- FineLine Saturn 4x2 with ClearPath, UC300-ETH / MB2

[DDgitfiddle]

Glad to hear you got your spindle sorted out. I had missed your update.

I squared my gantry with a 3-4-5 triangle. And just as a side note if, you make the triangle with UCCNC but then measure the hypotenuse with a ruler, you will go crazy trying to get it square. That little excercise will definitely highlight the differences in measuring sticks. Anyway, once I laid out 5ft with the machine and measured the hypotenuse with that, I got it square. With the CP servos, you should really set up hard stop homing on the gantry. After the hardstop home is run, the servos will keep it square. And I think this will synchronize the servos so they don't try to fight each. I've yet to do this on mine as I need to make the travel stops but it's on the short list.

Sent from my Pixel XL using Tapatalk

I am drilling holes in a square pattern then measuring the diagonals over 1/4" steel dowel pins. That way the actual machine calibration is less critical. I sized the square for 22" diagonals so that I could use my framing square for the measurement. That allows me to measure "over" the pins. I figured out a way to get a better measurement tonight. I used the framing square then used a 6" precision scale (accurate to 0.01") to measure from the pin to one of the marks on the square. That gave me a +/- value for the diagonals, and I can convert that into an amount to adjust the gantry. Right now I need to move the gantry by about 0.02" on one side.

When you say "hard stop homing," are you talking about using the torque-limiting feature built in to the ClearPath motors to bump up against the hard stops to indicate home? I suppose that could be more repeatable than using the prox sensors, but I have a couple of concerns about using that with both gantry motors. I tried using both prox sensors to home each motor independently with the idea that I could adjust the prox sensors to square up the gantry. I only tried it once, but one of the motors ended up stalling out and shutting itself down. My theory is that the gantry and bearings are stiff enough that the motor couldn't flex them. Sounds a little hard to believe that it couldn't make up less than a millimeter that way, but it didn't work too well. The other smaller problem is that my hard stops are pretty soft rubber, and might not make a great surface to home off of. I could replace those with harder plastic pretty easily and they would need to be adjustable. The other small problem is that UCCNC is expecting a home signal from something. I haven't really thought about how that would work.

So far, I don't THINK the motors are fighting each other. I have them slaved in UCCNC and that seems to be working OK. I feel like I am going to have to adjust the gantry a little no matter what. Right now I'm hoping I can just barely loosen the gantry bolts then jog one of the two motors by the right amount then lock everything down. There is probably an easier way to make this small of a tweak. It might be alot easier to use a dial indicator to tweak it in. I'm open to ideas.

Thanks,
Robert

[1Jumper10]

Hard stop homing is a feature in the CP servos and it works totally independent of UCCNC. I find the most beneficial use for it is to square the gantry. You could also use it on the X and Z axes and do away with home sensors altogether but I find that in actual use of the machine, its more convenient to home to a sensor with UCCNC.
You can tell if your servo's are fighting each other by plugging into them with the MSP software and seeing how much current they are using to maintain their position when they are just sitting there. It should be close to 0. But if theyre in the 10 to 40 % range then one wants to be in a slightly different position than the other. You can "relax" this condition by hitting your E-stop or Software Reset and then re-enabling. You'll see the motor current drop to near zero. But by doing this one motor will have moved a small fraction. This is where hard stop homing really helps. When initiated, each servo travels independently to an end stop that forces the gantry square and and each servo backs off the same amount (set to your preference). From that reference point the servos are synchronized and the gantry is square. You can initiate hard stop homing 2 ways: upon first power up or, upon first enable. I use it on my friends router and it works pretty well.

[ger21]

Not sure how you initiate the hard stop homing?
But if you don't setup any home switches in UCCNC, just do your hard stop homing, then the Home All in UCCNC, which basically homes in place.

[1Jumper10]

Not sure how you initiate the hard stop homing?
But if you don't setup any home switches in UCCNC, just do your hard stop homing, then the Home All in UCCNC, which basically homes in place.

