Lost steps newly upgraded machine <clearpath, resistors>

All things related to the Centroid Acorn CNC Controller

Moderator: cnckeith

MarkRH
Posts: 41
Joined: Thu Apr 09, 2020 5:20 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No

Lost steps newly upgraded machine <clearpath, resistors>

Post by MarkRH »

Latest version of CNC12 router
Clearpath SDSK motors
3kw spindle
Shielded vfd wire with ferrites, vfd mounted in a separate box, wire on a boom far away from signal wire.
Measurements and cut quality seem to be very good, out of the dozen files ive run so far 2 of them had lost steps, the first was flattening a countertop and it lost X entirely (about 30") and just came straight back, it caused no harm, reran it and was fine.
Today it lost some of Y, it ran the gantry off the rails and time for a new sheet of white oak ply.
The HLFB relay board never faulted the Acorn board.
IMG-0941.JPG
This machine operated for a year before the upgrade, I wasnt happy with the accuracy of .02" that it was with mach and direct drive but it did run pretty reliably.
My intentions are to add a EMI filter, and resistors to step lines when I find a diagram, maybe drive a new ground rod for the sub panel this is running off of and pay close attention for any ground loops, the motor wiring is not shielded its teknic spec.
I looked up any errors under utility right after it happened and it didnt show anything, only tool touch off.
Anything obvious in the report please let me know, I dont know how to read it.

Thanks
martyscncgarage
Posts: 9912
Joined: Tue Mar 28, 2017 12:01 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: Yes
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: Yes
CPU10 or CPU7: Yes
Location: Mesa, AZ

Re: Lost steps newly upgraded machine

Post by martyscncgarage »

How long are the cables from Acorn to your CP motors? Did you install the resistors?
Please post a picture of your control cabinet.
Did the machine keep going after those "artifacts" showed up?
Or did one of the servos fault, and if it did, did you use the MSP software to look to see what the error message was?

ALSO, and very important, CHECK the mechanics of your machine, nothing loose, couplers, ball nut mounts, thrust bearings etc.
Run this backlash test that cncsnw (Marc Leonard) put on his website(Thanks Marc!)
http://www.cncsnw.com/Lashhowto.htm

Marty
Reminder, for support please follow this post: viewtopic.php?f=20&t=383
We can't "SEE" what you see...
Mesa, AZ
MarkRH
Posts: 41
Joined: Thu Apr 09, 2020 5:20 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No

Re: Lost steps newly upgraded machine

Post by MarkRH »

Marty, thanks for your reply I watch all your vids on youtube.
Cable length 32 ft.
No resistors, I skimmed over that page 47-48 in the teknic literature but the Acorn board diagram doesnt show them, thats what was confusing.
Update; I called into Teknic today and spoke with Abe, he was very helpful and suggested the most likely culprit was no 1K ohm resistors between INPUT A- & INPUT A+, INPUT B- & INPUT B+ (2) resistors per motor, on step and direction signals. I will be doing that this weekend.
Will get a pic of cabinet and post, need to tidy things up a bit.
Yes the machine kept going, the Y motors ran the gantry off the rails its a V wheel machine. (after reading other posts and with Mike Richards help I just realized that I had the HLFB relay board jumpers on wrong setting and now it should work, the relays are closing as they should now.
I have not checked MSP software, too many irons in the fire atm (and need to read up on how)
All the mechanics are tight, they are easy to check on this machine.
Have backlash set to zero, its a 3.1 belt drive rack and pinion spring tension, these sdsk motors blow thru the springs even on the tightest setting like they are nothing.
Ordering a EMI filter now also.
Will keep you posted.
Thanks again!
Richards
Posts: 692
Joined: Sun Sep 16, 2018 9:01 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No
Location: South Jordan, UT

Re: Lost steps newly upgraded machine <clearpath, resistors>

Post by Richards »

The 1K resistors are almost mandatory with longer cables. I have not needed them with the 10-foot Teknic cables; however, I installed connectors on a 100-foot roll of cable to see what would happen. The cable was left coiled up in a roll without using resistors. I expected to have problems and I had problems. Adding the two resistors fixed things, even with coiled cable. It is very helpful to use an oscilloscope to watch for noise on the signal lines. Adding resistors, EMI filters and various other devices can reduce and even eliminate noise, but a visual image lets you know whether you solved the problem. I use a Hitachi dual channel scope that I bought several decades ago. Now you can buy PC based scopes that do a better job for $100 or less.

