Making Skins #74 - Brisbane Australia

[KenC]

Nice foot, do remember to remove them before you turn the table over.

[Surfcnc]

I never cease to surprise myself what I can forget. Searched for 30 mins today for my car keys as a recent example.

Working by myself, it is a case of nutting out all the moves. It takes a lot more work but I typically disassemble the base to move it even at this early stage.

Regards
Ross

[riesvantwisk]

Ross,

thanks for the info on the gauge, I have seen them here aswell for sale and I might get my hands on one of these one day. Anything I should look out for while buying?

Originally Posted by Surfcnc

Hey I liked your T slot base on the table. I have been looking at how I might doing something the same and along comes your photo. Did you use a T slot cutter with the boards in place or some other technique?

I will tell you my little secret
First what I did was surfacing the base board, needed to be careful because I had some screws in there so I used cheap bit to do the work (bit is still alive). I only was off 1mm on the right so it wasn't a biggy.
Second I did lay my spoilboard on my base board and wrote G-Code to cut the 20mm channels with a 6mm endmil, they are 8x20mm.
Then I would put glue on my spoilboard and glued it on my base board and let that set overnight. Next day we simply made the 6mm top channel so now I have a T-Slot

If you want I can send the g-code to make the 8x20 channels, I forgot to save the top channel, but that is just a single line

Ries

[Surfcnc]

Hi Ries

I looked very closely for marks to discover your technique. What can I say other than you now have a CNC machine so your solution involves a CNC machine, haha. An elegant solution and thanks for letting me (and the rest of the web) in on your secret.

The only caution about the little angle gauge is don't press the buttons accidentally while putting it away into it little pouch that it comes in. I have used up three batteries but they only cost about $2 each. By now I'm sure that you see it shares the precision innards of a vernier and boy don't those magnets cling on.

Cheers
Ross

[KenC]

Originally Posted by Surfcnc

Working by myself, it is a case of nutting out all the moves. It takes a lot more work .....
Regards
Ross

The experience (frustration) is mutual
At times, moving a simple object seems like a mega project...
BUT the reward is even greater when all went well.

[Surfcnc]

I mixed up my ladder panels a bit and had to re-establish their correct orientation - the lesson here is you just can't mark things well enough.

Although not much looks like it has been done, the x beams were rolled over in readiness for the final fit up and squared / leveled up again. I re-assembled the ladder frame and then checked my CAD drawings to see if it was going to land on the attachment plates. Placed the frame on the X beams and the attachment plates and was just 2mm out on the width on one end.

After a bit more measuring I realised that the 2mm was because the x beams were 2mm out of parallel. Fixed that error and the pics below tell the story of the "Perfect Landing", haha.

The moment the attachment plates were welded (and still some more to do there) the whole structure locked up solidly.

We have had some wet weather here so I washed down the Y beams with degreaser and a nylon scouring pad in the rain. I hand painted the Y beams in the shed and regret not spraying them on a nice fine day as the hundreds of holes in the beams needed constant vigilance to stop the paint running. Slow and steady won the race in the end and the job turned out well.

My next job - disassemble the base and then cut the angles on the end of the X Beams - Hope all that stuff about thin cutting disks on the forum is the go because I'm planning a freehand attack on them.

Regards
Ross

[Besser]

Just to let you know about the USB extensions and getting a hub near the y-car. http://www.data-alliance.net/-strse-...ter/Detail.bok
An active repeater cable is probably the go as a cat 5 solution is to costly. Running a double extension gives 10m plus the length of any hub lead.

Like the progress!

[Surfcnc]

Cheers - Ebay has a 5m cable out of Singapore for $8.99 incl postage (cheaper than a quality shielded cable here in Australia). I have some 3G wireless modem installations and a camera installation I can test it on to establish how they go.
Ross

[Surfcnc]

Rule number one is apparently - Stop and Think.

Pluged in my welder to fix the newly made plates onto the X beams and heard a tiny tick... then nothing. Checked the circuit breakers on the machine and the house fuse box and gave them a flick on and off ...still nothing.

Dropped the welder off for repair a few suburbs away only to be told I needed to bring the leads etc. Unperturbed I turned around and repeated the journey this time with all the leads.

