• This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn more.
  • PLEASE: Read the FORUM RULES BEFORE registering!

[Newbie] First Cnc Build : Hd Router/mill

Iceman

Active Member
Active Member
#1
I have never built a CNC machine, and it has been years (10+) since I have played around with one. I'm 30, and I am a Flight Engineer / Pilot (not an EE, I don't do industrial automation for work, etc.).

I grew up around airplanes, and I may have a few engineering courses under my belt. The machine is in the drafting phase, and the electronics (the heart) are starting to come together.

System voltage input: 240VAC split phase
Control/Safety Voltage: 120VAC, 24VDC
Motion Control Voltage: 5VDC
Stepper Voltage: 65VDC

Motion Control: Hicon Integra 6-axis motion controller
Stepper Drivers x4: Kelling KL8056
Currently planning Nema23 425oz-in 8 wire Steppers (on hand) - not 100% sure this will be sufficient for all axis.

Safety: Dold Safety Relay (controls main contactor)
-E-stop
-Start/Stop switches

The Teaser: 14012885_10153648073736237_176500348_o.jpg

(Shop Approved footwear?)
 
Last edited:

Iceman

Active Member
Active Member
#4
Thanks guys! I've started to partially power and test subsystems - I have yet to power that massive toroid up. I Finally have my comically overkill 100watt 200ohm resister to bleed the capacitors on shut down. I've also decided to add a suppression diode on the contractor coil. I will shortly be posting better pictures, and my schematic that I'm currently working off of (and on - it's very much a living hand drafted document). I will do my best to document this so that it may help others who decide to go this route. There are a few beautiful looking Home built CNC control panels out there, and the details and schematics are scarce, and often lacking in explanation. With everything going on - this project will be 2 years from start to finish. Hopefully I'll be able to capture that. Most likely there will be a youtube video with source documents at the end.

The disconnect I used was for a 3 phase system, so there is an unused leg. I've got 25A fuses in the disconnect which is being fed from a 30A breaker in the panel dedicated for this (10-4 wire). From the disconnect the power goes to a siemens isolator switch that will be mounted on the front of the panel - I could have probably located the parts to remote the main disconnect but that thing is a pain to actuate (probably a good thing) and the isolator switch feels like it was designed for a lifetime of continual abuse in hazardous locations.

After turning on the isolator switch, the next switch which will need to be actuated is the DC Safety and Control switch - a 120VAC switch that provides single phase AC through two MCB's properly sized for two switch mode power supplies - a 5VDC and a 24VDC. The 5VDC is the main power for the Hicon Integra, and the 24VDC is what I chose for control voltage as well as the voltage that will be used for sensors and other such things the system will need.

Controversial topic: the "bonding" of the neg(-) outputs of two different switch mode power supplies. At this point I have decided to the leave the outputs of both power supplies floating (not tied to ground), but tied together so that I only need a single Neg bus (really a 0VDC bus). It is too early to know if this is really an issue. I see for many applications that the negative output of a switch mode PSU is bonded to ground in control applications but not instrumentation (I think due to noise / ground loops).

Like I said - I've taken many classes, but not a PE/EE so, I know just enough to be safely dangerous. Electronics and electrical systems and engineering/machining has been a life long hobby, so sometimes I learn on the spot.

-IM
 

JimDawson

Global Moderator
Staff member
Director
#5
I Finally have my comically overkill 100watt 200ohm resister to bleed the capacitors on shut down.
A 10K, 2 Watt would be more than adequate. ;) But yours will work fine. If you look on the board, there are two bleeder resistors installed. Lower right corner of the board in the picture.

I always bond the - of power supplies to ground. Bring all of the grounds and shields back to a single point. Normally to a ground bus on the main panel. Only connect the shields on one end.

Nice looking panel. :encourage: Looking forward to watching your build.
 

Iceman

Active Member
Active Member
#6
Jim,

There are indeed resisters on output of the 65VDC power supply, my lack of experience had me feeling like these were too small - although I do have a tendency to overcompensate for my lack of practical experience by applying excessive overkill. The golden anodized heatsinks on the 100watt wire wound resisters were too good looking to pass up.

