet is room temp air blowing on the back of the g540 though. Of all the random faults and lost steps I have read about, this is the most likely fix.

alan_3301 wrote:

Electronics isn't my specialty, but I had the same configuration for a couple of years with no noticable problems. According to the gecko website:

The easiest factor in choosing a power supply is its current rating, which is based on your motor ratings. A motor control will always draw less than 2/3 of the motor’s rated current when it is parallel (or half-winding) connected and 1/3 of the motor’s rated current when it is series (or full-winding) connected. That is to say, a 6 amp per phase motor will require a 4 amp power supply when wired in parallel and a 2 amp power supply when wired in series. If multiple motors and drives are used, add the current requirements of each to arrive at the total power supply current rating.

What this means to me is a 3.5amp per phase bipolar motor will draw 1/3 of 7 amps. 2.333 amps * 4 motors = 9.3333 amps.
Now when I bought my setup, the retailer supplied the 7.3 amp supply with either 3 or 4 motor config.
I see now that they recommend the 7.3 amp with 3 motors, and a 12.5 amp supply with 4 motors.

I said in the other thread that your supply rating was ok, because I had no problems with the same setup, but based on the gecko suggestions, It wouldnt hurt to try a larger power supply.

I think the best bet is room temp air blowing on the back of the g540 though. Of all the random faults and lost steps I have read about, this is the most likely fix.

also dont forget to factor in cable loss etc , and this is all theory , ...
and that modern switchmode power supplys are not the ideal for this situation, you just cant beat a simple transformer and bridge diode and capacitor and of course experience ! ..

Hi Guys thanks for all the input.

I have been running my gantry for well over a year with a number of problems see U Axis Configuration under Advanced Config, which I recently resolved by repairing the Power Supply, I still have to configure "Gantry HS" though. So when you say "power supply power supply!!" I certainly hear you, but with all the tests I have done and the latest:-

With the A driver board in the Z driver board position on the G540 running for 20 minutes:-

Power supply at start voltage = 48.3 dropped to 48.2
after 8 minutes returned to = 48.3 for rest of test

After 20 minutes I lost 22mm in negative drift.

This all points to that particular board being faulty as it shows up in as common to all the tests, but is this the cause or is it the result of a faulty or under rated power supply in the first instance?

Rick G wrote:

That is a seems to be a very large change and a much greater demand on the motors.

One of my previous problems was loosing steps, getting following errors and RTAPI errors from running too fast on a small scale, that is now fully resolved with the new gearing system,no more following errors and RTAPI errors.

alan_3301 wrote:

Electronics isn't my specialty, but I had the same configuration for a couple of years with no noticeable problems. According to the gecko website:

[cnc basher wrote:[/b]

also dont forget to factor in cable loss etc , and this is all theory , ...
and that modern switch-mode power supply's are not the ideal for this situation, you just cant beat a simple transformer and bridge diode and capacitor and of course experience ! ..

Neither is electronics mine, but I have had a lot of experience in DC wiring and am well aware of voltage drop and power loss on under sized cabling, But what really gets me is this digital and switch mode stuff. I would agree a chunky transformer and bridge diode is the way to go but can the fancy electronics handle that kind of supply. I would be more inclined to go with a 48V battery and a decent battery charger to get a smoother supply ( now we are talking big money) and I am way out of my league here.
Unfortunately with the G540 it is almost impossible to use two power supplies as the drivers plug into the break out board via a multi-pin header and the power is supplied via the break out board.

My thoughts are now a 12.5 amp power supply a new driver board and a cooling fan (though I have never had a heating problem, a cooling fan is really the cheapest upgrade and certainly can't do any harm)

Cheers

Edit

Sorry cncbasher

steppers cant gain steps they can only loose them

I realized that, what I was trying to say was that the axis appeared to be gaining steps, "failing to hold" would have been a better description.

also try reducing your accelleration by say 10% ,

Thanks,I had already reduced as you suggested in the "Y Neg Drift Direction" thread, max velocity from 200mm/s to 100mm/s and max acceleration from 1000mm/s/s to 300mm/s/s with no improvement.

cheers

Hi BigJohnT

Not quite sure what you mean is that the setp stepgen dirhold and ~dirsetup value in Hal which is 15200

It is connected to the parallel port

It's not the stepper test but a program which runs up and down the X axis with the same X coordinates that is from 0 to 260mm offset on the Y axis by .3mm at each pass and varying Z depths between 0 and 18mm, a program I wrote in CamBam using

3D Profile Machining Operation horizontal scan-line

.

alan_3301 wrote:

I see now that they recommend the 7.3 amp with 3 motors, and a 12.5 amp supply with 4 motors.

Could you direct me to that site as my supplier in Australia is still advertising the kit thus:-

1 x G540 Geckodrive Stepper Controller
1 x PS-01 7.3A 48V Linear Power supply
1 x SW-01 Emergency Stop Switch
1 x SW-06 IEC Power socket/fuse/ switch
1 x CB-10 IEC Female Socket
2 x RL-04 48V coil 20A SPST Relay
4 x SM-03 387 oz.in Bi-Polar Stepper motors

cheers

www.cncrouterparts.com/nema-23-cnc-electronics-c-22.html

One of my previous problems was loosing steps, getting following errors and RTAPI errors from running too fast on a small scale, that is now fully resolved with the new gearing system,no more following errors and RTAPI errors.

O.K. but keep in mind the new gear ratio requires more torque from your motors to move and hold the gantry.


Rick G