[lilimike]

I don't know where to look anymore, here is my problem:

First off I am at my first steps in generating something with my home made CNC. I was able to test with engraving a picture using Microcarve V4 and the results were almost fair to good, enough to know it is 98% accurate.

Now using Lazycam I generated text and posted G-Code and started to engrave on Plexiglas at a deepness of 1/16" but at or near the third letter the Z axis starts going down to the point where it goes through the Plexiglas 3/16"

This is my code up to where I start having a problem:

N10 (Default Mill Post)
N15  G91.1
N20 G0  Z0.2000 
N25 M3 
N30  X0.1407  Y0.3134 
N35  Z0.1000 
N40 G1  Z-0.1000  F8.00 
N45  X0.1400  Y0.2650   
N50  X0.1393  Y0.1890   
N55  Y0.1329   
N60  X0.1407  Y0.0149   
N65  X0.1563  Y0.0142   
N70  X0.2544  Y0.0163   
N75  X0.2721   
N80  X0.2750   
N85  X0.2778   
N90  X0.2800  Y0.0156   
N95  X0.2821  Y0.0142   
N100  X0.2835  Y0.0135   
N105  X0.2842  Y0.0121   
N110  X0.2849  Y0.0099   
N115  X0.2856  Y0.0078   
N120  X0.2849  Y0.0064   
N125  Y0.0050   
N130  X0.2842  Y0.0036   
N135  X0.2835  Y0.0028   
N140  X0.2828  Y0.0021   
N145  X0.2814  Y0.0014   
N150  X0.2800  Y0.0007   
N155  X0.2785  Y0.0000   
N160  X0.2778   
N165  X0.2771   
N170  X0.2757   
N175  X0.2750   
N180  X0.2736   
N185  X0.2714   
N190  X0.2700   
N195  X0.2679   
N200  X0.0000   
N205  Y0.0867   
N210  X0.0007  Y0.1407   
N215  Y0.2011   
N220  X0.0000  Y0.2935   
N225  X0.0007  Y0.3134   
N230  X0.1227   
N235  X0.1407   
N240 G0  Z0.2000 
N245  X0.3725  Y0.3318 
N250  Z0.1000 
N255 G1  Z-0.1000  
N260  X0.3605  Y0.3304   
N265  X0.3505  Y0.3276   
N270  X0.3427  Y0.3240   
N275  X0.3363  Y0.3197   
N280  X0.3320  Y0.3141   
N285  X0.3292  Y0.3077   
N290  X0.3285  Y0.2999   
N295  X0.3292  Y0.2927   
N300  X0.3320  Y0.2864   
N305  X0.3363  Y0.2807   
N310  X0.3427  Y0.2764   
N315  X0.3505  Y0.2729   
N320  X0.3605  Y0.2700   
N325  X0.3725  Y0.2686   
N330  X0.3860  Y0.2679   
N335  X0.3924  Y0.2686   
N340  X0.3995   
N345  X0.4059  Y0.2693   
N350  X0.4116  Y0.2700   
N355  X0.4166  Y0.2714   
N360  X0.4216  Y0.2729   
N365  X0.4265  Y0.2750   
N370  X0.4308  Y0.2764   
N375  X0.4344  Y0.2792   
N380  X0.4379  Y0.2814   
N385  X0.4408  Y0.2842   
N390  X0.4436  Y0.2871   
N395  X0.4450  Y0.2899   
N400  X0.4464  Y0.2935   
N405  X0.4472  Y0.2963   
N410  X0.4479  Y0.2999   
N415  X0.4464  Y0.3077   
N420  X0.4436  Y0.3141   
N425  X0.4386  Y0.3197   
N430  X0.4322  Y0.3240   
N435  X0.4237  Y0.3276   
N440  X0.4130  Y0.3304   
N445  X0.4003  Y0.3318   
N450  X0.3860   
N455  X0.3725   
N460 G0  Z0.2000 
N465  X0.3285  Y0.2508 
N470  Z0.1000 
N475 G1  Z-0.1000  
N480  Y0.2423   
N485  Y0.1982   
N490  X0.3299  Y0.1592   
N495  X0.3285  Y0.0846   
N500  Y0.0000   
N505  X0.4457   
N510  Y0.0206   
N515  X0.4443  Y0.0682   
N520  X0.4457  Y0.1528   
N525  Y0.2082   
N530  X0.4472  Y0.2444   
N535  Y0.2508   
N540  X0.3285   
N545 G0  Z0.2000 
N550  X0.5830  Y0.3134 
N555  Z0.1000 
N560 G1  Z-0.1000  
N565  Y0.3020   
N570  Y0.2210   
N575  Y0.1855   
N580  X0.5823  Y0.1215   
N585  X0.5830  Y0.0085   
N590  Y0.0000   
N595  X0.4658   
N600  Y0.0107   
N605  Y0.0469   
N610  Y0.1130   
N615  X0.4672  Y0.1997   
N620  X0.4665  Y0.2466   
N625  X0.4658  Y0.2977   
N630  X0.4665  Y0.3134   
N635  X0.5650   
N640  X0.5830   
N645 G0  Z0.2000 
N650  X0.6446  Y0.3318 
N655  Z0.1000 
N660 G1  Z-0.1000  
N665  X0.6326  Y0.3304   
N670  X0.6226  Y0.3276   
N675  X0.6148  Y0.3240   
N680  X0.6084  Y0.3197   
N685  X0.6041  Y0.3141   
N690  X0.6013  Y0.3077   
N695  X0.6006  Y0.2999   
N700  X0.6013  Y0.2927   
N705  X0.6041  Y0.2864   
N710  X0.6084  Y0.2807   
N715  X0.6148  Y0.2764   
N720  X0.6226  Y0.2729   
N725  X0.6326  Y0.2700   
N730  X0.6446  Y0.2686   
N735  X0.6581  Y0.2679   
N740  X0.6645  Y0.2686   
N745  X0.6716   
N750  X0.6780  Y0.2693   
N755  X0.6837  Y0.2700   
N760  X0.6887  Y0.2714   
N765  X0.6937  Y0.2729   
N770  X0.6987  Y0.2750   
N775  X0.7029  Y0.2764   
N780  X0.7065  Y0.2792   
N785  X0.7100  Y0.2814   
N790  X0.7129  Y0.2842   
N795  X0.7157  Y0.2871   
N800  X0.7171  Y0.2899   
N805  X0.7185  Y0.2935   
N810  X0.7193  Y0.2963   
N815  X0.7200  Y0.2999   
N820  X0.7185  Y0.3077   
N825  X0.7157  Y0.3141   
N830  X0.7107  Y0.3197   
N835  X0.7043  Y0.3240   
N840  X0.6958  Y0.3276   
N845  X0.6852  Y0.3304   
N850  X0.6724  Y0.3318   
N855  X0.6581   
N860  X0.6446   
N865 G0  Z0.2000 
N870  X0.6006  Y0.2508 
N875  Z0.1000 
N880 G1  Z-0.1000  
N885  Y0.2423   
N890  Y0.1982   
N895  X0.6020  Y0.1592   
N900  X0.6006  Y0.0846   
N905  Y0.0000   
N910  X0.7178   
N915  Y0.0206   
N920  X0.7164  Y0.0682   
N925  X0.7178  Y0.1528   
N930  Y0.2082   
N935  X0.7193  Y0.2444   
N940  Y0.2508   
N945  X0.6006   
N950 G0  Z0.2000 
N955  X0.9112  Y0.1606 
N960  Z0.1000 
N965 G1  Z-0.1000

