[Titaniumboy]

Hmm, that's a tricky one. You might want to post on cncdrive's forum if somebody more knowledgeable than me doesn't pop up here.

Good idea. I posted a question on the CNCDrive forum and, after several hours of waiting until my post got approved, got an answer pretty quickly. It turns out that I had not read the UC300ETH manual closely enough, although in my defense I had convinced myself that the problem was solely in the UCCNC software.

I implemented their suggestion and was able to enter the proper Step, Dir, and Port settings for all of the axis drives and for the spindle. Their solution also let me change the altered Active Lows back to their original Mach3 settings.

I had plugged my cable into Port 2 of the UC300ETH whereas Mach3 had used Port 1 for everything. So the simple solution was to change all of the instances of Port 1 to Port 2. Here is the UC300ETH Port layout.

Attachment 445610



Port 2 and Port 3 on the UC300ETH has the same pinout as a standard LPT printer port. Ports 1, 4 and 5 have the same pinout as a bidirectional LPT printer port. When I was trying to set the Step/Dir pins for the X and Y and Z axis while using the Mach3 Port 1 designation, UCCNC was disallowing my entry because Port 1 pins 2 -9 are set up as Inputs and the Axis Step/Dir need to be Outputs. UCCNC was letting me select Pin 1 and Port 1 because that is an Output.

Pin Number	Ports 1 and 4 and 5	Ports 2 and 3
1	Output 1	Output 1
2	Input 2	Output 2
3	Input 3	Output 3
4	Input 4	Output 4
5	Input 5	Output 5
6	Input 6	Output 6
7	Input 7	Output 7
8	Input 8	Output 8
9	Input 9	Output 9
10	Input 10	Input 10
11	Input 11	Input 11
12	Input 12	Input 12
13	Input 13	Input 13
14	Output 14	Output 14
15	Input 15	Input 15
16	Output 16	Output 16
17	Output 17	Output 17
18-25	Ground	Ground
26	5 Volt Output	5 Volt Output

Changing all the axis drives and the spindle from Port 1 to Port 2 still wasn't allowing me to have any motors being actuated. I had to change EStop pin 10 and Charge Pump pin 17 from Port 1 to Port 2 before I finally got the table and head to move for the first time under the new UCCNC controller. That was a nice feeling.

The Mach3 XML Import completely missed Flood Coolant pin 9. Again, changing to Port 2 solved that.

I had mentioned on the CNCDrive forum that I "thought that the Mach3 Import would have caught the Port differences."
Gerry (ger21) replied "The Mach3 import just reads the Mach3 .xml file, and sets the exact same settings in UCCNC. It doesn't know what controller you are using, or how anything is wired." So the Mach3 XML Import function is not, just like everyone advised, a terribly smart Import. It does just enough to get you into trouble.

[Titaniumboy]

My Mach3 screen shot is showing an OUTUT #2 on pin 14 which the UCCNC Mach3 Import function missed. Does anyone have an idea of what Novakon used pin 14 for? In fact, does anyone have a complete listing of the Novakon pin designations?

I remember slowing down my motor velocities when I first received the mill and was learning Mach3. These are my present settings.

X and Y Axis are set to a velocity of 100 inches per minute with an acceleration of 20 inches per sec per sec.
Z Axis is set to 150 and 20, respectively.
The servo spindle is set to 75 and 3; respectively.

What are more reasonable values for Velocity and Acceleration? Remember that I mechanically have a NM-145 with the electronics of a stepper Pulsar.

[upnorth]

My Mach3 screen shot is showing an OUTUT #2 on pin 14 which the UCCNC Mach3 Import function missed. Does anyone have an idea of what Novakon used pin 14 for? In fact, does anyone have a complete listing of the Novakon pin designations?

I remember slowing down my motor velocities when I first received the mill and was learning Mach3. These are my present settings.

X and Y Axis are set to a velocity of 100 inches per minute with an acceleration of 20 inches per sec per sec.
Z Axis is set to 150 and 20, respectively.
The servo spindle is set to 75 and 3; respectively.

What are more reasonable values for Velocity and Acceleration? Remember that I mechanically have a NM-145 with the electronics of a stepper Pulsar.

Pin outs for your break out board should be in the back of your operators manual. Available ion the novakon site under the "more" tab. Click that then go to downloads.

Just wondering do you have or are you planning to use a pendant?

[Titaniumboy]

Pin outs for your break out board should be in the back of your operators manual. Available ion the novakon site under the "more" tab. Click that then go to downloads.

I took a closer look at the Novakon Pulsar BOB Rev. 0 Item Description where the pin outs are listed. Pin 14 is shown as the Dir for the spindle. My Mach3 screenshots confirm Pin 14 is indeed the spindle Dir, but also show pin 14 as Output #2.