You just set it up in the servo firmware to execute when they are first powered up. Or, when they are first enabled. It's definitely not as convenient as homing with UCCNC which can be done anytime. And technically, if the servos ever lose there Enable signal such as a Reset, then without the ability to home to a sensor, you'd have to power down and restart to have them hard stop home to re establish their square reference. That's why having a home sensor is convenient. You can get back to a known starting point after an E stop or Reset and continue on if you need to without having to power cycle the entire machine.

Robert, how did you home the Y axis independently to two sensor's in UCCNC? As far as I knew that feature wasnt available.

[ger21]

Robert, how did you home the Y axis independently to two sensor's in UCCNC? As far as I knew that feature wasnt available.

For Slaved homing, just set up a separate home switch for the master and slave, and it does it automatically.

[1Jumper10]

For Slaved homing, just set up a separate home switch for the master and slave, and it does it automatically.

Did not know that. Thanks.

Sent from my Pixel XL using Tapatalk

[DDgitfiddle]

Thanks for the responses, gents.

I will take a look at my motors to see if they are fighting each other, but I doubt they are right now. I am only using one home sensor right now, so it is not doing anything to square the gantry. When I power up, the motors shouldn't be fighting each other, so as long as they stay synchronized, they still shouldn't be.

As Gerry said, when I tried squaring the gantry with the two home sensors, I slaved the motors together, but selected a different home sensor for each. When I tried to home, the one of the pinions actually skipped teeth (the pinions are spring-loaded into the rack for anti-backlash), then eventually torque limited. I considered the possibility that the spring preload was too low, but I have no backlash, and I am trying to heed the warnings against over-tightening these springs to avoid excessive wear. It really seems like this gantry is so stiff that it doesn't want to square up this way.

I also like being able to control when the system homes, so I will likely stay with the home sensors, rather than using the hard stop homing. As I've been using my system, I like that it comes up fairly progressively from a power down and doesn't move until I command it through UCCNC. In my previous experience, it is common to home a robotic system prior to really using it, so that is pretty natural for me. The "Home All" sequence in UCCNC seems to work nicely, also.

I do want to look into whether I can set a torque threshold to protect the system, though.

I won't really have a chance to work on this until this weekend, and I want to check how perpendicular the spindle is first. I have a suspicion that the gantry extrusion is tilted backwards, so that would involve completely loosening it up and messing up any squaring I did, anyway.

Thanks,
Robert

[1Jumper10]

I lowered the torque limit in the CP firmware when I was getting my machine set up. Errors and goof ups were much less exciting when the motor would simply fault out instead of slipping pinion teeth. As I grew more used to it and started fine tuning I increased it in increments.

Good luck on getting it squared/trammed.

[DDgitfiddle]

I'm still working on squaring the gantry, but I discovered a few other "small" issues last weekend.

First, my spindle was running backwards. I could have sworn that I checked that, but apparently not. I discovered this after I stalled out the spindle while running a 2 3/8" surfacing bit. I couldn't see it running because I had the dust shoe on. When I took the dust shoe off, I realized the real reason for the stall. I fixed it by swapping a phase at the VFD since I'm not using a direction signal.

Second, I realized that the way I was monitoring for errors on the VFD wasn't working. I wired the error relay signal on the VFD in series with the HLFB signals for the motors. That signal triggers a Reset in UCCNC, which kills the enable signal to the motors. No sense in dragging a stopped spindle through the workpiece.

Third, I realized that I was still trying to use two home sensors for the gantry. Occasionally one of them would trigger a little before the other and cause a pinion skip. I fixed that and things are alot happier there, now.

Tramming the spindle was pretty easy. Front to back was spot on, but side to side was off by alot. With one adjustment I was able to get it right while using the surfacing bit. Should be really good when using smaller bits.

Squaring the gantry is turning out to be a little bit of a pain. There is very little play in the holes for mounting the gantry, so I can't just loosen it and bump it. I am trying to un-slave the gantry motors after loosening the bolts then using the motors to make the adjustments. I keep getting spring back when I do that, so it has been an iterative process. I haven't had enough time to iterate enough, and it got worse before it got better, so I'm still seeing about 0.030"-0.040" difference on my diagonals. I will think twice before I unbolt the gantry again after getting it squared.

-Robert

[LeeWay]

I would think that you could use shims for that variance to move one side of the gantry beam forward off of the gantry riser. That should then square the beam up while the trucks are in a relaxed state.

