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

[DDgitfiddle]

After reading more posts on this forum than I can count, I decided to pull the trigger on a FineLine Saturn 4 x 2. Thanks to everyone who has posted information this site. It has been very helpful. The machine has been sitting in my garage for close to a month now because I honestly wasn't expecting it to ship in less than a week. (Thanks again, Nate!) So it caught me a little unprepared in the electronics department. I now have most of the parts, so time to start the build thread!

To start off, I considered designing my own system. I'm a mechanical engineer with a background in automation, so it wasn't too much of a stretch...until I started adding up the costs and realized that I was designing something pretty close to a CNCRP or FineLine Saturn. Would have been fun, but I couldn't beat the cost of either system, and ultimately decided on the Saturn.

I want to add in a little commentary here about what I see as the key difference between the CNCRP Pro kits and the FineLine system: the frame. The frame for a CNC router is important for two main reasons that are pretty obvious: stiffness and positioning accuracy. These are related, of course. I do believe that the FineLine frame is probably stiffer than the CNCRP frame, but I doubt that it makes a real difference for cutting wood. I did some calculations on the gantry beams, and the difference in deflection was less than 0.001". (my unit has an aluminum gantry anyway). When you look at the straightness and twist specs for steel tube vs. aluminum extrusions, there is little difference, but steel can actually be a little worse. When you add in welding (even with stress relieving), I think that there is generally a higher chance that the aluminum frame will be a little straighter. Professional grade machines are often steel, but I suspect that many of them are post-machined to get nice flat surfaces for mounting bearing rails and gear racks. That's part of the cost of these machines. With all of this, for me, cost was the deciding factor because I didn't think there would be much of a performance difference. If I had built the system I was designing, I would have used the biggest aluminum extrusion I could find, and wouldn't have worried about stiffness.

I really wanted to try to get into a closed loop motor system. I considered adding encoders to steppers, then saw the ClearPath SDSK series motors. Not quite a full servo, but closer than an open loop stepper. I've read some justifications for these on cost, but for me it was more about performance, noise, and (hopefully) ease of implementation. I am using the enhanced CPM-SDSK-3421S-ELN for the three rack and pinion motors and the regular CPM-SDSK-3421S-RLN for the Z-axis. The Teknic IPC-5 power supply seems like a natural for driving these, so that was added to the list.

I also decided to go with UCCNC for controller software, which lead me to the UC300ETH-5LPT controller board. After reading about the fairly easy integration with the MB2 break out board (thanks, ger21!), I figured this also made sense.

The rest of the equipment is pretty standard stuff: Hitachi WJ200-022SF VFD driving a Chinese air-cooled 2.2kW spindle (still to be purchased), some small MeanWell power supplies (5V and 24V) for the controller and sensor power, Pepperl & Fuchs prox sensors, and a bunch of DIN rail mounted stuff for fusing and power/signal distribution. I didn't really want to mess with the possibility of leaks from a water cooled spindle, and it seems that a properly set up air-cooled will work just fine, minus some extra noise.

One thing that I decided was important for me was the safety circuit. The Hitachi VFD has a nice safety circuit built in for killing power to the drive electronics without pulling the main power, but the ClearPath motors really didn't without stepping up to the SDSC series (for close to $100 per motor more). So I was able to find a reasonable surplus safety relay and a pretty big contactor to implement a pretty legitimate safety circuit. It will kill motor power without losing controller power and should avoid power up transients from the controller because it requires a reset button to reset the safety relay. Probably not quite up to industrial standards, but safer than alot of options I have seen.

I will be posting more, including pictures, major component list, and rough schematics over the next few weeks/months. I thank everyone in advance for their helpful input.

Thanks,
Robert

Similar Threads:

• New Machine Build- Saturn 4'x4' My Second Build the first time with Clearpath Servos
• Saturn 4x2
• New Machine Build- Saturn 4x4
• Need Help!- Clearpath disabling
• My Fineline 2X4 R&P

[difalkner]

Looking forward to the build, Robert. I would point you to my build but I think Photobucket has hosed all the photos. The Hitachi VFD is really nice and I'm glad I went with that over the Chinese VFD. Good luck with your build and post lots of photos!

David

[DDgitfiddle]

Thanks, David! I have definitely learned a ton from your build log. I can't count how many times I referred to it as I was making decisions on components and starting to lay out my control box. Thanks for taking the time to post it here. I can't promise that I will be as thorough as you were in documenting everything, but hopefully this will be a good reference for someone in the future.

Thanks,
Robert

[DDgitfiddle]

OK, time to see if I can figure out how to post files and images.

Attached are a couple of rough schematics for the system electronics to show power and grounding distribution as well as the safety circuit. I apologize to any electrical engineers out there for the non-standard documentation...not my realm of expertise. I reserve the right to update these in the future, but I will try to post modifications when they happen. Also attached is an image of the panel layout I am currently thinking about.