I'm curious to see if any noise is added to the power (and if it even really matters) in this case if I tie all the - to ground (see what I did there?). For some reason there is much controversy over which way is correct (or if indeed it is purely situationally dependent - is there a code that spells out which is correct? I will mess around with the O scope, I already want to witness first hand (because theory isn't enough, trust but verify) the suppression diode reducing the voltage spike from the collapsing magnetic field when the contactor is turned off. Come to think of it - that massive toroid is a fairly large inductor - do people do something similar with suppression/flywheel diodes on power supplies like this to reduce the voltage spike created (on the supply side) when powering down?

-IM
 

JimDawson

Global Moderator
Staff member
Director
#7
The golden anodized heatsinks on the 100watt wire wound resisters were too good looking to pass up.
They are purdy ;)

When mixing power and control signals, I have found it better and less chance of noise if everything is referenced to ground. There is no code that I'm aware of.

It is common to put suppression diodes across relay coils, especially when switching them with transistor logic circuits. It would be uncommon to put a surge suppressor across the primary of the toroid, but you could put a MOV across it. https://en.wikipedia.org/wiki/Surge_protector#Metal_oxide_varistor_.28MOV.29
 

Boswell

Hobby Machinist since 2010
H-M Supporter-Premium
#10
Iceman, you look like you are making good progress since your first set of pictures. I'm a sucker for instrumentation so I like seeing the panel meters. Are these going to measure current to the steppers of each axis?
 

bpratl

Active User
H-M Supporter-Premium
#11
Iceman, that's a very nice layout, looking good. What is the LxWxH of your Cabinet/panel and the source or did you fabricate it.
Bob
 

Iceman

Active Member
Active Member
#12
The enclosure is 24x20x8 made by Yuco (?) found on Amazon. It isn't the highest quality, but it didn't break bank - came a horrible beige color, nothing two cans of protection enamel spray paint couldn't fix. Accidentally signed the front panel with a thumb print this morning as a small section is still curing.

The meters you see (I am also a sucker for instrumentation) are Volt meters, to measure the split phase and phase legs on the input, and the the 24VDC control voltage. I have though about measuring the current each driver is drawing with ammeters, I bought a Shurite Edgewise series meter to scope it out, and it requires a rectangular mounting hole (what a pain). Any good tips - wish there were rectangular knockout tools like the round ones I used on the holes. I also have fuse holders for the stepper drivers, not exactly sure on where the stepper driver ammeters and fuse holders are going to go.

The panel is a good size, and I know it looks like it was planned, but I feel like it is still too small. I've seen it written before, find the size you think you need, and add 4-6 inches on both sides - you'll almost never curse it being too big, but slightly small creates headaches. I think this enclosure will just barely work - I may need to come off with conduit to a small secondary box for stepper driver instrumentation, and eventually with a third box for the VFD/Spindle controls.

Slowly making progress, it will come in waves as I suspect these things often do.

-IM
 

Boswell

Hobby Machinist since 2010
H-M Supporter-Premium
#13
I agree with the logic of larger is better than smaller in most cases. for your rectangular holes and depending on the gauge of the sheet metal you could use a nibbler after drilling as much as you can and then clean up with a file. You can find Nibblers under $50 that say they are good up to 18 Gauge steel. If you use the nibbler be sure to put some tape down to protect your paint finish from the tool. Also if you do add current meters, be sure to put on on the Spindle motor. When I was using a smaller mill I used one and it was a great way to keep me from over loading the motor.
 

Iceman

Active Member
Active Member
#14
I've decided to hold what I have - this is after all version Mk 1. Mod 0. The ammeters would be great, but I don't want to clutter the panel. The next Version, or perhaps down the road I can implement some type of current transformer/ pickup system and touch screen user interface that monitors the system in conjunction with what is already there. Always looking to improve the human/machine interface.