Got out my "toy" welder to just tack it up then realised that either both my welders and now even my grinder had all failed or something else may ibe the cause of all is inactivity !!
In fact the newly installed 15Amp circuit had given up the ghost. Called the sparky to fix up his work and gave up for the day.

Note to self - Check EVERYTHING before doing ANYTHING, hehe.

Attached is a photo of the plates and one of the X Beams with a nice 45 degree angle cut on it.

[Claudiu]

Ross, what are these plates for?

[Surfcnc]

Hi Claus

The plates cover the four cut ends of the two X beams. Mentally slide one of the plates forward in the photo and that is where it fits.

Regards
Ross

[Claudiu]

Ok, I was just too blind to make the connection and see it from the photos.
Good luck with your build...

[AuS MaDDoG]

Looking good Ross!! I guess the thin cutting discs done the job then
Did you bend the plates up, or did you have someone bend them for you?

Cheers
Tony

[Surfcnc]

To bend the plates I clamped a guide across the width of the plate and then cut most of the way through with a normal thickness cutting disk. The cut is then V'd out a bit more on the edges so it can close cleanly to the 45 degrees. The plates were then put in a bench vice and "gently" hit with a lump hammer.

After the plates were welded into place, I ran a filler bead along the inside of the cut to close it up. Photos show my first effort then I had a second go at it with a much tidier result

Yes Tony those thin disks work really well, the best hint I can give you is to start from the middle of the cut then work outward from both directions. Used that technique on my second through to fourth cuts and it went very well. I clamped a guide and cut along it for all the cuts.

Dodgy First Go
DSCF0001.jpg DSCF0002.jpg

Final Result
DSCF0003.jpg DSCF0004.jpg

Regards
Ross

[Surfcnc]

This rainy weekend, myself and my electrician took to the control box to see what we could achieve. We segmented our knowledge into our allotted voltages, me the low voltage and him the 240 voltage and together we came up with a great result. Definitely not finished at this point as the Estop, the run button and the pause button are not wired up yet.

My sparky takes credit for the box layout and acute attention to the voltage separation. He is also a dab hand with the side cutters and the boot lace crimper. I work in slow motion by comparison with my wire strippers and the crimper. An issue for further research by the sparky is the possibility of fitting an electrical safety switch.

The end result - I connected up a little Atom 330 PC (the one I built up to prove my self wrong in a previous post), the parallel port external cable, one stepper on the X axis and ran the Mach3 trial cutting the road runner.... it moved. In case your thinking this is just to good to be true, the next test was to jog the stepper and it just pulsed back and forth. Re-tuning the motor to a slower speed helped and away it went, spinning smoothly in response to a jog command.

Regards
Ross

[javeria]

Ross have you mentioned anywhere the size of the box you have used to house the electronics.

[Surfcnc]

Hi Irfan

The box is 600 x 600 x 200 mm. It is all steel including the backplate that I'm using to heatsink the stepper drivers.

Regards
Ross

[Regnar]

The wire ducting does make it look nice! I could be wrong but by using a metalic plate around your printer port that comes out the side I believe this now grounds your breakout board with the rest of the system ( no longer opto isolated ). Again I could be wrong but I thought I read it somewhere around here.

[Surfcnc]

Thanks Russell - I had all the gear to insulate around the expansion socket but ended up doing it this way. My reasoning at the time (*note I welcome any factual input to the contrary) was...

1. the tang is normally touching all the metal on the PC case where it would usually go.

2. looked at how the plug connected and could not see how any connection / short to the pins could be made to the box due to the plastic in the middle of the external connector and the plastic shield at the back of the cable end were the pins run into the inside of the box as a flat cable. In summary, the modified tang literally holds the connection in place and has no electrical connection to the PC or the break out board other than the earth in the box itself.

Over to those that know about these things for their comment.

Regards
Ross

[riesvantwisk]

Ross,

Russel is right here. You don't want the printer port be 'grounded' on your case.

The reasoning is to prevent ground loops, if you ground your printer cable like this you get the following ground loop (in it's simple form).

220V Socket => computer case > printercable => controlbox => 220V Socket

You need to break that loop by not letting the shield of the printer cable to touch the control box. The tricky thing is that this setup might work for you, but it's a potential source of problems.

You could also not ground your computer But the control box is called a dirty ground, a dirty ground can have potentially lot's of 'signals' in there from router, Gecko's etc... You don't want to have these signals around your printer cable. So the best is to have the printer cable, computer and bob be electrically separated from the controlbox.