A few details...

The box is probably larger than really necessary. I suspect that will help with access to parts, wire routing, and ventilation. The down side was a little cost and the overall size. Sounded like a good trade-off to me. Never heard a hobby CNC builder say that they really wished they had bought a much smaller enclosure.

I split the 24VDC power into two separate power supplies. One is pretty much dedicated to the safety circuit, and the other is dedicated to system control and sensing. I did this due to a comment I found in the ClearPath manual regarding power for the motor control inputs. It indicated that switching inductive loads with the same power supply could result in voltage spikes that could damage the motor controller. The contactor seemed like it could fall into that category...maybe I'm being overly cautious, but it didn't cost much for that peace of mind.

The controller is laid out in an effort to keep AC lines away from DC signals as much as possible. I haven't drilled any holes yet, so this could change. The one thing I'm not super happy about is having the VFD right next to the motor power supply and sort of close to the controller. It does allow me to keep the spindle cable length inside the enclosure to a minimum, though, and helps keep the AC power separated. I'd appreciate some thoughts on this.





Thanks,
Robert

[ger21]

You may want to consider using the Charge Pump output from the MB2 into your safety relay, so that you can't power up the system unless UCCNC is running?

[1Jumper10]

OK, time to see if I can figure out how to post files and images.

Attached are a couple of rough schematics for the system electronics to show power and grounding distribution as well as the safety circuit. I apologize to any electrical engineers out there for the non-standard documentation...not my realm of expertise. I reserve the right to update these in the future, but I will try to post modifications when they happen. Also attached is an image of the panel layout I am currently thinking about.

A few details...

...

The controller is laid out in an effort to keep AC lines away from DC signals as much as possible. I haven't drilled any holes yet, so this could change. The one thing I'm not super happy about is having the VFD right next to the motor power supply and sort of close to the controller. It does allow me to keep the spindle cable length inside the enclosure to a minimum, though, and helps keep the AC power separated. I'd appreciate some thoughts on this.


Thanks,
Robert

I'd say you're separation is fine. You're mounting the VFD and the IPC500 to a metal ground plane. The CP servo's use 75VDC so their power inputs are inherently somewhat noise immune and I think I remember their manual stating they do fine in electrically noisy environments. My setup is similar but with a little less separation than yours I think. I haven't bench tested all components together but running the VFD and the MB2 didn't cause any problems. My laptop computer was closer to the VFD than the IPC500 was and it didn't have any ill effects from either the spindle or the VFD.

Speaking of spindle speed control, how do you plan on controlling your spindle speed?

[DDgitfiddle]

You may want to consider using the Charge Pump output from the MB2 into your safety relay, so that you can't power up the system unless UCCNC is running?

That is a good point Gerry. I'll take a look at that. The safety relay I am using looks like it has two inputs, and I am only using one for the e-stop. Should be pretty easy to wire up.

Thanks,
Robert

[DDgitfiddle]

@1Jumper10, thanks for the input. I was thinking that the 75V power supply would be OK in its metal case, but never hurts to ask people with a little more experience. The enclosure is 24" square, and I'm trying to get everything grounded correctly, so hopefully minimal problems with electrical noise.

Spindle speed control is still up in the air. Currently it would be manually set on the VFD (not a great solution). Modbus would be the nicer solution at some point, but I haven't figured out the connections yet. Advice on that with this hardware would be appreciated. I would prefer to avoid running a separate cable from the PC (laptop) if possible. Is there any way to convert one of the extra LPT ports on the UC300-ETH into an RS-485 output? (and have UCCNC control it?)

Thanks,
Robert

[ger21]

The easiest way is to use the analog speed control on the MB2.

[DDgitfiddle]

The easiest way is to use the analog speed control on the MB2.

Is there any down side to this? It obviously needs some setup in software, but no more than MODBUS (probably a lot less). Any particular concerns with electrical noise? The wire run inside the box would be pretty short, so is twisting the wires sufficient to avoid noise issues in this case? If I remember correctly, this is a 0-10VDC signal, but is it PWM'd out of the MB2 or does it use a D/A converter?

I know I could convert to MODBUS with a later upgrade, once I figure it out, so this might be a quicker way to get up and running. Still lots of things to work through that I haven't thought of, I'm sure.

Thanks,
Robert

[ger21]

Is there any down side to this? It obviously needs some setup in software, but no more than MODBUS (probably a lot less). Any particular concerns with electrical noise?

Depends on your VFD, but I'd try this first.
Takes less than 1 minute to setup UCCNC. Just assign the PWM port and pin, max and min rpm and frequency.