Attachment 445626Attachment 445628


I am now thinking that Pin 14 being shown on Output #2 is some sort of mistake? (On a side note, I’ve always wondered why the Y Step/Dir and Z Step are shown with Active Low set, but X Step/Dir and Z Dir do not.)

Just wondering do you have or are you planning to use a pendant?

Great minds think alike. I do not have a pendant but have been wanting one. I was going to ask about pendants on the forum after I had UCCNC fully debugged and making parts.

I was thinking about the UCCNC UCR201, but that was me just being lazy and wanting the pendant to work with minimum fuss. Not sure about the fact that the UCR201 is wireless and whether that will be reliable or not.

https://www.cncdrive.com/UCR201.html

[upnorth]

I don't really remember how Mach 3 works. I would have to do some research to figure it out. Its beyond what I can remember.

As for a pendant I have a Vista CNC P1A-S pendant. It works with Mach 3 or UCCNC and a few others. Unfortunately it does not work with a Centroid acorn. I really don't need it anymore so if you want a deal on a pendant send me a P2P. My acorn will be getting a touch screen and after that I will make the decision about replacing the pendant or not.

[Titaniumboy]

I'm not making very fast progress, but I am making some progress.


1. I have verified operation of the spindle and coolant pump.

2. I found that the X-axis was going backwards. The X+ jog key would send the table to the right, but this sends the spindle to the left, which is wrong. Changing the X-axis Dir pin 2 to Active Low fixed the problem.

3. During No. 2 above, I sent the mill into the X limit switch. I got to learn how to use the UCCNC "Override Limits" button.

4. And during No. 3 above, I discovered that the X and Z Limit Switch inputs into the breakout board were swapped with each other. I moved them into their correct positions. This wiring error has evidently been in place since Day 1.

5. I continue to become more familiar with UCCNC.


This is the cable between the red UC300ETH board and the Novakon breakout board. I thought I had been clever by installing the CNC4PC C84 IDC26-to-DB25 adapter board. But this caused the cable to go to the right and then it had to make a sharp bend downwards. This was causing more strain on the EC300ETH than I was comfortable with.

Attachment 445734



So I bought these two little goodies for around $11. I removed the stamped metal from the DB25 connector end of the ribbon cable. The second item is a DB25 Male-to-Male gender changer.

Attachment 445736



This is the After pic. Much less excess cable and no strain on either of the circuit boards.

Attachment 445738



I had to steal the screws from another DB25 connector in an effort to get the new cable connected to the gender changer. I found that the new screws wouldn't tighten down properly, probably because I was installing them on the side of the DB25 connector that had held a captive screw. No problem - I whipped out a couple of custom spacers on the lathe.

Attachment 445740



And here is the spacer and stolen screws being installed in the DB25 connector. Notice how the black plastic case has a gap? Tightening the new screws securely reattaches the black plastic case while also securing the entire cable to the gender changer and the Novakon BOB.

Attachment 445742



While my OCD is still in gear...


Before pic of the ethernet cable between the computer motherboard and the UC300ETH.

Attachment 445744



After pic of the ethernet cable. I cut off one end of the cable and used my handy dandy RJ45 crimper to crimp on another connector. I hadn't noticed until doing the cutting and crimping that this particular ethernet cable only had two wire pairs inside instead of the normal four pairs.

Attachment 445746



I was very pleasantly surprised that everything still worked afterwards.

[Titaniumboy]

...plugged those into UCCNC, then I calibrated my steppers both for steps per unit and accelerations. Your accelerations and velocities probably will be different in UCCNC than in Mach.

Open Mach then write down your Steps/unit from the calibration screen. You can plug those into UCCNC and it will get you close. You'll still have to calibrate it later. I use my 8" calipers for XY, and my height gage for Z. Step and direction shouldn't change port/pin, but you may need to adjust your step pin to active high or low. This will depend on your drivers, but having a look in mach will tell what that should be, it will be in the Mach motor tuning menu.

So at the time that CL_M posted this it went over my head. But now that I am slogging through setting up and learning UCCNC, I revisited all the past advice regarding UCCNC.

Why would the steps per unit change between Mach3 and UCCNC? As far as the stepper drivers are concerned, steps are steps are steps. Ditto for acceleration and velocity between the two controllers.

What kind of numbers are you using on your G0704 for Velocity and Acceleration? How about Homing Speed Up and Homing Speed Down?

[CL_MotoTech]

For steps per unit I just wanted to be thorough. Theoretically it shouldn't change, but why not go through that process? It can be done in a half hour for all three axis. I clamp my digital caliper to the table and then move the lower jaw with a rod in the spindle. For acceleration I think I am running 20, and 160 for velocity on XY&Z. I'm no expert, but from what I have read, Mach didn't implement acceleration and velocity very well. Whereas UCCNC has much better motion planning, but that means your acceleration and velocity settings will differ despite the hardware being the same. For homing speed I think I used 30 up and 10 down.