[DDgitfiddle]

I would think that you could use shims for that variance to move one side of the gantry beam forward off of the gantry riser. That should then square the beam up while the trucks are in a relaxed state.

I think this is actually one of the shortcomings of the FineLine Saturn design. I have one of the earlier systems with an aluminum extrusion gantry beam. The beam is directly bolted to a horizontal carriage plate (truck) that the two linear bearings are also bolted to. There is not really a surface that holds the beam perpendicular to the linear rails. So, unfortunately, there is nothing to shim against. The machine is an interesting set of tradeoffs, as all machines are. In this case, the carriage plate is simple and easy to machine, saving cost on the machine at the expense of this adjustment.

It is possible that the machine was nice and square when I received it. I had to remove the gantry to get the machine into my shop. There are a few things that could be done to make it easier to adjust, but that would add cost. Its a pretty capable machine for the price. I will get the adjustment right eventually, and then I won't have to touch it again. I'm not complaining, I just need to find the time to get it done.

-Robert

[FineLineAuto]

I think this is actually one of the shortcomings of the FineLine Saturn design. I have one of the earlier systems with an aluminum extrusion gantry beam. The beam is directly bolted to a horizontal carriage plate (truck) that the two linear bearings are also bolted to. There is not really a surface that holds the beam perpendicular to the linear rails. So, unfortunately, there is nothing to shim against. The machine is an interesting set of tradeoffs, as all machines are. In this case, the carriage plate is simple and easy to machine, saving cost on the machine at the expense of this adjustment.

It is possible that the machine was nice and square when I received it. I had to remove the gantry to get the machine into my shop. There are a few things that could be done to make it easier to adjust, but that would add cost. Its a pretty capable machine for the price. I will get the adjustment right eventually, and then I won't have to touch it again. I'm not complaining, I just need to find the time to get it done.

-Robert

Robert,

This was one of the things we addressed with the new Saturn 2 series. The steel gantry was redesigned to include the supports and make it easier to realign if necessary. If you look in the image, you can see the gantry uses a total of 10 screws (6 top, 4 under the gantry directly into the beam) and all of the holes are slotted. This makes it far easier to shift around and realign.

[DDgitfiddle]

Nate,

I think this will make it easier to access some of the bolts for adjustment and will allow the gantry to move better as a single unit. That might make it easier to tap it into position. The older design suffers a little from a stack-up of tolerances between the holes that need to line up to bolt the gantry to the carriage. It all goes together, but there is not much extra play for adjustment. When I re-mounted my gantry I tried to use a framing square to make it perpendicular to the main bearing rails. I got it within about 0.030" over the length of the gantry that way. The trick was then to figure out how to move one end by that small amount without making it worse.

I have read that people will often use the two homing sensors on the gantry to square it up. This just doesn't seem to work on this machine. I think that is the result of overall system stiffness in this case. Given a choice between the two problems (stiffness vs. ease of adjustment), I'd rather have a stiffer system. Adjustment is a one-time problem. Stiffness affects every move the machine makes.

I managed to get the gantry adjusted to within about 0.002" over the length of the gantry last night. That translates to about 0.004 degrees and a difference of about +/-0.01" in the 22" diagonals across a 15.563" square. I am approaching the limits of what I can measure, so I am going quit while I'm ahead. When I start thinking about the resolution and accuracy of the rest of the system, the fact that I'm cutting wood, and the fact that all of the cuts in that direction will be less than 26" (the range of motion), I think this is good enough. I try to think about this in terms of furniture joinery. If two parts are out of square by 0.002" over 40" it is not going to make much difference. You can see way more movement just due to humidity. I can always go back and tweak it one more time if I find that this justification is wrong.

[FineLineAuto]

Glad you got it square. Looking forward to seeing some parts come out of this.


Sent from my iPhone using Tapatalk

[DDgitfiddle]

So this was not what I figured my first "real" part would be, but what the heck...



This is a plastic storage box that I have drilled a 9x14 grid of tiny holes in on a 20mm pitch. The holes were drilled with a 0.4mm dia. end mill. I have no idea what size the holes actually measure in diameter, but I can say that each one is actually a hole. This is a quick and dirty test rig to simulate rain for my job. At about half full, I get a nice steady drip from 126 little holes.

