[CarveOne]

Ah, I am using the G540. I have an estop hooked up, but long ago took off the home and limit switches. Have a relay for the router which turns it on and off. But no other external stuff. Been considering putting a home on the x and y again just to have a constant base point to start from.

Oh, G20 is inch mode.

I wouldn't want to be in the room if you try the gantry test.

Having at least the X,Y,Z homes would be beneficial, so I'll wire them sooner or later. Also a Z axis touch plate, or something better.

There's plenty of room here to hide behind a work table to one side of the machine. If it takes out a wall and the roof collapses, just try to get under the work table.

CarveOne

[widgitmaster]

Hey Carv'e!
Check out this link, it shows the wire gage and current limits for all AWG cables!
Widgit

[CarveOne]

Hey Carv'e!
Check out this link, it shows the wire gage and current limits for all AWG cables!
Widgit

That's some good info. I haven't run across a list of skin effect frequency value for each wire gauge before. Just a formula.

I usually consult my old ITT "Reference Data for Radio Engineers" or Sams "Reference data for Engineers: Radio, Electronics, Computer, and Communications" books for wire gauge and current carrying capacity charts.

Some of the web sites have some very useful online calculators and formulas. I used to use the UltraCAD Design site a lot for high speed backpanel design work. Circuit board hobbyists can find some useful calculators there. There have been a number of improved calculators added since I last used them.

Thanks,

CarveOne

[R L Harris]

That's some good info. I haven't run across a list of skin effect frequency value for each wire gauge before. Just a formula.

I usually consult my old ITT "Reference Data for Radio Engineers" or Sams "Reference data for Engineers: Radio, Electronics, Computer, and Communications" books for wire gauge and current carrying capacity charts.

Some of the web sites have some very useful online calculators and formulas. I used to use the UltraCAD Design site a lot for high speed backpanel design work. Circuit board hobbyists can find some useful calculators there. There have been a number of improved calculators added since I last used them.

Thanks,

CarveOne

Ok enough with brain halting give me headace electronics charts lingo and calculators and lack of no do it your self
cnc wiring for dummies handbookwith pictures!
Is 22awg ok because thats what i bought to lol plus need picture of wires from breakoutboard to gecko 203V drives!(chair)

[CarveOne]

Ok enough with brain halting give me headace electronics charts lingo and calculators and lack of no do it your self
cnc wiring for dummies handbookwith pictures!
Is 22awg ok because thats what i bought to lol plus need picture of wires from breakoutboard to gecko 203V drives!(chair)

22awg is what most folks use for the step motor wires. It's fine for the motors that we are using. The Gecko FAQ says it's ok for everything but the PSU to Gecko drives power wiring, which needs to be 14awg or 16awg. 14awg is hard to insert in the Gecko terminals, and our motors don't really need it.

Use 16awg PVC automotive hookup wire for the wire that goes from the power supply (PSU) to the G203V drive modules. Twist red and black wires together. Red = + vdc, Black = - vdc. Run a twisted red/black wire from the PSU terminals to one G203V. Do the same for the other G203Vs. DO NOT connect these from G203V to G203V to make the wire runs shorter. (known as daisy chaining, and it doesn't work very well in this application) Route these wires away from any other wires as much as is possible. Be sure that red is positive and black is negative at all connections. The PSU has more than one output terminal for +vdc and for -vdc. They are tied together on the circuit board under the connector.

I hesitate to show my existing wiring scheme for the control box because it may have some wiring that is causing the step-over problem. I wouldn't want you to inherit it. How yours is to be routed will be determined by your layout of the modules and breakout board anyway. I can just say at this point to keep the wires from the BOB to the Gecko drives short and direct as possible while keeping them away from each other and the PSU wiring.

The Step and Direction wires from the BOB to the Gecko drives are the most critical. There are all kinds of opinions on how to run them and what kind of wire/cable to use. I'm sure that I didn't initially do it "right" by most of those opinions. After hooking up the motor cable this weekend and doing a test cut (and seeing no step-over fix) I'm going to rewire my BOB to Gecko drives step and direction signal wiring and see how it works. If it doesn't fix the step-over I'll try another way.

From the Gecko site FAQ:
------------------------

Q: Can I use twisted pair wire for my step and direction signals?
A.) You may use twisted pair wiring, but you can not run step and direction through the same twisted pair. Use one pair with STEP and COMMON, use another pair with DIRECTION and COMMON. Connect both pair's COMMON to the G203V COMMON terminal.

------------------------

I'll try unshielded twisted pair wires first since I don't have shielded pair on hand. Some folks use shielded twisted pair cable and tie the braided shield to the chassis ground only at the BOB (signal source) end of the cable.


