Last One - 4'x6' Steel, Epoxy and ClearPath servo's

[1Jumper10]

Weerasak was kind enough to help me with the E-stop circuit to. You put all the normally closed inputs that you want included in your safety circuit on inputs X105 - X115*. Then assign X102 as your E-stop input in UCCNC. If any input is not OK, then the UB1 will cause an E-stop input and UCCNC will go in to Re-set. It also is part of the Output Signal Switching Device , OSSD, and it can open and close a power contactor if desired. Its described pretty well in the UB1 owners manual but it wasn't making sense to me until I saw how it would work with the Clear Path HLFB.

I don't know if the limit override Software button in UCCNC will let you recover from an alarm condition with this circuit? Or, if it requires an actual switch? I see the Override input on the UB1 for this purpose but I guess I'll have to experiment a little with it and report back.

Thank you Weerasak at http://www.cncroom.com for the assistance.

* - This requires the solder bridges on the back of the UB1 to be soldered closed to be included in the OSSD circuit.

[ger21]

I don't think the Limit Override will do anything in this case.
From what I've heard, , when you hit a limit, you need to still click the Reset button. The override button is necessary until the limit switch input is cleared. So you need both. In your case, there won't be a limit switch input active, so the override doesn't do anything.

What you probably want is a switch with two contact blocks. One going to the OVR input, and one to another input, and use an input trigger for Reset. So one switch resets the UB1, and UCCNC at the same time. I think this will work.

I plan on doing something similar. Using the safety circuit in the UB1 for the fault lines from my servo drives. But my servo power contactor is controlled by a PILZ safety relay, which uses the charge pump signal, and is in my E-Stop circuit. So I can't power up the servos until UCCNC is running and out of Estop.

Its described pretty well in the UB1 owners manual

I can see now where Port 1, Pin2 (X102) can be assigned to E Stop, but a little more clarity wouldn't hurt.
But yes, Weerasak is very helpful.

[1Jumper10]

I don't think the Limit Override will do anything in this case.
From what I've heard, , when you hit a limit, you need to still click the Reset button. The override button is necessary until the limit switch input is cleared. So you need both. In your case, there won't be a limit switch input active, so the override doesn't do anything.

What you probably want is a switch with two contact blocks. One going to the OVR input, and one to another input, and use an input trigger for Reset. So one switch resets the UB1, and UCCNC at the same time. I think this will work.

I plan on doing something similar. Using the safety circuit in the UB1 for the fault lines from my servo drives. But my servo power contactor is controlled by a PILZ safety relay, which uses the charge pump signal, and is in my E-Stop circuit. So I can't power up the servos until UCCNC is running and out of Estop.



I can see now where Port 1, Pin2 (X102) can be assigned to E Stop, but a little more clarity wouldn't hurt.
But yes, Weerasak is very helpful.

Recovering from an E stop might require 2 overrides: When in Reset. the servos aren't enabled so the HLFB "OK" signal is absent. The "Override limits" software button should re-enable the servos. But the OSSD override will still have to be engaged/depressed/input to finally clear the E-stop input. I wonder if I could map an output in UCCNC to the "override limits" button and tie that output to the "OVR" input? If thats possible, then the "Override Limits" software button would be overriding the E-stop and the OSSD circuit? I think this is similar to what you meant by using an "input trigger", only its a software solution? This is almost exactly what you described isnt it

Reading the owners manual I see the charge pump signal can be tied to the K3 relay. I'm going to use the CP signal through this relay to control a solid state contactor that turns on the servo power supply and the VFD. This was less of a safety feature than a method that will let the whole machine power up in two phases instead of one surgilicious "KATHUNK"! Should work OK as long as the CP signal doesnt disappear when the software is in re-set.