Nothing too tough or exciting, but it does show a few things about the system. First, the runout on the cheap Chinese air-cooled spindle must be pretty good because the bit ran pretty true, and I didn't break even one (I had 9 backups just in case ). Second, I used a really slow plunge rate (5 in/min) to avoid breaking the bit, and the z-axis did a nice smooth job with it. The 200 in/min moves between holes were almost instantaneous. All the time was spent drilling the holes. And it should go alot faster than that when I get the courage to do it. I'm also impressed at how well these bits cut polypropylene for something like $7 for ten of them.

I'm still working on the spoil board, so I just taped this down to a piece of plywood clamped to the frame. Hopefully I will have time to machine the spoil board pieces this weekend.

-Robert

[FineLineAuto]

So this was not what I figured my first "real" part would be, but what the heck...



This is a plastic storage box that I have drilled a 9x14 grid of tiny holes in on a 20mm pitch. The holes were drilled with a 0.4mm dia. end mill. I have no idea what size the holes actually measure in diameter, but I can say that each one is actually a hole. This is a quick and dirty test rig to simulate rain for my job. At about half full, I get a nice steady drip from 126 little holes.

Nothing too tough or exciting, but it does show a few things about the system. First, the runout on the cheap Chinese air-cooled spindle must be pretty good because the bit ran pretty true, and I didn't break even one (I had 9 backups just in case ). Second, I used a really slow plunge rate (5 in/min) to avoid breaking the bit, and the z-axis did a nice smooth job with it. The 200 in/min moves between holes were almost instantaneous. All the time was spent drilling the holes. And it should go alot faster than that when I get the courage to do it. I'm also impressed at how well these bits cut polypropylene for something like $7 for ten of them.

I'm still working on the spoil board, so I just taped this down to a piece of plywood clamped to the frame. Hopefully I will have time to machine the spoil board pieces this weekend.

-Robert

Looks good! How was the accuracy between holes?

[DDgitfiddle]

Honestly, I didn't measure. It wasn't the sort of thing that was too critical. I'll try to take some measurements tomorrow to see. Just looking at it, though, the holes lined up pretty nicely. Once I get the spoil board in and faced off, I'm planning to try some accuracy tests (small circles, maybe some dial indicator tests)

-Robert

[rcmcdee]

Hi Robert,

Based on this thread I am about to build my first CNC machine. We have ordered the Saturn 2 and I am ready to get the electronics on order. I went back through the thread and did not find a parts list as you had indicated at the top of this thread. I would greatly appreciate a full list (including the miscellaneous bits and connectors) if possible. If I just missed the list in the thread please let me know where it is. I really look forward to the benefit of your experience,

Thanks In advance,
Robb

[DDgitfiddle]

Hi Robert,

Based on this thread I am about to build my first CNC machine. We have ordered the Saturn 2 and I am ready to get the electronics on order. I went back through the thread and did not find a parts list as you had indicated at the top of this thread. I would greatly appreciate a full list (including the miscellaneous bits and connectors) if possible. If I just missed the list in the thread please let me know where it is. I really look forward to the benefit of your experience,

Thanks In advance,
Robb

Robb,

I started out with the best intentions of fully documenting the electronics side of the build, but I have been procrastinating on updating everything. This is a good excuse for me to work on it, so I will try to get the bill of materials and schematics updated and in one section of this thread as soon as possible.

-Robert

[rcmcdee]

Robb,

I started out with the best intentions of fully documenting the electronics side of the build, but I have been procrastinating on updating everything. This is a good excuse for me to work on it, so I will try to get the bill of materials and schematics updated and in one section of this thread as soon as possible.

-Robert

That would be great. Your build and one other is what convinced us to go down the ClearPath route. We started with the idea of building a CNCRP machine that included the integrated control package. After reading various threads in this forum now believe that our goal of optimizing the quality of finished part (not the speed of production) is best served by the extra money and work involved in your approach. I am new to this CNC game but long on 3D design and printing. My experience is that the quality of output goes up the deeper I get into the guts of the process.

Thanks again for the thoroughness and attention to detail you have provided so far.