Hardware_Control_Project

Post Reply
jsheley
Posts: 19
Joined: Sun Feb 17, 2019 7:27 pm
Model of CNC Machine: expensive_pile_of_junk

Hardware_Control_Project

Post by jsheley »

Hardware_Control_Project

I once again ask for the help of Vectric’s forum members to help me find the value of a motor control project idea. If you know of a forum better than Vectric’s “Hardware Forum”, then please tell me. I am not asking you for money. I would like to read your opinion on the practical usefulness of my idea.

I value the opinion of people who are knowledgeable in machine control. Many of Vectric’s forum readers (i.e. you) have experience with machine control, such as a 3axis or a 5 axis mills or 3D printers. What I would like from Vectric’s forum is an idea if my hardware control design is worth the time, effort, and money to develop.

My basic idea is to relocate the motor control (control signals, firmware, and power supply conditioning) from a centralized box to each motor. From an external hardware perspective, I want to replace the central NEMA 23 control box and multiple cable schemes with a single USB-A cable (5 ampere rated). I do not want to use USB signals, just the USB-A cable design.
A four conductor HVAC cable can be used for higher current loads such as NEMA 34 motors. HVAC cables are a viable option, but are less desirable because of printed circuit board connector issues.


(Please see attached image "proposed_3_axis_design_V0X05.png")


The basic benefit(s) for the end customer:
1) Vastly simplified motor connection(s). Power and control are combined in single cable. No central box with multiple cables.
2) Simple MODBUS type expansion upto 64 control nodes (motors, end stops, probes, coolant jets, ect). Just configure one 6 bit DIP switch for each control node.
3) Cheap, ubiquitous, commercially available replacement parts. (USB cables, HVAC cables, USB to RS485 converters, power supply)
Does the above described product exist on the market?
Do the above benefits sound appealing?
If this is a good idea, then which motor should I develop first “NEMA 23” or “NEMA 34”?
Is the smaller NEMA 23 motor relegated to “niche” hobby markets?

. I do not know if “NEMA” standards exist outside of USA. (? :?: ?) If you do not know what a “NEMA 23” motor or “NEMA 34” motor is see attached data sheets for examples of a "NEMA 23” motor and a “NEMA 34” motor.


Please share your constructive thoughts with me.
Attachments
Image of proposed design
Image of proposed design
N23_Industrial_Grade_Motors.pdf
Example of NEMA 23 motor
(501.62 KiB) Downloaded 89 times
N34_Industrial_Grade_Motors.pdf
Example of NEMA 34 motor
(270.36 KiB) Downloaded 77 times

User avatar
Rcnewcomb
Vectric Archimage
Posts: 5864
Joined: Fri Nov 04, 2005 5:54 am
Model of CNC Machine: 24x36 GCnC/WinCNC with ATC
Location: San Jose, California, USA
Contact:

Re: Hardware_Control_Project

Post by Rcnewcomb »

What are the advantages of this system over EtherCat, for example?

- Randall Newcomb
10 fingers in, 10 fingers out, another good day in the shop

jsheley
Posts: 19
Joined: Sun Feb 17, 2019 7:27 pm
Model of CNC Machine: expensive_pile_of_junk

Re: Hardware_Control_Project

Post by jsheley »

Rcnewcomb,
I thank you for your constructive question!
The video that you furnished portrays a “token ring” communications protocol similar to BAC/NET used in the HVAC industry. The advantages of what I would like to develop is resides in hardware not the communications protocol. (hardware versus firmware)

The disadvantages of “token ring” Ethernet based motor control versus what I want to develop:
1) Power management.
2) Product commissioning.
3) Expense of printed circuit board parts.
4) Regulatory (security) burdens.

Power management:
The biggest detriment of Ethernet motor control resides in the cable construction. Ethenet conductors are designed to propagate high speed AC signals over long distances. Ether net cables are not designed for large currents needed to propel NEMA34 or NEMA23 motors. The current capabilities per conductor is approximately 300mA. Full load NEMA 34 motors need instantaneous currents in the neighbor hood of 1A or more. Poof goes the Ethernet cable.
Current burdens at the power source of a token ring design are additive.