[OlfCNC]

Recovering from an E stop might require 2 overrides: When in Reset. the servos aren't enabled so the HLFB "OK" signal is absent. The "Override limits" software button should re-enable the servos. But the OSSD override will still have to be engaged/depressed/input to finally clear the E-stop input. I wonder if I could map an output in UCCNC to the "override limits" button and tie that output to the "OVR" input? If thats possible, then the "Override Limits" software button would be overriding the E-stop and the OSSD circuit? I think this is similar to what you meant by using an "input trigger", only its a software solution? This is almost exactly what you described isnt it

Reading the owners manual I see the charge pump signal can be tied to the K3 relay. I'm going to use the CP signal through this relay to control a solid state contactor that turns on the servo power supply and the VFD. This was less of a safety feature than a method that will let the whole machine power up in two phases instead of one surgilicious "KATHUNK"! Should work OK as long as the CP signal doesnt disappear when the software is in re-set.

I think you can attach outputs to any buttons. Take a look at the output triggers page in the I/O triggers.
Press the Set button and select code 27. Limits override and assign an output pin.

[ger21]

Reading the owners manual I see the charge pump signal can be tied to the K3 relay.

Yes, if you change a solder bridge.

Should work OK as long as the CP signal doesnt disappear when the software is in re-set.

There's an option in UCCNC for charge pump always on.

[1Jumper10]

I think it can be done.

Gerry and OlfCNC thanks for the advice. I should get the UB1 this week.

[1Jumper10]

After a considerable setback I've begun to make progress again. Just to re-cap; I burned out all the outputs on my MB2 breakout board trying to use a sinking output as the source for the FWD RUN command on the Hitachi WJ200 VFD. I contacted CNCroom for some assistance. They were very helpful and in fact, instrumental in getting the VFD connected properly to my new breakout board, a UB1. But it even took them a couple times before they got it right. Nice to know I wasnt in fact as stupid as I felt. The UB1 is a better board than the MB2 if you are using a UC300ETH and that eased the pain of a replacing it.

Moving forward I've got my circuit planning 90% completed. I've decided to keep the homing sensors. There are just times where re-homing with out having to completely power down the system is a good option. Besides, they were already installed, just not connected. In addition to the UB1 I added a 25amp solid state contactor to control power to the servo power supply and the VFD. The contactor, in addition to zero point switching, has a built in snubber circuit to tame the power surge when it first turns on. And its now controlled by the Charge Pump signal so no power to the servo's or VFD until UCCNC has control. And with the UB1 I was able to eliminate a 5 volt power supply.


Safety Circuit

- Servo Enable signals are disabled until the Charge Pump signal is present. This is a UB1 feature selected with a solder bridge. Outputs Y302 - Y307, Y316, Y317 are then controlled by the charge pump signal.
OSSD
- All HLFB (Servo On) signals are required
- VFD power on (AL0/AL2 termial on VFD)
- E-stop switch closed (normal)
Still thinking about this but I think I'm going to use the OSSD output signal to close a relay that the servo enable signals will pass through. If the OSSD becomes not error free, the relay will open and disable the servos. OSSD works independent of software so I guess it would be reliable but I prefer to route at least one enable signal through the E-stop switch .

Recover from a Re-set condition:

- OSSD override switch will be K1 relay and mapped in UCCNC to the Override Limits software button. Pressing the SW "Override Limits" button will:
-enable servos and this will return the HLFB "ok" signals to the OSSD
- re-normalize the OSSD.
I tested the K1 operation when pressing the "override limits" software button and it doesnt stay closed very long. Hopefully it will be long enough for everything to get restarted.


Power circuit:

Main 220V contactor closes -
- 24 vdc On
- 12 vdc On
- UB1 on
- cabinet fan On.
Software startup
- Charge pump signal closes K3 relay (solder bridge UB1)
- K3 sends 24vdc to Solid State contactor in turn supplying 220Vac to Servo power and VFD.
- Z axis servo hard stop homes upon power up getting it up out of the way.
- "override limits" button and "Reset" button enables servos and the Y axis hard-stop homes. X axis will home with software control, no hard stop homing.


The pic is the WJ200 VFD front circuit board after it was all wired up and working with the pulse train frequency control. The resistor is visible across P24 and EA terminal.

[NIC 77]

Nice to see you're back on track!