On my Huanyang VFD, I do apparently get a noise issue, which causes the RPM to fluctuate. But I believe that this is more of a Huanyang VFD issue, as the analog voltage from the MB2 is very steady. And yes, the MB2 converts the PWM to analog.You may not see this issue at all.

One other issue is that if you supply 10V to the MB2, you only get ±9.8V back out, which limits your top speed. Your VFD may have a few ways to fix this. Either by upping the 10V VFD output, or adjusting the RPM range to get max rpm at 9.8V. Since my VFD couldn't do either, I used a 3rd option. I configured the min and max frequency in UCCNC so that the MB2 actuall gave me a 0-5V signal, and set the VFD accordingly. This gave me the full rpm range.

You also need to use one of the relays on the MB2 to start and stop the spindle.

What I ended up doing in the end, is using Modbus to control RPM, and the MB2 relay to start and stop the spindle, as I couldn't get a steady rpm with the analog voltage. But again, I think this is a VFD issue, and would use the analog if I could.

[1Jumper10]

The hitachi VFD will accept a pulse stream input as speed control and I set up mine to use the step and direction spindle output of UCCNC. I connected the pulse stream input on the VFD it to the C axis pins on my MB2. It bench tested great. 2 wires, digital speed control so some noise immunity, no MODBUS adapters or plugins to implement.

[DDgitfiddle]

I was wondering if I could get the pulse steam to work on the Hitachi, especially since it can be a 24V signal. That sounds like a good option. It's on the list of things to work on...along with more things than I can count. Thanks for the help!

[ger21]

The step pulses from the MB2 are 5V, not 24V.

[1Jumper10]

I think the Hitachi VFD takes 3 - 27max volts for its pulse train inputs. I just looked for my DVD Manual to verify but couldn't find it. It works with mine and we have the same setup UCCNC - UC300ETH/ 5LPT - MB2 - WJ200.
If you go the Pulse stream route, when you get to it, I can tell you the configuration settings that are working for me.

[DDgitfiddle]

You may want to consider using the Charge Pump output from the MB2 into your safety relay, so that you can't power up the system unless UCCNC is running?

I was looking back at my schematic and realized that I was feeding the Charge Pump for the MB2 through the reset button for the safety relay. I *think* that this will do what you are suggesting Gerry, but if it doesn't work, I have the option of treating the charge pump as if it were another e-stop.

Thanks,
Robert

[DDgitfiddle]

I think the Hitachi VFD takes 3 - 27max volts for its pulse train inputs. I just looked for my DVD Manual to verify but couldn't find it. It works with mine and we have the same setup UCCNC - UC300ETH/ 5LPT - MB2 - WJ200.
If you go the Pulse stream route, when you get to it, I can tell you the configuration settings that are working for me.

Thanks, for the help. It looks like it should work, I just made the mistake of thinking that the MB2 put out 24V on the step pulses. I will take you up on the offer to get the config settings when I get there. I still need to get the machine on a base and into my shop (I'll save that for a different post), and try to assign the I/O on the MB2 to sensors and other inputs. If I try to wire this up without following some sort of diagram, I'll screw it up for sure. Progress will be a little slow for another week or so, but I'm hoping to put a lot of time into this when I have a week off coming up.

Thanks,
Robert

[DDgitfiddle]

The step pulses from the MB2 are 5V, not 24V.

I was looking at the MB2 manual tonight. There is a jumper that can be de-soldered to allow an external power supply to power the Vcc for the motor outputs. Do you know if the MB2 can tolerate using 24V instead of 5V there? The manual does indicate that the external power supply would be 5V. This is more of an academic question, than anything else...probably not necessary for me to to this. 24V logic can be more resistant to noise, and the ClearPath motors will accept 24V on the inputs.

[ger21]

I was looking at the MB2 manual tonight. There is a jumper that can be de-soldered to allow an external power supply to power the Vcc for the motor outputs.

Are you sure? My understanding is that the jumper disables the 5V that normally provides 5V to power the ESS on pin 26 of the ports.

Most of the board circuits runs on 5V, from a different voltage regulator.

[DDgitfiddle]

Hmmm...on a second reading, that sounds right, Gerry. I wasn't thinking about the fact that the onboard DC-to-DC converter was also supplying voltage to the electronics that run the MB2 (and optionally the ESS if that was what I was using). Using an external 24V power supply would probably release a lot of magic smoke. Thanks for keeping me on the right path.

For the sake of anyone who might pick up this thread in the future, I should let people know that a big part of my day job involves very creative problem solving. I will probably throw out a number of ideas on this thread that will turn out to be dead ends. For me, this is a way to find creative solutions and I really appreciate when people explain why something won't work. I learn, and often find better solutions because of that feedback. So, thanks again for offering the benefit of your experience to me and others.

Thanks,
Robert