I have the holes drilled for the cannon plugs, and I need to dill/punch the last hole for the main power feed before I can reassemble the cabinet and drop in the bulk of the electronics. I can then start to mount the front panel components, and continue on with the wiring and testing. I am including two 14 pin cannon plugs along with the four 4 pin stepper cannon plugs to accommodate I/O needs, position sensors, limit switches, etc. I also need to figure out how I'm going to get a DB15 pin cable in to the panel for the MPG I have that works with the Hicon. Looking for a panel mount for the DB15 hasn't yielding much. I also need to panel mount the ethernet - that or go for a cable gland system. Le sigh.

I should be so fortunate that I am able to frustrate myself with entertaining challenges on the side.

-IM
 

Iceman

Active Member
Active Member
#15
IMG_2114.jpg

Disclaimer: The holes are not perfect : especially the cannon plug bulkhead screw holes - I feel like it's going to be a rat tail file kind of assembly (I.E. "Assembly went smooth, but it was somewhat of an organic process").
 
Last edited:

Iceman

Active Member
Active Member
#17
Ask and ye shall receive - alas, I need a old school DB15, not the db-HD15 - Picky, I know. I had read some bad reviews on Allied (after my 5th- or 6th order) - but have had no issues with them, and they seem to have a good stock selection.

Thank you for taking the time to help source parts. The cat 6 port looks perfect for my needs. Phoenix Contact also makes some interesting panel mount ports (industrial ethernet - way overkill)

-IM
 
Last edited:

Iceman

Active Member
Active Member
#19
Now thats a thing of beauty. Wish it had cat5/6 too - it would be perfect. The more I research, the more interested in the M12/M8 cable standards for sensors and such - but that will have to be a V2 thing.

I think that second find is going to be the win.

Thanks for that find.

Tomorrow I'll hopefully get to a point when the electronics/electrical drop into the main enclosure. It will make some of the wiring harder, but enable the final wiring at the same time - the old chicken and the egg dilemma.

-IM
 

Iceman

Active Member
Active Member
#20
The main switches other than the E-Stop (which was made by Fuji) are Mfg. by IDEC. I have limited experience, but the first few switches I bought were so cheap (overseas poorly made IE Uxcel - stay away) -( but hard to pass up) - ended up feeling terrible. Poor actuation feel, just did not feel proper for such a project (or maybe even safe). The IDEC switches are superb. They feel like they will outlive this project.
 

Iceman

Active Member
Active Member
#24
Ok - Here is V8.something of the working schematic.

Here is my "theory" of operation. Remember my middle name is overkill...and danger....hmmm, danger overkill has a nice ring to it...

Power from my main US distribution panel, two pole split phase 30 amp breaker, 10-4 wire out to a 4 prong NEMA 35 amp rated outlet.
From that outlet is essentially a generator extension cord also rated at 35 amps. From there the machine plugs in with a 25 foot service cord 10-4 again, heavy duty designed to survive an apocalypse.

Power will enter the panel via a metal wire strain relief, and connect to the first set of terminal blocks. Although since I removed the fuse monitoring device to make room for 4 phoenix contact high voltage DC breakers to take the place of the original fuse idea for the stepper drivers, this connection may be made directly to the 3 phase fused disconnect. 25 amp class CC fuses are in the disconnect on two legs, one leg is unused. From there the hot legs will go to the panel isolator on the front panel, and back to independent terminal blocks creating power "rails". The neutral is switched at the initial disconnect but from there will go directly to a neutral bus. Ground and Neutral are tied together in the service panel, but kept isolated in the machine.

So essentially we have very well protected power in the box at this point, split phase 240/ two 120 "legs". One of the 120 legs will be used to power the control and safety systems which essentially means to power two switch mode power supplies. This is accomplished by routing power through a "control and safety master" switch which provides power to a separate 120v bus. From this bus power is routed to circuit breakers ("C" trip curves, I chose this based on considering SM PSU's as relatively inductive) one each for the respective power supplies. From the circuit breakers the power is routed directly to the SM PSU's. The 5VDC power goes directly to the Hicon Integra motion controller (hence the control aspect of the switch name). The 24VDC goes to a 24VDC bus located in a somewhat more strategic wiring location. The "negative" or zero outputs of both DC SMPSU's are tied together to remove any difference in potential between these two systems. Not done yet but likely to happen will be connecting this negative bus to ground to create a zero reference among all of the equipment in the cabinet (ground loop issues will be looked at if they pop up).