Obvious you do need to use a plug for your computer with ground for safety, So the control box and computer are grounded/connected to each other outside the control box only avoiding ground loops.


Ries

[Surfcnc]

Thanks Ries

I can always make the hole a little bigger and use a plastic insulator.

Cheers
Ross

[Surfcnc]

I have put a multimeter on the across the two cable shields at each end of the parallel cable that I am using to connect the control box to the PC.
No continuity is evident, that is, there is no connection across the two cable shields on the control box and on the PC to make the earth loop.

Checking the pin outs for the DB25 connector shows lots of signal grounds as per the image attached. I assume these grounds will always be made from the control box to the PC and contained within the DB25 connector itself.
DB25 printer pins.jpg

I can't seem to understand the difference isolating the control box shield end makes to the ground loop, can anyone assist me to understand?

Thanks
Ross

[riesvantwisk]

Ross,

if you don't measure a connection between the two ends of your parallel port connectors measuring on the metal part of the connectors then your cable might not be shielded in the first place. So essentially you will not create odd ground loops with THIS cable.

Just be aware if you change parallel port cables that does have this shield, then you might introduce odditiy to the system just be aware of that.

As a side note and tip, is your cable fully wired?

Ries

[Surfcnc]

Thanks again Ries, a good explanation so it is nice and clear now. Different cables are made in different ways.

Is my cable fully wired? good question. I think it is, but thinking so is not enough evidence to actually know it is. Tomorrow I will check the pins for continuity and get back to you.

Regards
Ross

[Gerald D]

Irfan also had problems with a cable that had too high a resistance.

[javeria]

yes I had problems with high resistance cables and a 0.1 to 0.3 ohm resistance is what you should look for.

[riesvantwisk]

Irfan,

although I trust your findings I find 0.1 to 0.3 ohm resistance really low (measured). My multimeter (not one of these fancy yearly calibrated meters) does measure 0.5Ohm just by shorting out the wires from the multimeter itself I know you can correct this in your measurements but read on...

Looking at this table : http://www.epanorama.net/documents/w...esistance.html (see Copper wire resistance table)

For a AWG 28 it states 43.4Ohm/100Meter = 4Ohm per 10 meter, my cable I just tested measures around 5.5Ohm for a 10 meter printer cable (again I have a cheap multi-meter) . Unless you had a really short really short cable, in the order of 50cm then you would measure these low values (may be a AWG 26 cable??)

But, the input resistance of PC->BoB PMDX122 is in the order of 10KOhm pull-up if I remember correctly, so I highly doubt that even 10Ohm or 100Ohm cable resistance will make the system fail on just cable.

If I would needed to trouble shoot this problem, I wouldn't think of a resistance problem of the printer cable when measured < 10Ohm.

Ries

[Surfcnc]

Ries

Tested the cable and it is fully wired on all the pins, so hopefully that means I can set Mach3 to do what ever it wants down the pins. Also checked the resistance of the cable, it is 6.2 ohms for a 3m cable.

My cheap and cheerful multimeter also records around the same value of .4 - .5 ohms on the 200 ohm scale when the probes are shorted. As it has no infinity icon it reports 1 and a few spaces back to the decimal point when open circuit.

Spot on with the info there Ries.

Regards
Ross

[javeria]

Ries - my multimeter is a fluke 87v, which was calibrated 2 years back, now if that matters or not I don't know-

this is where I have documented it http://www.mechmate.com/forums/showthread.php?t=1180

In my case I did not find find any difference in cuts when I did smaller 2 d works - but on a detailed cut sheet the missed steps was noticeable.

[riesvantwisk]

Irfan,

I am still surprised that your cable is 0.03Ohm, I re-checked the specs here : http://www.mwswire.com/barecu6.htm and 1000feet of awg26 is 40Ohm, one meter would result in 40Ohm/(1000feet/30)=0.833Ohm/meter, that's not even close to 0.03Ohm!

I fail to see the difference between 3Ohm and 0.03Ohm while a breakout board with a good input impedance would be around 10KOhm, I wounder what teh Specs of your BoB say about this...

Anyways, your systems works, that's what counts.. and Ross knows hat to lookout for!

Ries