Clough42 ELS on PM1130V

[Just for fun]

• I finished my ELS install a week or two back. Here's a picture.

In adding the picture, I see that I still need to do a anchor for my cable to the controller/display (so not quite complete). But, it's in and it works. I ran for a bit with the original firmware from Clough42. But, I've already switched over to another branch of the code. The version I'm running now lets you

• show the angle of the spindle (in place of showing RPM).
• It allows entering "any" pitch for a thread, rather than a select few.
• And, it allows a "Thread to Shoulder" function that's pretty slick. Once you set the shoulder (stopping) position and the beginning position, it moves the carriage to cut up to the shoulder, then stops the carriage. At this point, you need to stop the spindle and then retract the cross slide. The firmware then moves the carriage back to the beginning location (even though the spindle isn't turning) and when there, you wet the cross slide (and or compound) for the next cut. Turn on the spindle and it cuts the next pass.

This version of the firmware is available here:
https://github.com/kwackers/electronic-leadscrew

I'll also note that the correct v-belt to go direct from the servo motor to the gearbox input shaft for my PM-1130v is a belt that SDP-SI doesn't carry. I started with their 575mm long belt and it is definitely too long. From Vbeltmart.com, I ordered a 570mm and a 565mm length. The 565mm is too short, but the 570mm is perfect.
View attachment 474284

The Branch code sounds pretty cool, but not being code savvy, I looked at the link and I'm not getting much out of it yet. I currently have my system on the work bench working with the Clough42 code. I'm slowly working on brackets for my lathe.

 

[javaduke]

• I finished my ELS install a week or two back. Here's a picture.

This looks really good, much better than my setup!
Have you figured out STEPPER_MICROSTEPS_FEED and STEPPER_RESOLUTION_FEED yet? I have a suspicion that my math is wrong and the feed is too fast. Wondering what values you came up with

 

[rronald]

• My leadscrew is 8 TPI.
• Stepper_Microsteps is 3 (because I run with the gearbox in the 3 to 1 reduction settings).
• Stepper_resolution is 1000 (to match the servo)

The above numbers need to be perfect to get threads to work.

• Stepper_Microsteps_Feed is 4.71 (has the 3 to 1 of the gearbox and whatever additional gearing the carriage includes)

My attempt at guessing the reduction gearing for the carriage was way off. So, I determined microsteps for the feed experimentally. Turned the spindle by hand and measured how far the feed moved per revolution (I now have a DRO, but a dial indicator would work). I did 10 turns of movement to improve measurement accuracy versus a single turn. I got it as close as I could using a sharpy mark for determining angular position at the start and finish.

Since I now have the new firmware, I could measure more precisely since I could have the ELS firmware display the spindle angle to better indicate the start and stop of my 10 spindle rotations. The sharpy I marked the spindle with is only precise within a couple of degrees. If I was off by 2 degrees to one side at the start of the test and 2 degrees on the other side at the test finish.....I'd be just over 1.1% off on a single revolution but only .11% off when measuring 10 revolutions.

The new firmware can show degrees and tenths, so the ratio could be calibrated about 40x more accurately if one were so inclined. However, one percent accuracy on the feed rate is plenty close enough for me and even if I had a better number, I'm not sure I'd bother to re-compile and load the code..

 

[mcdanlj]

I'll also note that the correct v-belt to go direct from the servo motor to the gearbox input shaft for my PM-1130v is a belt that SDP-SI doesn't carry. I started with their 575mm long belt and it is definitely too long. From Vbeltmart.com, I ordered a 570mm and a 565mm length. The 565mm is too short, but the 570mm is perfect.

For anyone else (not @rronald!) who ever needs a closed loop belt size that they can't find, I have posted a DIY solution that can be made to work easily and accurately for any belt tooth form that interdigitates, and which is in my testing very strong.

Making custom length timing belt loops

When you need a closed loop belt of a length that you can’t buy, there have been several techniques recommended, including diagonal cuts with pins through teeth, and cutting through the middle of two teeth and sewing them together. However, the strength of the belt is in the fiberglass or steel...

forum.makerforums.info

That firmware sounds great - especially ‘thread to shoulder’ .I will investigate.

My branch is based on kwackers' branch, but adds more thread pitches, including pitches that can't be cut with the original or kwackers (by a 10x increase in precision). It also serves as an example of how to work with odd gear ratios by finding integer solutions for STEPPER_RESOLUTION by providing a one-liner python script to find solutions. (In my case, I need STEPPER_RESOLUTION 1925 to get integer solutions for my 6:1 input combined with 275:144 gearbox settings, so I needed to program 168 steps/rev in the servo!)

GitHub - johnsonm/electronic-leadscrew: Lathe electronic leadscrew controller (clough42/kwackers)

Lathe electronic leadscrew controller (clough42/kwackers) - GitHub - johnsonm/electronic-leadscrew: Lathe electronic leadscrew controller (clough42/kwackers)

github.com

 

[rronald]

I've just completed a kit for a knurling tool and it's surfaced an ELS issue.

The servo doesn't have sufficient torque to do the knurling. This is an operation at low speed for a low feed rate and quite a bit of tool pressure. The operation begins and then I hear a long beep. Not sure if it's on the Clough42 control board or the servo itself, but the meaning is inescapable. The servo is now "on strike" and the lead screw stops moving. Power cycling gets the servo/ELS back to operational status. But the load imposed reliably trips it off.

I'm thinking that I need some additional gearing.

Any ideas?

 

[WobblyHand]

I've never done power feed knurling. Hand feeding works ok for me. I've not knurled using my ELS, guess it never occurred to me. I'd bet I'd need to up size my system for something that I don't do that often. At least for me, it's not worth the expense. Might be for you, though.

Is your knurler a scissors type?

 

[rronald]

Yep. It's the almost embarrassingly elaborate Hemingway Sensitive Knurler.

 

[mcdanlj]

I'm thinking that I need some additional gearing.

The only time I stalled my servo (that is, triggered overcurrent protection shutdown) was recently, trying to rotary broach 6mm hex in tough 4140. I then ran the carriage by hand and it was quite difficult. It's otherwise been going fine.

As mentioned above, I'm running 6:1 gears x 275:144 in the transmission, so nearly 11.5:1 from motor to leadscrew.

(I have said Hemingway Sensitive Knurler kit in parts on my upcoming projects shelf. ☺)

 

[rronald]

I'm not expecting to knurl much, so I would definitely not upgrade just for that.

I may do some testing to see how much margin I have for normal cuts. It does appear that stepperonline has (or will have) a 400 watt servo in the same frame as the 180 watt one. So, if I wanted to "upgrade" to about 1/2 HP (from my current 1/4HP version), I might just need a bigger power supply and the 400 watt servo.