CarveOne

[CarveOne]

I made some more progress today. I moved one of the Gecko heatsinks about 1/4" to give more room to drill holes and installed some metal spacers that will allow connecting the motor cable shields to the ground plate. Notches were cut into the oak rear panel for the motor cables to exit through. Now I can finish the motor cable installation tomorrow.

I also painted the new Moore Gear racks with the textured silver paint. The gear teeth are not painted. While I'm at it I may as well paint the rear side racks too before I'm finished with this.

My local CNC friend is looking at the San Jose gcode in detail to see if there are any settings in the gcode that does not match the settings in Mach3, and would cause the kind of weirdness I'm having. He certainly knows more about all this stuff than I do. He'll cut something with it also.

CarveOne

[toughtool]

I would suggest you use your shielded 4 wire cable for the motor coils (2 for "A" and "X", 1 for "Y" and 1 for "Z" and use one (1) stranded CAT5 cable (24 gauge) for the switches. Is there enough room in your cable chain for that many cables?
One CAT5 cable from the driver board to the gantry will give you 4 twisted pairs, each pair with a different twist and impedance, for any switches you are using. You can ground one of each of these pairs if you have induced noise, leaving 4 wires for switches.
Also don't forget you have a parallel cable from the computer to the BOB. May have a bad one. Joe

[Jay C]

My 2 cents, I cut it at half depth in softer wood - no issues
Arbo cut it full depth in foam - no issues

Have you tried to cut it in foam or half depth? It can't hurt right? I enjoy reading your adventures so please take my post as it is intended - just wanting to help.

-Jay

[CarveOne]

I would suggest you use your shielded 4 wire cable for the motor coils (2 for "A" and "X", 1 for "Y" and 1 for "Z" and use one (1) stranded CAT5 cable (24 gauge) for the switches. Is there enough room in your cable chain for that many cables?
One CAT5 cable from the driver board to the gantry will give you 4 twisted pairs, each pair with a different twist and impedance, for any switches you are using. You can ground one of each of these pairs if you have induced noise, leaving 4 wires for switches.
Also don't forget you have a parallel cable from the computer to the BOB. May have a bad one. Joe

I already have added the four shielded cables and the IGUS e-chain is just wide enough for four of them. I zip-tied the router power cable onto the "fingers" at one side of the e-chain.

I can run the CAT5 cable but it has to be zip-tied to the e-chain "fingers" at the other wall of the X axis e-chain.

The Y axis e-chain has room to place it with the shielded motor cables but I'll probably do the same as for the X axis in order to keep the cable farther away from the motor cables, which are running on 72vdc.

So far, the e-chain is flexing acceptably if I zip-tie it to one link and skip two links. Adding the CAT5 cable may change that a little. I may need to open up the bend radius to fix that. It's doable though.

Thanks, I'll use CAT5 when I wire the switches. I'm considering using romanlini's design instead of the mechanical switches I have.

CarveOne

[CarveOne]

My 2 cents, I cut it at half depth in softer wood - no issues
Arbo cut it full depth in foam - no issues

Have you tried to cut it in foam or half depth? It can't hurt right? I enjoy reading your adventures so please take my post as it is intended - just wanting to help.

-Jay

No problem Jay. I plan to do just that once it is back running again. I could do it on my smaller machine though. I don't have any blue or pink foam on hand. Maybe I can find something else though. On the big machine I wonder if I am using an accel number that is too low. It has a 175 lb gantry and I used 30 accel. It probably needs to go higher. Maybe it isn't accelerating fast enough during jogs.

You must mean mis-adventures.

CarveOne

[CarveOne]

I decided to go ahead and rewire my step and dir signal wires per the Gecko FAQ, and rearrange the connections while I was at it. Once I hook up the shielded motor cables it becomes much harder to move the control box and to rewire these interconnections. The photos show the before (yellow, green, brown, orange, black wires) and after (red and blue wires). Now no signal wires cross other signal wires and each signal wire has its own common wire. They are all separated from each other as well. They can't get any shorter either. The four heat sink grooves just happened to be convenient for routing the wire pairs and keeping them mostly surrounded by ground potential since the heatsinks are attached to the ground plate with stainless steel screws.

The blue wires are all common, the red wires are either step or dir. I only have those two colors. The order I used from top to bottom of the BOB connector is step then dir. Both common wires go into a single common terminal. I used common door bell single conductor wire.

The five hex spacers are where I will ground the motor cable shields. One spacer connection for each cable shield, or I can tie them all just to the center spacer. Since my new wires are smaller, and lay closer to the chassis, I can replace the 1" hex spacers with some shorter ones I have.

CarveOne

[FandZ]

I gotta say, the before and after shots of that wiring looks great. Hopefully it's all in the wiring.

[CarveOne]

I gotta say, the before and after shots of that wiring looks great. Hopefully it's all in the wiring.