Please look at image named “simplified_token_power_schematic.PNG”.

For a single point power supply attachment and 3 motors, the current through the power conduit is 3X the individual motor load. The frail Ethernet cable conductors have no hope of furnishing the necessary power to a three motor NEMA34 application.

What I would like to implement is a “Star” type of cable connection with cables designed for high current. Each individual motor has individual a cable to the power supply. Each cable is only responsible for one defined motor load. If the cable is an HVAC cable less than say 10 meters then I anticipate no problems with NEMA 34. 5A USB-A cables should be OK with NEMA 23 motors.

Please look at image named “star_power_schematic.PNG”

Product Commissioning.
“Product Commissioning” is excitement and frustration that happens between the when the product arrives in boxes and when your computer begins to speak to your hardware. Customer frustration can easily kill a multimillion dollar product line.
Ethernet products, in general, are much more frustrating than USB based “serial” devices. Ethernet based products require IP addresses (DHCP or static). You sometimes need to furnish port numbers. The average mill owner has better things to work on than learning the intricacies Ethernet works, so trying to commission an Ethernet device resembles:
1) Install CD.
2) Click “yes” when prompted to assent to legal gibberish.
3) More obedient “yes master” clicks.
4) Restart computer twice.
5) Try to access Ethernet product.
6) Mutter, grumble, pray that health insurance covers stress induced heart attack.
7) Wait 45 minutes on hold for customer support.
8) Repeat steps 1-7
Commissioning an Ethernet on a company network also places you at the mercy of the IT departments whimsical security edicts.

A USB to RS485 is much easier to commission than Ethernet. Some USB to RS485 serial ports are “plug and play”devices. End user only needs to furnish the motor control software “COM” number generated by operating system.
Each additional control node only needs to have a unique 7 bit switch address. A DIP switch sets the address. A DIP switch is a simple set of switched on each control node that you manipulate with a small screw driver.
Please look at image named “DIP_switch.PNG”


When I worked for “S”-electric, men out of high school routinely commissioned RS485 networks at HVAC installations in parking garages.

In compared to Ethernet based products “Plug and Play” USB devices are “Chimp Simple” to commission.


Expensive printed circuit board parts:
A high end RS485 chip and PCB connector cost approximately $4. (>1000 unit quantities)
A reliable Ethernet connector (with transformer) ,and chipset are in the neighbor hood $8.
An additional cost of $4 may not sound like a lot to normal people. Scrofulous curs (project managers) will put a project on hold because someone wants 25 cent part increase on a product that sells retail for $150.

Regulatory (security) burdens:
Another big nasty problem with ether net based products is “security” . When I worked in the HVAC industry, the end customer wanted assurances that our networked products were “hack proof”. How do you design for the unforeseen?
For a simple application such as motor control, a RS485 MODBUS type network is safer than a “token ring” Ethernet network. Most RS485 networks are so primitive that hackers do not even bother with attacks. More fundamentally, RS485 networks generally do not access financial records so there is not much motivation to develop attack methodology.
Attachments
simplified_token_power_schematic.PNG
simplified_token_power_schematic.PNG
star_power_schematic.PNG
star_power_schematic.PNG
DIP_switch.PNG
DIP_switch.PNG

jsheley
Posts: 19
Joined: Sun Feb 17, 2019 7:27 pm
Model of CNC Machine: expensive_pile_of_junk

Re: Hardware_Control_Project

Post by jsheley »

Rcnewcomb,
I thank you for your constructive question!
The video that you furnished portrays a “token ring” communications protocol similar to BAC/NET used in the HVAC industry. The advantages of what I would like to develop is resides in hardware not the communications protocol. (hardware versus firmware)

The disadvantages of “token ring” Ethernet based motor control versus what I want to develop:
1) Power management.
2) Product commissioning.
3) Expense of printed circuit board parts.
4) Regulatory (security) burdens.

