Rigid Tapping Trouble <resolved, wrong encoder>

[cnckeith]

Is there a full spec for the usable rotary encoders? I see NPN, PNP and "line drivers" The latter using 8 wires as I see on diagrams to wire the Acorn 9 pin plug.


PFM Dave

Acorn Encoder post.
https://centroidcncforum.com/viewtopic.php?f=63&t=921


https://www.automationdirect.com/adc/se ... tial%29%22

[cnckeith]

Keith: did something change in the rigid tapping cycle?

It used to be, the cycle would wait for the index pulse, then begin feeding Z down in proportion to the A/B counts.

Once it had started feeding down, there used to be no dependence whatsoever on the index pulse.

In the current software, what is the mechanism whereby invalid signals on the index channel could cause "438 Spindle Slave position error"?

no changes in tapping cycle.

i have no idea how having index on all the time would results in a 438, i just know you want the index once per revolution so lets solve that one first!

at this point in the debug process i would have swapped out the encoder long ago.

[PFM Dave]

Update,

I tried swapping the Z lines at the DB9, at this point the * is off and blinks when the spindle is turned by hand. The RPM displayed a negative value, so I moved to a -8000 in the parameters and got the RPM to display positive. Ran a rigid tap test getting the same 438 error. So I will go with Keith and replace the encoder. AS for the would have swapped it out long ago... If I had one on the shelf sure however I do not have one on the shelf.

More after I get a new encoder.

PFM Dave

[PFM Dave]

Another update:

Replaced the original encoder with a new 1000 line encoder from automation direct. I wired it up to a fresh DB9, yes included the shield to the DB9 shell. The "*" in the PID screen now only blinks 1 time per revolution. The absolute position displays as a negative count, the spindle setup is set to -4000 and the RPM when running clockwise is a positive number. I then used Intercon to create a new rigid tapping program for a 10-32 tap. I did edit out the tool height offset and just used the tap to the top of block for zero Z. The first run did NOT generate a 438 error but did not tap the block. Making the height adjustments did produce a tapped hole! however the second attempt produced the dreaded 438 spindle position error. I stopped for the day.

Today I reran the test 10 times to see what would happen and to my surprise did make a successful tapped hole on try 9 of 10. Not great odds but the operation ran perfectly, tap moved to top of block, threaded the hole, stopped paused and reversed out of the hole. But why 1 in 10 times. I only cleared the error(s) see picture, and re ran the same program. I do see a new error on some not all of the failed tapping attempt's an error 327. I am not sure what causes this error or why it would be followed by the 438 error. This does appear to be a new piece of the puzzle.

I believe I have included all the usual required files from the current configuration of the machine. It is getting closer and shows much better signs of future success.

Thanks in advance for any additional help. I am going off to research the 327 error now.

Regards,

PFM Dave

[cncsnw]

Interesting.

When the "438 Spindle slave position error" condition happens, had the Z begun to feed down (from the +0.100" clearance plane), as before?

Roughly what range of Z positions did it stop at, when it stopped with "438 Spindle slave position error"?

I do not have any good reason to think this would help, other than that it is loosely related to getting up to speed at the start of a tapping cycle, but:
Try changing Parameter 240 from 0.10 to 0.05 and see if that makes any difference.
Alternately, you could try changing Parameter 240 to 0.20.

[cncsnw]

A more promising avenue would be to reduce the value you have in Parameter 69.

Right now you have P68 = 600 RPM and P69 = 3 seconds. You also have P82 = 500 degrees.

That says that, when approaching the bottom of a hole in rigid tapping, the control should:
1) Reduce the spindle speed to 600 RPM
2) Run at that speed for three seconds, to ensure that the commanded speed has been attained
3) Turn off the spindle motor
4) Expect the spindle to turn another 500 degrees while winding down to a stop.

Those things are used to calculate how far above the target final Z to begin the slow-down-and-stop procedure.

Since three seconds at 600 RPM is 30 turns, and 500 degrees is 1.38 turns, the control is being told to begin preparing to stop 31.38 turns above the bottom of the hole. With a 32 TPI thread, that would be 0.98". Since your hole is only 0.80" deep (measured from the R plane), the software may be thinking it is too late to stop in time, even before it has started.

Try reducing P69 to maybe 0.5 seconds.

You might also ask for a slower RPM in P68, but you would need to change the minimum speed you allow on the Control Configuration screen before that would do any good.

[PFM Dave]

CNCSNW,

I can go make some changes and test again, I still do not understand why if a hard parameter is wrong why or how it successfully tapped 1 in 10 holes. I seem to think this is odd. With my setup motor and pulley size 600 RPM is about as low as the machine can run. I can change the "minimum" but the mill will likely not go under 600 RPM. Thank you so much for explaining the settings and how they effect the operation of the mill.

I will go reduce P69 and give that a go.

Regards,

PFM Dave

[PFM Dave]

CNCSNW,

Well I set P69 to .5 as suggested and it ran 10 out of 10 tapped holes successfully! It does make me crazy that this long road began with a "bad" or maybe better to say wrong encoder. I hope to put together a post that helps cover this topic without so much back and forth. The key that got the encoder identified as the main culprit was the * on the PID page. This should only blink on one time per revolution, mine was on most of the time and went off one time per revolution. All other aspects were reasonable in my opinion. Displayed RPM correctly etc etc. But this error clearly messed up the tapping cycle.

One other item of interest is in Intercon, if the machine is set for rigid tapping the floating tap option is not available, turning off rigid tapping the float option becomes available but not the rigid. I for one would like to see all options available all the time.

A BIG thank you to all who helped me get to a successful conclusion to this long trail. With this and CHIPS running on the machine I am very pleased with the functionality. I hope to expand tapping to thread milling at some point in the future as well, the videos I have seen make this an interesting option especially for larger thread sizes.

Keith,
Thanks for keeping me moving forward.

Regards,

PFM Dave

[Coolidge62]

Glad you got it sorted. I bookmarked your thread for future reference.

[suntravel]

IMHO there is no problem with rigid tapping in CNC12, it is perfect doing its thing setup ok.

The encoder must be working and setup in the right direction.

Since a DIY cnc builder is just a cnc builder it is up to him to set up the parameters for his machine.

Mine are a lot faster than the default settings, but they work perfectly from 800-3000 rpm on my machines.

Uwe