[1Jumper10]

Cleared a big hurdle tonight. I finished hooking up the circuits to the UB1. I connected up all the servos but i didn't install the Y axis motors in the mounts. I just placed them on each side of the table by the gantry ends and faced them toward each other. I wanted to make sure they were turning in the right direction before I installed them and that was easier this way. Plus they still need to be auto-tuned. Took a deep breath and powered it up. Everything worked as expected . I set my ports and pins, checked the homing on the X axis, tested my sensors, tested the E-stop and the Reset... after the last couple weeks, it was gratifying to have everything working like its supposed to Really happy with the UB1 board! I could see all the enable outputs turn on and off, I could see the status of the OSSD inputs, I could see the axis movement commands. Nice. And I'm not sure why, I'll have to double check, but recovering from a Reset or E-stop didnt require pushing the Override Limits software button. I could see the servo OK signals were gone from the OSSD circuit and the servo enable outputs were off. The Reset button definitely re-enabled the servos and it reset the OSSD circuit. Normally, to clear a reset, the software override button lets the servo's re-enable which provides the "Servo On" HLFB signal, which UCCNC detects as "everything is OK" and then you can clear the re-set. Apparently that interlocking function is dealt with in the UB1. To recover, you just hit the Reset. Weerasak, you magnificent bastard!

[1Jumper10]

Control box is done except for the cover and an upgraded ventilation fan and a higher wattage 12vdc power supply. Fan arrives tomorrow. PS hopefully by Friday. Wiring is done except for the 2 potentiometers for the FRO and SRO. I'll have to have a control cabinet to mount them in. The button is the override switch for the outlet right next to it. I'll plug my dust collector in to it and be able to control it with UCCNC. The outlet is split so the bottom one is always hot while the top one can be switched on and off.

The Z and X axis servos are tuned and working good. The X is set at 800 IPM. Z, a bit slower. Both axes are crisp and responsive.

I measured the backlash on the X axis today. The 10:1 GAM planetary gearboxes were sourced from eBay so their quality was an unknown variable. If I did this right the measured backlash was .0005". I used a Turlen indicator with .0005" graduations fastened to the gantry and touching the X carriage. I zeroed the X, moved it 18" left then back to 0. Repeated it several times and every time it returned to .0005" less than 0.


Sent from my Pixel XL using Tapatalk

The two meters just to the left of the VFD are the coolant flowmeter for the spindle on the top. The bottom one is the current output of the VFD. The flowmeter isn't scaled correctly for the lower flowrate I want but the effect is that any reading means I'm flowing enough coolant.

[matth]

Good work.

Although the control box looks weird without any servo drives in it!

[1Jumper10]

It's all the holes in it from all the redesigns isn't it?

Sent from my Pixel XL using Tapatalk

[1Jumper10]

Yesterday I auto tuned the Clear Path's for the gantry using a dummy load on the X axis. Loaded the configuration files into the Y axis servo's and got them installed. Tested movement, set up homing and soft limits. Settled on 800 IPM for the X and Y axes with acceleration of 30 in/sec^2. Everything seemed fine. Tested things a little more until I was sure the limits and E-stop would work if I needed them. No problems so I loaded up some G-code and ran my first program. Just air cutting, not actually making chips, still need a cutting bed. I was very happy with how well it worked. I ran it twice I like it so much.

It was kind of a landmark day. I started this build about 10 months ago and worked on it almost daily since then. There were a couple periods I just had to wait on stuff but for the most part, outside of my day job I was designing or building. And today it worked like I built it to.
I've still got to fab and install end stops, make a work surface (probably a vacuum table), cabinet cover, refine the work station, make a touch probe and a fixed probe for auto tool changes, get Gerry's 2017 screen set and a few other things. But for the most part, its done. When I get a bed on it I'll post a YouTube video of it in operation and update this thread. I'll probably post a few more pics now and then just to wrap it up. Thanks to those of you who posted in this thread with suggestions tips or just encouragement. Especially, TeknicTom, Nic77, LouieAtienza, Mactec54, and Ger21.