NOTE: I do not recommend using cables longer than absolutely necessary.

Also, remember to use MSP to set the IN RANGE POSITION for the HLFB to work properly. Page 51 of the ClearPath manual briefly refers to setting the IN RANGE POSITION:

"The HLFB output asserts (conducts) when the motor is enabled, not
shutdown, and the measured motor position is In Range (±X encoder
counts) of the current commanded position. The In Range parameters are
set in the HLFB Setup dialog."

You may need to experiment a little with various settings to find a reasonable setting for each axis. Keep in mind that if the setting is too low, the motor will fault during normal acceleration and deceleration. I have my Acorn boards set for 0.5 sec ramping, which gives the servos plenty of time to accelerate without too much lag between commanded position and actual position; however, I played around with the ramping and the HLFB setup to verify that the motors would fault if pushed too hard. In particular, I have one of the smallest SDSK-23xx motors that I bought to see what Teknic's smallest motor could do. Even that small motor is very impressive, but I was able to make it fault by accelerating too quickly or by using it to move too heavy of a load. In my application, I use HLFB to stop the servos only when a gross error has occurred. Your application may demand that the motors fault with a smaller following error.
-Mike Richards
carboncymbal
Posts: 88
Joined: Tue Dec 15, 2020 11:12 am
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: 0035FF8A1900-10082038
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No

Re: Lost steps newly upgraded machine <clearpath, resistors>

Post by carboncymbal »

Richards wrote: Sun Feb 14, 2021 3:05 pm The 1K resistors are almost mandatory with longer cables. I have not needed them with the 10-foot Teknic cables; however, I installed connectors on a 100-foot roll of cable to see what would happen. The cable was left coiled up in a roll without using resistors. I expected to have problems and I had problems. Adding the two resistors fixed things, even with coiled cable. It is very helpful to use an oscilloscope to watch for noise on the signal lines. Adding resistors, EMI filters and various other devices can reduce and even eliminate noise, but a visual image lets you know whether you solved the problem. I use a Hitachi dual channel scope that I bought several decades ago. Now you can buy PC based scopes that do a better job for $100 or less.

NOTE: I do not recommend using cables longer than absolutely necessary.

Also, remember to use MSP to set the IN RANGE POSITION for the HLFB to work properly. Page 51 of the ClearPath manual briefly refers to setting the IN RANGE POSITION:

"The HLFB output asserts (conducts) when the motor is enabled, not
shutdown, and the measured motor position is In Range (±X encoder
counts) of the current commanded position. The In Range parameters are
set in the HLFB Setup dialog."

You may need to experiment a little with various settings to find a reasonable setting for each axis. Keep in mind that if the setting is too low, the motor will fault during normal acceleration and deceleration. I have my Acorn boards set for 0.5 sec ramping, which gives the servos plenty of time to accelerate without too much lag between commanded position and actual position; however, I played around with the ramping and the HLFB setup to verify that the motors would fault if pushed too hard. In particular, I have one of the smallest SDSK-23xx motors that I bought to see what Teknic's smallest motor could do. Even that small motor is very impressive, but I was able to make it fault by accelerating too quickly or by using it to move too heavy of a load. In my application, I use HLFB to stop the servos only when a gross error has occurred. Your application may demand that the motors fault with a smaller following error.
What resistor value are you using? I have seen Teknic suggest using 10k on 24v acorn outputs.
martyscncgarage
Posts: 9912
Joined: Tue Mar 28, 2017 12:01 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: Yes
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: Yes
CPU10 or CPU7: Yes
Location: Mesa, AZ

Re: Lost steps newly upgraded machine <clearpath, resistors>

Post by martyscncgarage »