Power management:
The biggest detriment of Ethernet motor control resides in the cable construction. Ethenet conductors are designed to propagate high speed AC signals over long distances. Ether net cables are not designed for large currents needed to propel NEMA34 or NEMA23 motors. The current capabilities per conductor is approximately 300mA. Full load NEMA 34 motors need instantaneous currents in the neighbor hood of 1A or more. Poof goes the Ethernet cable.
Current burdens at the power source of a token ring design are additive.

Please look at image named “simplified_token_power_schematic.PNG”.

For a single point power supply attachment and 3 motors, the current through the power conduit is 3X the individual motor load. The frail Ethernet cable conductors have no hope of furnishing the necessary power to a three motor NEMA34 application.

What I would like to implement is a “Star” type of cable connection with cables designed for high current. Each individual motor has individual a cable to the power supply. Each cable is only responsible for one defined motor load. If the cable is an HVAC cable less than say 10 meters then I anticipate no problems with NEMA 34. 5A USB-A cables should be OK with NEMA 23 motors.

Please look at image named “star_power_schematic.PNG”

Product Commissioning.
“Product Commissioning” is excitement and frustration that happens between the when the product arrives in boxes and when your computer begins to speak to your hardware. Customer frustration can easily kill a multimillion dollar product line.
Ethernet products, in general, are much more frustrating than USB based “serial” devices. Ethernet based products require IP addresses (DHCP or static). You sometimes need to furnish port numbers. The average mill owner has better things to work on than learning the intricacies Ethernet works, so trying to commission an Ethernet device resembles:
1) Install CD.
2) Click “yes” when prompted to assent to legal gibberish.
3) More obedient “yes master” clicks.
4) Restart computer twice.
5) Try to access Ethernet product.
6) Mutter, grumble, pray that health insurance covers stress induced heart attack.
7) Wait 45 minutes on hold for customer support.
8) Repeat steps 1-7
Commissioning an Ethernet on a company network also places you at the mercy of the IT departments whimsical security edicts.

A USB to RS485 is much easier to commission than Ethernet. Some USB to RS485 serial ports are “plug and play”devices. End user only needs to furnish the motor control software “COM” number generated by operating system.
Each additional control node only needs to have a unique 7 bit switch address. A DIP switch sets the address. A DIP switch is a simple set of switched on each control node that you manipulate with a small screw driver.
Please look at image named “DIP_switch.PNG”


When I worked for “S”-electric, men out of high school routinely commissioned RS485 networks at HVAC installations in parking garages.

In compared to Ethernet based products “Plug and Play” USB devices are “Chimp Simple” to commission.


Expensive printed circuit board parts:
A high end RS485 chip and PCB connector cost approximately $4. (>1000 unit quantities)
A reliable Ethernet connector (with transformer) ,and chipset are in the neighbor hood $8.
An additional cost of $4 may not sound like a lot to normal people. Scrofulous curs (project managers) will put a project on hold because someone wants 25 cent part increase on a product that sells retail for $150.

Regulatory (security) burdens:
Another big nasty problem with ether net based products is “security” . When I worked in the HVAC industry, the end customer wanted assurances that our networked products were “hack proof”. How do you design for the unforeseen?
For a simple application such as motor control, a RS485 MODBUS type network is safer than a “token ring” Ethernet network. Most RS485 networks are so primitive that hackers do not even bother with attacks. More fundamentally, RS485 networks generally do not access financial records so there is not much motivation to develop attack methodology.

User avatar
Adrian
Vectric Archimage
Posts: 14503
Joined: Thu Nov 23, 2006 2:19 pm
Model of CNC Machine: ShopBot PRS Alpha 96x48
Location: Surrey, UK

Re: Hardware_Control_Project

Post by Adrian »

jsheley wrote:
Wed Dec 30, 2020 11:40 pm
If you know of a forum better than Vectric’s “Hardware Forum”, then please tell me.
I would have thought CNC Zone would be a better place.

jsheley
Posts: 19
Joined: Sun Feb 17, 2019 7:27 pm
Model of CNC Machine: expensive_pile_of_junk

Re: Hardware_Control_Project

Post by jsheley »

" Adrian "

cnczone is an excellent suggestion! I thank you.

Post Reply