[Titaniumboy]

That is a good idea. I had checked the mill on all axis with a dial indicator, but that is only over an inch or less. I had never thought about using a digital calipers over a much larger distance.

So you have velocity of 160 and acceleration of 20 on all three axis. I’m assuming you are using steppers? Do you only see velocity 160 on rapids? What is your fastest speed actually in a cut so far?

My settings from Mach were velocity of 100 on X and Y, and 150 on Z, and acceleration of 20 on all three. I changed the Z axis the other day down to velocity of 100, but that was just to get the Z to match the X and Y. Perhaps I will be able to bump up the velocity on my mill as I get more time on it.

Right now the Rapid Speed Up is 50 and and the Rapid Speed down is 25, but those numbers came from Mach also. These speeds sound really high compared to yours, but my mill uses proximity switches so no damage if it overruns a tiny bit.

I have a G0704 also. The original plan was to do a Hoss CNC conversion but I finally realized that was going take a really long time. It has been handy to have the G0704 around while getting the Novakon into commission. Funny thing is that our G0704’s table is longer than my cnc mill table by four inches.

Dumb question. How do you shut down UCCNC? I have been clicking on the UCCNC icon at the top far left, but that feels weird.

[CL_MotoTech]

Yes, steppers. I am using 5tpi screw as is common. 160ipm*5tpi=800rpm. Steppers lose their torque rather quickly. I've run as high as 200ipm, but once I got the table loaded up I occasionally would lose steps. For reference, I believe stepper driven Tormach machines are around 140ipm on XY and like 100ipm on Z.

Your rapids are the maximum velocity your machine will run, so for me it's 160ipm. Your homing speed is different, and you want that to be reliable (I assume you just mixed up the terms here). Since I generally only home once, a little slower isn't a big deal, and if I need to re-home for some reason it's good to have a nice reliable homing process. I use proximity switches on my G0704 as well. That said, I am using cheap Chinese imitation proximity switches that are like $2 a piece, so they aren't likely as repeatable as other switches, so the slower speed helps.

I actually have a base, table, and column of a NM145 that I was given. I hope to someday convert it to linear rails and get it back running, but for now I am stuck with the G0704. I've also considered just making it into a fancy drill press, as I miss having a manual drill press.

I always hit the X on top right, I didn't even know you could click the UCCNC symbol.

[upnorth]

Another easy way to verify your calibration is with 2-4-6 blocks. I put two of them end to end and then use a 1-2-3 block in the vertical position. They are held together with a pair of threaded rods and some washers. Then you align the blocks with whatever axis you want to calibrate. Then using a last word indicator you pick up the edge of the 2-4-6 block. This is zero. Then raise the Z axis to clear the block. Then command a move in the direction you want to calibrate that is the same as your 2-4-6 block length. The last word indicator will then indicate on the raised part of the last block. First time you do this its important to make sure that the indicator is above the last block so it can't touch it. This is done in case something is very wrong with the initial numbers.

I can post a photo of the setup in case it is not clear from the written description.

[Titaniumboy]

CL_MotoTech
Thanks for the thoughts on Rapids and Homing speeds. Yeah, I goofed and referred to Rapid Speed Up/Down instead of Homing Speed Up/Down - good catch.

There must be an interesting story how you ended up with a NM-145 carcass.


Upnorth,
Thanks for the good ideas on calibration. I don’t have 2-4-6 blocks , but you may have given me the excuse I need to go buy a set.


———————————

More slow progress on my UCCNC transition from Mach3.

I was having trouble with getting the actual spindle speed to match the requested Speed. I held my breath and swapped some cables and went back into Mach3 today to see exactly what Mach3 was doing with spindle speed. I have even brought out the digital tachometer to measure the spindle speed.

In Mach3 the spindle had a setting of 10,000 steps per revolution and this gave a good match between the requested Speed and the actual measured spindle speed. In UCCNC I had to, by trial and error, get to a setting of 166.67 steps per revolution in order to get the requested Speed to match the actual measured spindle speed.

As a side note, under Mach3 the spindle rpm’s were varying by 20 to 30 rpm’s at 4000 rpm. Under the UCCNC and UC300ETH combo the rpm’s are rock solid, varying by maybe 1 rpm. Very very impressive.

Any idea why the spindle Steps per Revolution are so diffferent between Mach3 and UCCNC?


EDIT
I just realized that 10,000 divided by 166.67 equals 60. That seems like a suspiciously round number, but I’m too tired to realize what it means?!?

[Titaniumboy]

Ok, that’s not a good sign. I stumped the head guy over on the CNCDrive forum.

