[David Lucas]

Hi everyone

As the title explains I am losing steps on all axis; .5mm using 1/4 micro stepping after a 10,000 lined program. The error only reduces (but still present) if I mico step down to 1/8 or 1/16 and of cause becomes a higher loss if i step up to 1/2 stepping.

I think I have read every post covering this issue over the past 6 weeks and tried every possible fix, a short list follows:

Reduced motor velocity & Acceleration cross various ranges to almost zero.
Changed the kernel speed.
Tried Sherline.
Been over the PC for hidden programs and run tests, disabled the tool path preveiw.
Tried all pulse widths.
Checked all wiring many times.
Checked voltages.
Tried backlash compensation.
No mechanical binding.

This has always been the same problem from day one, any changes or tests still leaves me with the same consistent missed steps?

I am wondering if ordering another BOB would be the next step as I do not have a scope for testing, ideas anyone??? :drowning:

3 axis CNC
CNC DB25 breakout board
Nema23 249oz stepper motors
CW230 3A drivers
350w 24v pwr supply
Desk top PC Intel Celeron 1.70GHz 224MB Ram
XP 2002 service pack 2
Mach3 R3.043.022

Thank you

[awerby]

It sounds like there might be a basic mismatch between your motors and your driver/power supply combination. What's the nameplate voltage on those steppers? Do you know how much inductance they've got? 24 volts is really low for getting any degree of performance out of a stepper, particularly those relatively high-torque steppers of yours. The easiest thing to try might be to switch to a 36 volt power supply (more amps would be good too, since you're only getting about 1.6 out of a maximum 3.0), and see if things got better. I'd do that before I changed out the breakout board, anyway.

Andrew Werby
ComputerSculpture.com — Home Page for Discount Hardware & Software

[109jb]

awerby is right. In researching steppers I have yet to see one that really wants less than about 40v. The general rule of thumb is to have voltage at sqrt(inductance)*32. I'm going to guess your inductance is about 2 mH for a voltage of 45 v that the motors really want. You could probably do 48 v without problem and could di that by getting another 24v power supply and wiring in series. However, your drivers need to be able to handle this too.

On my machine I initially lost steps and found it was due to the pulse width that the drivers wanted. I use linuxCNC and there are specific settings for this. Not sure about Mach3 or other software. I increased those settings and no more lost steps. Incidentally, I am running 570 oz- in nema 23's that can take 50v and 5a, but i am running them at 24v and 3a without problem.

[David Lucas]

Hi awerby & 109ib

Thanks for your replies, this is an area I had not covered due to the lack of knowledge but I had had thoughts about going up to a 36v power supply. My drivers are listed to 36v so maybe 48v maybe to much?

Can you have alook at my spec's and advise...

The motor specs:
Nema23 249Oz
Holding torque 1.8Nm
Voltage 4.2v
Current 3A
Resistance 1.4 Ohms
Inductance 3.5 - 20% mH

CW230 Drivers:
24 - 36vdc
Current 0.9-3A

Power supply S 350 24:
AC input 90 - 132VAC / 180 - 264VAC
DC output 24V
Rated current 14.6A
Rated pwr 350.4W

[109jb]

With those specs your motors would perform best at 60v but your driver rating is limited. Speed is about proportional to voltage and torque is about proportional to amperage. At 24 v you are at about 40% of what your speed could be. What type of screws (ballscrews, acme, etc) and what pitch do you have on the screws?

[David Lucas]

Hi 109ib

Chinese 16mm ballscrews, single start, pitch 5.

I am happy to upgrade the drivers if it will solve my problem as well as improving the performance of the machine.

Thanks again

[ger21]

The general rule of thumb is to have voltage at sqrt(inductance)*32

That's the maximum voltage you should use. The machine should still run fine at 24V, but it will only run half as fast as the motors are capable of.

Since you're losing steps on all axis, my guess would be either the PC or the breakout board. Can't guarantee it's either of those, though.

If you can find another PC to try for free, I'd try that first. Then I'd try the breakout board if it's not the PC.

[David Lucas]

Hi

I'm at my 2nd home for now (work) but will try another PC tonight and let you know?

Thanks

[David Lucas]

Hi

I have tried a 2nd CPU and no luck it still has lost steps, if no other suggestions then I will order another BOB???

Thanks

[neilw20]

Have you checked the clock phase? Might need active high/low reversed.
Mostly at the lower voltages, you are just speed limited.

[David Lucas]

Hi

Clock phase, can you explain a little please.

I have tried changing the active high/low settings without any luck?

Thanks

[neilw20]

Yes, That is active hi/low on outputs.
Refer to the driver manual for requirements, and be aware that the phase may swap depending on the design of the BOB.
When the direction signal changes, it is important that the step pulse does not change at the same instant in some drivers. ie. There is a setup time for the direction signal, so the clock does not occur in the wrong (axis) direction.