I've learned in my years in electronics that it's not safe to say "This is going to fix the problem for sure.". Especially if your next paycheck depends on it.

The door bell wire is stiff enough that I can lift it out of the heat sink grooves and it will stay suspended in air if the signals don't like being in the grooves due to capacitance or something. Let's hope it works. There are no vibrations going on that will shake it enough to wear on the insulation. If it becomes a problem I'll just put some clear silicone rubber sealer in the grooves and push the wires into it.

Moving the control box around will put twisting motions on the shielded motor cables, and these cables don't like to twist. I put temporary notches in the rear panel, but will see if I can find some of the cable grippers for this size cable tomorrow at a local electrical supply house before I do the final hookups.

CarveOne

[toughtool]

Carverone,
That is a very nice layout. I don't see any thing that would cause a problem with either wiring versions. I am surprised though to see you did not put it all in a metal box. Shielding the electronics and wiring from the welder and other machines in the room.

Is it possible for you to run a different test than the San Jose test and post a photo? Your first cut and final cleanup cut are under very different loads. Maybe two concentric circles, something like 2-1/2" and 3" in diameter, one at the original depth and one very shallow cut? This could show any mechanical flex problem between two different loads. May be helpful figuring out what is actually happening, using very simple G-code. Joe

[CarveOne]

Carverone,
That is a very nice layout. I don't see any thing that would cause a problem with either wiring versions. I am surprised though to see you did not put it all in a metal box. Shielding the electronics and wiring from the welder and other machines in the room.

Is it possible for you to run a different test than the San Jose test and post a photo? Your first cut and final cleanup cut are under very different loads. Maybe two concentric circles, something like 2-1/2" and 3" in diameter, one at the original depth and one very shallow cut? This could show any mechanical flex problem between two different loads. May be helpful figuring out what is actually happening, using very simple G-code. Joe

The oak box is another one of my early "cheap out" attempts. The welder is rarely used, and is never used when the CNC computers are on or when cutting anything. I'm also out in the country well away from high power transmitters. Next trip into town I'll get a sheet of blue foam and some more MDF. By then I'll have the control box back in operation again. The only other things is a small AM/FM/CD boom box and the electronic ballast overhead 8' lights. They noise up the FM radio. Maybe I should cut something with the lights off too.

Eventually, or sooner, I'll transfer it all to a commercial metal enclosure.

Concentric circles at two depths is a good idea.

The step-over difference happens before I even see the cutter go down to the wrong starting point for the cleanup pass. If there is a loading issue it has to be in the first pocket cuts.

CarveOne

[CarveOne]

I found some electrical cable clamps in my junk repositories and removed the nuts. The box was bored to the depths of the threads. The counter bores were too loose to thread them into the oak. Before installing them I drilled slightly oversize holes for the 3/8" diameter cables. I then epoxied them in place with Locktite's version of JB Kwik.

The cables should be constrained well enough with these instead of the plastic cable grippers. I'll only tighten the clamp plates enough the cables don't slip in and out easily.

This saved me a trip into town this morning. Now I can feed the cables through after stripping the ends beforehand. These are not ideal, but I have some pieces of 1/16" thick neoprene rubber sheet that can be cut and placed between the cable and the clamp plate.

CarveOne

[CarveOne]

I went to Lowe's late this afternoon and picked up two sheets of MDF and a sheet of 1/2" thick blue foam to do test cuts with.

I then completed the final connections to the computer and fired up Mach3. I bumped the feed rate to 200ipm and set the accel to 50 for X and Y. I set the Z for 100ipm and accel to 50. Then I powered up the BOB and Geckos.

Every motor came to life with no issues. I jogged each axis to make sure the motors go in the correct directions. A previously cut piece of MDF was placed on the table and the zeros were set. The router bit was a little below the top surface, but was close enough for me.

Turned on the router and dust collector and started the cut. I watched it do the cuts with no peeking under the dust collector. It either works correctly or it doesn't. When it completed the text and raised for Tool2 change I stopped it and moved the gantry out of the way. A little sanding off the fuzz and IT WORKS!! No step-over this time!

Looks like ole CarvelessOne finally found the cheese. I hope so anyway. Not exactly sure where the cheese was but I'll take what I can get if it fixes this problem. More cutting will tell for sure.

The actual depth of cut in the text pockets measures 0.226" due to less than perfect zero reference. It was supposed to be 0.150". The router and cutter handled it just fine.

CarveOne

[Arbo]

Nice and congratulations!

It's good when things work out, and yeah, the cheese not so important... once the issue is gone.

[R L Harris]

Cool can't wait too see some of your projects
now that your up going with out step over problem!

[CarveOne]

Nice and congratulations!

It's good when things work out, and yeah, the cheese not so important... once the issue is gone.

Thanks Arbo. One cut doesn't really prove it but let's hope it has been cured.

CarveOne