This is what Clearpath recommends for systems with long cable runs:
10K resistors:
Clearpath Resistors.JPG
Reminder, for support please follow this post: viewtopic.php?f=20&t=383
We can't "SEE" what you see...
Mesa, AZ
Richards
Posts: 692
Joined: Sun Sep 16, 2018 9:01 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No
Location: South Jordan, UT

Re: Lost steps newly upgraded machine <clearpath, resistors>

Post by Richards »

Teknic has changed the recommended value of the resistor since I downloaded the ClearPath manual. As Marty noted, Teknic now recommends using 10K resistors. The manual that I have lists 1K. More importantly, to me at least, is why the resistor is used. I've snipped the equivalent circuit from page 47 of the ClearPath Manual. Note that the resistor is being used as a "pull-up" resistor. A "pull-up" resistor is used to assure that a signal is in a well-defined state. We usually think of a digital circuit, such as the pulse or direction circuit going to the driver, as being either ON or OFF. At relatively low speeds, that is the case. As the speed increases, the state can "float". That means that the voltage level is neither in the ON state nor is it in the OFF state. The Acorn Spec manual doesn't list the On/Off state voltages for the H2-H3 connectors, but it lists the voltages for the DB25 connector, where the High or ON voltage is 3.66V to VCC and the Low or OFF voltage is 0.0V to 0.44V. Voltages between 0.44 and 3.66 are "floating". They are not defined as being either On or Off. The same thing happens with the H2-H3 connectors. At a range of voltages, the circuit is Off. At another ranges of voltages, the circuit is On. In between those two ranges, the voltage is "floating".

Adding a properly sized "pull-up" resistor, insures that the circuit quickly moves through its "floating" state.

The value of the "pull-up" resistor depends on the length of the cable (built-in resistance) and the signals transition speed.

An oscilloscope is used to see the effects of using resistors with different values. A clean, square shouldered signal is what we're looking for.

To me, that's where the fun is when building controllers. That's also why I spend months bench testing components before being satisfied that the design is good enough When I know that each signal is performing at its optimum level, I know that I've designed a proper electrical interface.

Substitute 10K for 1K in the following screen shot from page 47 of my ClearPath manual (or use a 'scope to see what works best in your circuit). The resistor "pulls-up" the voltage of the diode's cathode to a high level. A diode cannot conduct current unless there is a voltage drop across the diode.
Resistor.png
-Mike Richards
martyscncgarage
Posts: 9912
Joined: Tue Mar 28, 2017 12:01 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: Yes
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: Yes
CPU10 or CPU7: Yes
Location: Mesa, AZ

Re: Lost steps newly upgraded machine <clearpath, resistors>

Post by martyscncgarage »

Mike, Teknic indicates that a long cable creates capacitance and that by the time the step signal reaches the drive, it is no longer a nice square wave, it is more of a shark tooth pattern. The resistor is said to help correct this. Have you found this when your oscilloscope traces?
Marty
P.S. Mike, check your PM messages
Reminder, for support please follow this post: viewtopic.php?f=20&t=383
We can't "SEE" what you see...
Mesa, AZ
MarkRH
Posts: 41
Joined: Thu Apr 09, 2020 5:20 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No

Re: Lost steps newly upgraded machine <clearpath, resistors>

Post by MarkRH »

Mike, Marty,

Thanks for the help, I am sure this will help many in the future also, I have the same diagram Mike has with the 1K ohm resistors, notice how the diagram says step and direction and after just calling, Abe said to use them on both and step and direction and mentioned that they were working with Marty.
Marty after your post and not having any 10k on hand I wired the 1k in series to make up 2k just on step and ran 5 files and everything came out good.

I am wondering what a 10K resistor would do to the 24v, would it bring the voltage down to 2.4?
The current has to increase, or does this matter on a pull up resistor?

Thanks
MarkRH
Posts: 41
Joined: Thu Apr 09, 2020 5:20 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No

Re: Lost steps newly upgraded machine <clearpath, resistors>

Post by MarkRH »

22 gauge wire should only be .516 ohms at 32ft
Post Reply