Chris

[1Jumper10]

Got the sub spoil board surfaced and epoxied. Epoxied both sides and the bolt holes with some really thin epoxy that soaked in pretty well. Should seal it from moisture intrusion and vacuum loss. The top will be a vacuum table with some T slots. It will span all the way across the I beams to the outer edges.

Sent from my Pixel XL using Tapatalk

[NIC 77]

Nice work! As always.

[Bongodrummer]

Awesome to behold! How did the servo tuning go for the duel x drive? Do the servos track each other perfectly when slaved? I am making a duel ballscrew x-axis, and am umming and ahhring about whether to mechanically link the two screws with a belt, or not to bother... Looks like I will be going the clear path rout too, so would be cool to here your views...
Thanks.

[wallyblackburn]

Yep, that's the reason. Just like water it seeks its own level. It's a low molecular weight, low viscosity, slow curing epoxy with some incredibly small shrinkage percentage of cured versus uncured. So now in effect, everything black in the pic is on the exact same geometric plane.

Sent from my Pixel XL using Tapatalk

I know this was a few months ago, but can I ask what kind of epoxy you used?

Thanks,
Wallace

[1Jumper10]

Awesome to behold! How did the servo tuning go for the duel x drive? Do the servos track each other perfectly when slaved? I am making a duel ballscrew x-axis, and am umming and ahhring about whether to mechanically link the two screws with a belt, or not to bother... Looks like I will be going the clear path rout too, so would be cool to here your views...
Thanks.

Hi BD -
The auto tuning for the gantry drive motors took me a couple of attempts. Part of the reason was getting a realistic dummy load that accurately replicates the actual loads the drive motors will be moving when they are finally installed. My dummy load added some oscillation that caused my first attempt to error out. 2nd attempt used a more compact load and I limited the torque to 75% of max and that was successful. When I loaded the auto tuned motor config files into the gantry motors I set the torque limit up to 100% they worked fine. I even left the RAS setting at 16ms which is pretty short. Sometimes, I'll hear them hunt for a position but its pretty rare and very short.

In operation they move the gantry which weighs about 400lbs at 800 IPM, with accel set at 35 in/sec^2. It could go faster, but right now thats fast enough. My first half hearted attempt at running some G-code resulted in a ruined 60 degree V-bit, 1/4" shank. I'll post a pic. I was cutting a V-carve design in a piece of particle board, 1/2" DOC, 500 IPM. I set zeros and hit the start button. Stopped. Anchored my work piece more appropriately and started over. It was just hogging the material out! I had to slow down to 450IPM because my spindle was drawing 7.5amps and i didnt want it to stall. After a minute or so I hit cycle stop because I wanted to try the "Run from here" command in UCCNC on a work piece that was just scrap anyway. I thought I had everything set appropriately and hit Cycle Start. The axis started but the spindle did not. The Z axis drove the cutter 1/2' deep into the board and the the G code kept running. Before I could hit stop the cutter shaft was bent at about 45 degrees. I learned 2 things: 1) I need to re-configure the alarm output on my VFD so it shuts down the program under similiar conditions in the future. 2) The Clear Path servo's didnt care a bit that they had turned my 60 degree V-cutter into a drag knife and that it was plowing a 1/2" deep groove in particle board. At least for a couple inches before the shaft bent.

If I were you, I wouldnt bother linking the ballscrews with a belt. The servo's wont need the belt to maintain their position relative to each other. I would recommend that you set up hard stop homing for two motors slaved to the same mechanical system. I will be doing that in the near future but i've got to make some end stops first. On second thought, you could easily drive two ballscrews with a belt and one ClearPath servo. That would really simplify your auto-tuning.

[1Jumper10]

I know this was a few months ago, but can I ask what kind of epoxy you used?

Thanks,
Wallace

Wally - I used SC15P epoxy from here Precision Epoxy Products

Mike is a good guy to deal with but he only takes checks.

[1Jumper10]

V-cutter after the machine used it as a drag knife.

Sent from my Pixel XL using Tapatalk