So I figured my Z axis is missing some steps therefore I ran some test using this code:

F8
G20 G90
M98 P1234 L50
M30 (stop)
O1234
G1 z-0.1
G1 z0.2
G1 z0
M99

But after the 50 loops, Z is at zero and accurate.

What can be my problem?

Thanks for any help.

Mike

[ger21]

You're losing steps on the G0 moves, and your testing with much slower G1 moves.

[lilimike]

Yes you are right, I tested again replacing G1 with G0 in my test code and I am loosing 0.4073 after 50 loops
Why am I loosing steps on the G0 moves and not on G1?
Do I have to slow down the velocity and play with acceleration?

Thank you

Mike

[ger21]

Cut the acceleration in half.

[lilimike]

My acceleration was set to 2, I tried to set it at 1 and also at 4 but still got the same results.

I had previously started another thread and I guess I should have kept it going but I figured I had a different issue and this is why I started this thread.

Actually I am starting to think my issue is I am getting old because I keep running into different problems and this all doesn't seem logical to me, it looks like there is no specific setup procedure that applies to everyone.

So going back to that other post you had suggested I go from 35KHz to 25KHzs which I did but it made things worse however I tried 40KHz and all was going right (until this present thread)

So now it is getting complicated therefore I have started from scratch. The seller of my driver board suggests I use 35KHz so I set 35KHz. Note that I have tried the drivertest and I get "System Excellent" from 25KHz up to 100KHz.

I have used the Axis calibration "Set steps per unit" on the settings screen and got steps per = 31659.90338 for accurate distance travel.
I set Acceleration to 5 and I tested velocity for maximum which was 33. Going to 34 started going slower so I reduced by 20% and set velocity to 27.2 and reduced acceleration to 2.