The rest of the wiring in this schematic is to work with my Dold Safety Relay, creating a E-stop circuit, and manual start/stop operation. This will likely vary depending on the design of the individuals safety system/relays used. But ultimately the goal is to control a Siemens contactor relay which will power the 65VDC 20A power supply which will feed power to the steppers. There is /now/ a suppression diode on the contactor relay to reduce the voltage spike from de-energizing the contactor relay.

Not drawn yet, but the power will go from the 65VDC PSU to 4 Circuit Breakers then on to the stepper drivers. There is also two 24vdc legs that go to the hicon to power sensors, not yet drawn.

The power for the 65VDC 20 Amp PSU is routed first through double pole 20A MCB then through the contactor, then to the PSU. Two of the coil wires are connected together to operate in series per 240V operating guidelines on this particular PSU. I know typically the MCB current rating is calculated at something like 125% of normal operating current, but I don't see this thing drawing that much current with the Nema 23 steppers I have....currently... (it could be a learning moment).


It's a start - and very much a living process.

-IM
 

Iceman

Active Member
Active Member
#25
Quick shop update: The front panel controls and indicators are wired up (finally). Life gets busy.

I would like to believe one of the reasons we decide to tackle projects of this magnitude is because it affords us with the opportunity to solve unforeseen problems.
-
tumblr_mso9cmrWmI1qz5q5lo1_500.gif
(points for the movie reference)

-Discovered that the massive toroid (with 2 10,000uF capacitors) trips the 25A MCB (b curve) I have wired in for it. Planning on implementing an inrush current limiter circuit which includes the following:

Negative temperature coefficient resister (NTC)
Time Delay Relay

Will post back with results.

-IM
 

Iceman

Active Member
Active Member
#26
Every time I think about a time delay relay it reminds me of a specific aircraft write up-

"No Ice" deice time delay relay inoperative.

-IM
 

Iceman

Active Member
Active Member
#27
Parts arrive tomorrow, though I'll be flying, so maybe Tuesday night will be my first successful power up.

we will see.

-IM
 

brino

Active User
H-M Supporter-Premium
#28
the first few switches I bought were so cheap (overseas poorly made IE Uxcel - stay away) -( but hard to pass up) - ended up feeling terrible.
I tried some cheap ebay switches once.....and quickly found out that you cannot even solder to the leads without the entire terminal melting out of the plastic housing. Yeah that saved time and money! Doh! :mad:

-brino
 

Iceman

Active Member
Active Member
#29
Still Alive. Changing the MCB for the main DC TR unit from a 25A B-Curve to a 30A D-Curve. I have a 30A main breaker as well, it might be pushing things a bit - but after all the breakers are there to protect the wiring - so I feel that I maintained a high level of safety. This "Should" or solve the issue. I scored an awesome Siemens time delay relay for 13 bucks (normally way more expensive) - so in the parts bin for the next project.
 

Iceman

Active Member
Active Member
#30
Ok - had one of "those" moments.

Swapped in the 30A D Curve and blew a couple sets of slow blow fuses on the fused disconnect (ops check good!). this thing is acting way too much like a direct short....I did ohm out the transformer leads before I hooked it up...right? I couldn't believe it - this thing has two primary coils, and I ignored that VERY important fact and while I had a 50/50 chance ( we don't leave things to chance do we?) I had it wrong. So one of the coils was energized/shorted while the other was shorted/closed by itself. I reinstalled the 25A B-Curve, and corrected the wiring SNAFU - all systems go. As it was originally intended.

It was a good exercise in futility because no amount of inrush current mitigation would solve that. With the large wattage resister the breaker did not blow because there was now a large hot....oh, never mind.

the problem is solved! (no matter it was one I made for myself).

-IM