The steps per value for the spindle should be the same for MAch3 and the UCCNC, because it is defined by the spindle drive to how many steps are needed for one revolution of the spindle motor.

It looks like something about seconds and minutes, but the steps per value has nothing to do with time, so I don't know.

[Titaniumboy]

So I tried to make the problem a little smaller by asking a slightly different question over on the CNCDrive forum.

Why did the Mach3 XML Import take the Mach3 value of 10,000 spindle steps per revolution and turn it into a value of 500 in UCCNC?

Remember that I actually ended up with a setting of 166.67 to make everything match up.

No replies so far...



In other news at the Last Melon Engineering world-wide headquarters, we managed to make this.

Attachment 446000


You may be thinking that it looks a lot like the eight pieces made at the end of May for the flood coolant shower enclosure. And you would be right, since it was made using the same GCode.

But it was made under the new UCCNC/UC300ETH control system !!

It is probably my imagination, but the cutter and stepper motors sounded smoother than they did under Mach3.

Thanks to everyone who helped me on this UCCNC controller migration. I really appreciate all the help.

[Titaniumboy]

Building on my successful first part run under UCCNC on Saturday, I jumped back onto the coolant system.

And it works !! No more coolant tank leaks. And I have a nice strong stream of coolant hitting the cutting tool. Thank goodness for the shower enclosure.

Like others have observed with the stock Torus (not Torus Pro) and NM-145, the chip tray wants to pool coolant away from the right rear drain. I figure there is about a quart, maybe a tad more, that is pooling in the chip tray and table and not making it to the drain. I see a hole saw in my mills near future.

I was a little surprised to see that all of the pooled coolant had evaporated less than a day later, leaving behind oily residue.

The drain line back to the coolant tank lands in a metal tray that sits into the top of the coolant tank. It has a bunch of holes in it, and some of the delron chips were being caught in the metal tray. Not all of the chips were being caught by the tray, however. Does anyone put anything in the metal tray to help keep chips out of the coolant tank?

[Titaniumboy]

I have noticed that our Internet service tends to go out when the mill drives are energized. The mill is located right next to where the AT&T phone boxes are located on the outside wall. The internet service from AT&T is provided by these phone cables. In fact, we lose both the phone and Internet.

Any ideas or thoughts?

[Titaniumboy]

The drain line back to the coolant tank lands in a metal tray that sits into the top of the coolant tank. It has a bunch of holes in it, and some of the delron chips were being caught in the metal tray. Not all of the chips were being caught by the tray, however. Does anyone put anything in the metal tray to help keep chips out of the coolant tank?

I found this 24” x 36” x 1” washable HVAC filter at Home Depot for $9. I should be able to get at least six chip filters cut out of this one HVAC filter.

Attachment 446222

https://www.homedepot.com/p/True-Blu...4361/202195941




Let’s talk about probes for a little while. I had noted that UCCNC has some nice probing routines.

I see that Cliff Hall of Hallmark has a nice probe, as does Drewtronics. But Cliff’s probe is $750 while the Drewtronics is $170 (non-TTS) or $250 (TTS).

Are there other probe options to consider? Or is a probe a complete waste of money?

[upnorth]

I have a Drewtronics probe. Buy one it works great. If you completely screw up it will cost under $40 to rebuild it. So far it has been incredibly accurate. The price is great. Its made in the USA. When I was using Mach 3 I had loads of problems with losing position and got frustrated with resetting the reference position. I used edge finders and 123 blocks. With the probe its easy. I have the non TTS version. I have it set to a reference height but it is in a TTS holder. I honestly think buying a probe was one of the best things I ever got for my mill.

[upnorth]

I have noticed that our Internet service tends to go out when the mill drives are energized. The mill is located right next to where the AT&T phone boxes are located on the outside wall. The internet service from AT&T is provided by these phone cables. In fact, we lose both the phone and Internet.

Any ideas or thoughts?

Is the case for our milling machine earth grounded?

[Titaniumboy]

I have a Drewtronics probe. Buy one it works great. If you completely screw up it will cost under $40 to rebuild it. So far it has been incredibly accurate. The price is great. Its made in the USA. When I was using Mach 3 I had loads of problems with losing position and got frustrated with resetting the reference position. I used edge finders and 123 blocks. With the probe its easy. I have the non TTS version. I have it set to a reference height but it is in a TTS holder. I honestly think buying a probe was one of the best things I ever got for my mill.

Don’t be shy. Tell us how you really feel about the Drewtronics.:banana:

Is the case for our milling machine earth grounded?

The incoming ground is connected with a screw to the electronics case. And then there is a ground wire from there to the ground bus. I would think my machine is grounded identically to how the Pulsars are grounded.