Now if I run this code using 9 loops, its all good:

F8
G20 G91.1
M98 P1234 L9
M30 (stop)
O1234
G0 z.5
G0 z0
M99

If I run it 50 loops, looking at the diagnostics screen while Queue Depth cycles from 20 to 19 and back to 20 it is also all good until the computer beeps and at the same time as the Queue Depth cycles from 19 to 18 and keeps going down. At that moment the stepper screams and looses about 0.23 worth of steps.
I changed Acceleration from 2 to 1 and also tried 4 with no apparent improvement. I also tried reducing velocity to 20 but still the same problem.

My steppers have 4 wires, I've tried inverting each set of wires going to coils with no change.

Is there anything I can test perhaps with an oscilloscope?

Thanks again for any help and patience

Mike

[Dorsal]

Check for a loose setscrew on the z-axis drive couplers. If there are none already, file some flats on the shafts and be sure the setscrews engage those flats.

[lilimike]

There is nothing loose. The missing steps can be heard when the stepper screams as soon as the computer beeps when the Queue Depth is no longer buffered.
I can set my code to do 100 loops and it will do at least 80 without loosing steps but almost near the end the problem happens.

Mike

[ger21]

How fast is your PC, and what does the drivertest look like? Is it a perfectly flat line, or are there spikes in it?

[lilimike]

I have a Pentium 4 processor 3.4GHz with 1 GB of RAM

I do see a few spikes in the drivertest
With 35KHz it looks like this:

And with 100KHs it looks like this:


I do not understand much of what the drivertest shows other than the status "System excellent"

Mike

[UUU]

Try doing your loops with G0 in one direction (say up) and G1 the other way (down). If you G0 both ways you may be losing the same number of steps both ways (OK the Z is special because up is harder than down). But if you lose steps on the rapid, but not on the slow, it will show up. Also this mimics the way the Z is moved in your engraving code.

[mactec54]

lilimike

Try changing the G91.1 to G90 this may make the control work for you better as well

[Crevice Reamer]

Voltage= speed. Double the Voltage, double the speed. Are you running your 52V motors on 12V? If so, try using the 24V PSU again. It should double your rapid speed--and stop the missing steps problem at 33IPM.

Maybe you should think about upgrading to a G540.
http://kelinginc.net/GeckorDriver.html

CR.

[Khalid]

CR.. is right.. Double the Voltage or Triple the voltage...

[lilimike]

UUU and Mactec,

Your suggestions might work but I feel like I would be going around the problem instead of resolving it. At this point I realize something is obviously wrong and it is either the power supply, the driver board or my steppers but my goal is to fix it so that it will work properly no matter how the G-Code is written.

CR and Khalid,
Before I had a 12 Volt supply I was using a 24 volts and I had even more issues. On my Chinese driver board there is an indication for the supply (12V to 24V) but I am now thinking that 12V is possibly the source of my problem and I have a feeling the Chinese board might actually start giving up at 23.xxxx V and/or my 24V supply may be giving out 24.xxxx + which is over the limit so I guess I will investigate the power supply.

So far I have tried 2 switching power supply (12V and 24V) I don't know if I have read or if I had a dream but can no longer find any indication, is it possible that I should be using a linear power supply instead of the switching type?

I've been 2 weeks looking at problems so i guess I can continue for a little longer until I will most likely replace all the electronics by something more serious.

Mike

[Crevice Reamer]

Many switching PSUs have an adjustment for Voltage. Maybe you could adjust it to 23~.

CR.

[lilimike]

Taking another look at the driver board it indicates 12V to 36V.
So I put back my 24V supply which can be adjusted from 23V to 25V but really this is more troublesome, the stepper sounds like a stepper when velocity is set to 2 and the max before it gives hissing sounds is 9 and even at 9 it doesn't sound healthy and it is actually loosing steps coming down. I really think now my issue is this Chinese driver board. I figured I would spend $50 instead of $150 but it turns out I will spend $150 on top of the $50 I just paid. I will wait until I have the other driver board and see how it goes before I pull any more hair out for what is left of it!

Thank you all for your help.

Mike

[lilimike]

If anyone is browsing this thread in search for a solution to fix the multiple problems of Chinese driver boards, see this thread and read it to the end as there are 2 updates from the first post.

Mike

[neilw20]

Read and follow section 5.5 of the Mach3 manual.
Get this right and you will have no troubles.:withstupi

[lilimike]

Hi Neil,

Thanks for your suggestion but that was my first approach but really my issue was the cheap Chinese driver board. With the fix (see other thread) it's working at half the speed but its working.

Mike