I want to build a 6 axis

[DOMDEC]

I got the opportunity to buy a industrial Robot.
This machine used to assemble portable telephones.
It was build in october 2000 and has 2 X_Y_Z configurations.
These are driven by YASKAWA 400 W AC servo motors and come with controllers. These can be pulse driven.
The X-Y-Z system is NSK ROBOT MODULE.
There are also 2 steppers (VEXTA ASM66AC) to control the C axises
I am already busy to dismantle 1 of the 2 systems, I will keep these for spare parts.

The only problem was the thing is a bit to big for my workshop.
Had to take it as it was however.
On the other hand it is beautiful made and comes complete with safety doors and contacts, way to much sensors, pneumatical systems, explosion safety lights and so on.
Fist goal is to try to make a functional 3 axis CNC router with Mach 2 software.
However the final goal must be 6 axis with automatic tool change.

Once finished i will use the CNC for making molds and parts for my classic bike (AJS 1955) and car (sunbeam alpine IV 1967)

Still a (very) long way to go.

Best regards, Dominique

[DOMDEC]

In the past days I started the project.
Made a small interface board (not PCB but rather messy wiring experimental board, witch is to be poured over with poleyster resin for strenght) with some 74LS244 to protect the computer . This will split the 12 outputs from the 5 inputs from LPT 1.
Hooking them up to the YASKAWA servo drive was not so easy because there are a lot of wires coming from the drive (connection CN1) and as I did not have a proper break out box for them I have to measure out wire after wire.
Finally made it working. It is now possible to let the motor drive left and right (Jogging mode) with the keyboard.
In the YASKAWA drive the inputs are opto isolated so I don't think it necessary to protect it further more.
The outputs for cooling and other stuff will be trough an extra board with opto-couplers.
Next thing to to is to build the control cable a bit more proffesional as this was just a experimental set up.
However the following question remain.
As with 2 LPT ports I will only have 10 inputs. 6 axis with reference and limit switches (limits + and - together) will already take up 6 x 2 = 12 inputs, how can I do this ?
Can I also use the Z-puls from the encoders to reference very accurate ???
The motor from the Z-axis comes with a brake, so I had to connect an extra 24V to hold it open. Schould I leave it like this or can I use the brake for stronger holding force, and if so schould I use an extra output from the LPT ports to control it ?

[ger21]

Originally Posted by DOMDEC However the following question remain. As with 2 LPT ports I will only have 10 inputs. 6 axis with reference and limit switches (limits + and - together) will already take up 6 x 2 = 12 inputs, how can I do this ? Can I also use the Z-puls from the encoders to reference very accurate ??? The motor from the Z-axis comes with a brake, so I had to connect an extra 24V to hold it open. Schould I leave it like this or can I use the brake for stronger holding force, and if so schould I use an extra output from the LPT ports to control it ?

Check out http://www.newfangledsolutions.com You can use modbus to add inputs to Mach3. Get in contact with Brian at newfangled for more info, or check out the Mach2/3 Yahoo group.

[2muchstuff]

Are you sure it won't fit in the shop. It's only the size of you Sunbeam.

Just move the shop into the garage. It looks like it's already set up for fire suppression. Are those fire sprinkler heads I see along the ceiling.

BTW... Very nice find, lots of nice looking parts and pieces.

[DOMDEC]

ELectronics are a real pain, I spend today wit the next step of the project, hooking up the motor to the lineair slide that came with it.
Everything kept on working fine, and I started with calculating the pulses to revolution ratio.
It seemed all was well, however after writing a small program to test everything i noticed a accumulating error.
So I decided to put a gauge to the slide to check it out and jogged 1 mm at a time.
In the direction of Z- everything was perfect, in Z+ there was only a movement of 0.97 mm. Strange.
First I suspected Mach 2 (still demo version) to foul me.
Then I downloaded OMRON WMON software for monitoring what was going on.
The drive was receiving the same amount of pulses from the electronics in Z+ compared with Z- .
I went to see the gauge again and now it worked perfect ???
Took the serial connection out and the problem was back ??
To make a long story short, I needed to ground the drive to my computer (you can see a small yellow wire on the photo).
Strange is that the pulses were only lost in one direction ??

By the way, those tubes in the garage are no sprinklers but air pressure distribution, I still have to paint it.

[DOMDEC]

Since last time I have been working hard to get around some problems.
As I had written a small program I noticed that MACH 2 got a RESET about once in 15 miutes. Figured out this had nothing to do with Windows XP but with spikes getting in trough the parallel printer port.
So I decided to rebuild the interface with opto couplers.
It had been over 15 years that I made some PC board but I managed.
Far from clean (the design has been altered several times by trail and error) but it works and no more unwanted RESET buttons coming up.
In meanwhile prepared everything for the other axis coming up soon.
As I have been collecting all sort of stuff over the last years I got a smile on my face of finally using these things that were laying around.
The big PC board is an industrial power supply 220 V to 12 V (+/-) 5 V and 24 V. The connecting points comes from printer switch boxes that everybody seems to throw away these days.
Also managed to hook up the inductive limit switches that came with the robot. These inputs are fed in the YASKAWA drives directly.
By using the Z pulse from the encoders i can also very accuratly reference the axis in Mach 2.
Next step will be to place everything in the machine and to move 3 axis.
However next week i will be on holliday in the French alpes so work will have a stand still.

[DOMDEC]

Glad I can continu with the project. Placed everything in the machine, but did not find a good place for the computer, so this one will sit in the working area for now.
Took al lot of time to wire everything up, and some connections are still temporary.
As I wanted to test cutting something I needed a temporary spindle, in the future this should be on the B and C axis but for now it is only a 3 axis.
To speed things up an ordinary drill will do for now.
The first experiment was to cut a square and a circle (2mm deep) in a piece of oak.
It ended with the mill drilling deep into the oak at a certain point.
It seems that the Z dropped down, ???, but there was no fault of the servo drives, nor from the computer, hmm strange !
Must admit that this temporary construction causes way to much vibation so I shall have to solve that first of all.

[DOMDEC]

I couldn't leave it alone for a day, just had to get a positive result.
To solve some of the vibration in this test fase I just pushed a piece of plastic between the drill and the back plate and tensioned it up with a strap.
This worked beter. After redoing the oak piece I made a smaller piece in a sort of plastic that was laying around, worked out very fine and then decided to go for aluminum (is a must for me ).
Things went very well exept for the end of the piece were the mill diggend in some 2 mm deeper then Programmed.
Hmm, it seems that because of the strong vibrations the Z-axis servo can not hold position and drops down, no holding power anymore.
I was planning of moving the Z-axis higher up but then I have to remake a piece to hold it against the upper slide.
Next thing to do first of all however is wire the emergency stop and the drive enable's to the front of the machine so I can stop the machine if something goes wrong. As the Robot came with some buttons already mounted this schould not be a problem.
The bottom plate of the machine is now only a 3mm plate witch is only supported at the sides, this acts as a trampoline making vibration even worse.

[acondit]

Endmills and drill chucks are generally a bad combination. The chuck teeth generally can't generate sufficient holding force to keep the endmill from moving. This can result in the endmill being pulled down into the material being cut (hence deeper cut than desired).

Peace,
Alan

[DOMDEC]

I know, but this was not the case, the mill was still hold on in the same position.
It was the z-axis losing all power to hold the axis.
The last time I gave (very short after the mill got stuc k) the command to jog up, nothing happend , then all at a sudden the Z-axis shoot up like a rocket.
It seems that the servo drive goes into a sort of overload, but still counting pulses, however the red light on the front does not come on ?
Schould read the error code signal from the drive, unfortunatly did not wired it trough the connectors.
The drill chuck is only temporary, as is the vice on the table, but I have to start somewere.

[DOMDEC]

I wanted to make the A-axis strong enough so vibration would no be a problem.
I had some steel plate ( 8 mm ) laying around so that would do.
Welded a strong support were the motor-reduction could be bolted onto.
Motor is a 1.5 KW YASKAWA servo connected to a 1:18 reduction.
As this reduction has some very heavy bearing in it that would just be fine.
As for now I leave it as it is but the axis still has to be centered by means of 3 setscrews.
Main setback is that this drive is needing 3 x 220 V so I had to arrange for that also,having a transformer 3 x 380 V to 3 x 220 V for my lathe all I had to do was making an extra outlet.
Next thing to do will to make the adapter piece for a 4 way jaw chuck.
Before I could mount everything in the machine I had to make the base plate as well.
Made a base construction of some 25 x 25 mm steel rods. On top of this I mounted a white plastic-hardbord plate (10 mm).
By now the 4 axis is connected to Mach 2 at 100 pulses for 1° so that should be OK.

[DOMDEC]

After the first test i decided to go for a z-axis with more travel and a bit more rigidity. It is not that I want to make real big pieces but if you want to mill something to a existing piece that is already like 250 mm high then you need te travel.
So I scrapped another Robot with some nice slides.These came with an OMRON AC servo motor 400 W and uses the same YASKAWA controller as the previous axis so that is nice.
Toke some milling and drilling to make it fix but eventualy it worked.
Next step was to replace the drill with a small router. I managed to get a old one that is still in very good shape and most important that was easy to mount to the z-axis.(600 W 45000 rpm collet 6mm)
A wooden (double plate) was quickly made to give it a try.
(The wooden multiplex plate is only temporary to check everything out)
However , there is just to much flex in the plate, with one hand you can pull the plate almost 2 mm out of line.
Initial testing milling MDF went very well, milling Aluminium was possible but vibration set in.
So next step will be to replace the wooden plate by some Aluminium.