The MB6582 Panel Group Buy : FINALLY!!

[NorthernLightX]

As you can see on the right panel the screening is too far to the left, almost off the panel. Also the screening on the left side of the panel overlaps with the holes, making it unusable.

@Twin-X: please read my PM and reply about what to do with the faulty panel.

@Altitude: thanks for the effort, but the factory frakked up bigtime. I'd hate to be in your shoes on this one...

P.S.

Sorry for the long wait, I have been absent from the forum for some time, I hope to be a regular again from now on  ;)

[Sasha]

As I worked in screen printing industry for few years I hardly can believe somebody could actually not see this kind of error or let it out of the company like this. Not to say anything about very bad screening ink. Everybody here on the forum would do better job "hand-screening", without use  of any printing table. Outrageous!  :o

[Kartoshka]

Finally, the cutting was done today.

everything was done using one 3mm cnc bit. eventually it survived!

Videos on youtube and some of images here, additional are on facebook's Midibox group.

Guys, i need your help.

our first trial.... reveals some strange oddities!

i'll explain:

solidparts were exactly as they should have been;

dimensions of front- and rearplates are exactly the same as PACTEC provided to us.

(front 232.41x163.27; rear 217.93x52.58)

(also as they are, i have one of these boxes waiting for the pcbs so smash if you're reading this, keep that in mind please).

what happened at solidCAM stage (or CNC), is beyond my ability to describe (nor that i made solid parts as well  ;)

what happened, is some strange scaling on both of plates:  ???

on rearpanel - on X axis it substracted 3mm resulting in different proportion of distribution of holes

on frontpanel - substracted 3mm on both axes, like it should be wider and taller by 3mm on each axis resulting in different distribution as well.

in addition, we mistakenly deleted one of the leds in solidpart while deleting typography.... but that can be easily fixed, while scaling is more problematic and should been solved at higher priority.

anyone with CNC/solidCAM/solid knowledge, is welcome to drop some lines advicing how to avoid this mistake in near future.

final word,

i'll be cutting the plates again next week.

Kart.

[Sasha]

on rearpanel - on X axis  it substracted 3mm resulting in different proportion of distribution of holes

on frontpanel - substracted 3mm on both axes, like it should be wider and taller by 3mm on each axis resulting in different proportion as well.

To me it sounds like calibration problem. As far as I know in CAM software you should calibrate all axes. Or maybe it is the case only for DIY routers, and factory made machines have that done automatic. That is first thing that I would check.

[Sasha]

Where is the link for youtube videos you mentioned? I tried it`s search engine but you probably didn`t tag it as I expect.

[Doug Wellington]

Here's my guess...

You're off by 3mm, right?  And your cutter is 3mm, right?

How did you define the hole locations and the cutter path?  Does your CNC program require a tool path and bit radius offset or do you have to calculate the actual path yourself?  In other words, if you define a path and run the 3mm cutter centered down that path, the metal will be actually cut 1.5mm off to the side of that path.  If you then go to the opposite side of the material and go down the path, the 1.5mm will be cut off in the other direction, compounding the error.  Depending upon whether you're doing an inside cut or outside cut, you'll end up with the metal either 3mm too small or 3mm too big...

Doug

http://www.analognotes.com

[Kartoshka]

Does your CNC program require a tool path and bit radius offset or do you have to calculate the actual path yourself?

i think on the front it was done automatically while on the rear the path was set manually... but i have to make sure with people who actually done it, i'm not that deep into industrial design and machinery knowledge.

if you define a path and run the 3mm cutter centered down that path, the metal will be actually cut 1.5mm off to the side of that path.

that's what i thought also... but they told me that it never makes cutting on the path itself, only inside or outside... maybe they're wrong  :-\ then how on rear it shrink only on X, ignoring Y-height  ???

also, what bothers me is that holes don't exactly match preprinted image (scaling! not offset)

sunday i'll have caliber at hand and will make exact measurings of holes, errors, dimensions, etc.

[Doug Wellington]

Nice video, but it scares me that I don't see a lubricant/coolant spray...  That tool won't last as long without it, and the quality of the machined surfaces won't be as good...

Doug

http://www.analognotes.com

[LaRssA]

Hi Kartoska!

I used to work as a CNC-miller... I might be a bit rusty, but I was pretty much hardcore at EdgeCAM and Solidworks in my day...

Depending on the machine language you are using, check the CNC-code for the following code after the tool moves down on the Z-axis..

It should have "G41" when it moves in X or Y after that... That's the code for tool-compensation (left side.. G42 is right, and G40 cancels it) If it's not there, check your CAM-software. You are supposed to be able to add compensation there before compiling the code. Or you could add it manually if you know enough CNC to know the compensation-syntax.

Regarding the cutting - I've done some panels myself, and it's a bitch to get it 100%... I'm assuming it's aluminium, so here's a few tips:

    - Use a tool with two cutters (this can also be used as a drill, but I'm guessing you know this already)

    - Use really high rpm, and a low feeding

    - Use A LOT of coolant. Not only will it help to cool it of, but it will help you remove flakes, and stop the tool from being packed.

        Since it's aluminium, the melting-temperature is really low, so if it gets too hot, the flakes will be packed around the tool, and it'll

        break. Also, use a steady tool-holder and make sure the plate is clamped really good and tight. Clamping it on a piece of PVC also

        work really good. You might even mill it in parts to make sure it's clamped good enough...

Looks to me like you have "edges" after you've finished milling.. That's cause you spread it around like butter instead of actually cutting it. I might be wrong on this, though.. just looks like it.

Let me know if I can be of any more help, and good luck! :)

Lars

[Kartoshka]

Peeps, thank you!  :-* that's the pleasure in forums where people gather by interests and not by their physical locations.

I posted another video from beginning of cutting. btw, the speed was set at about 1200RPM (maybe i mess smth as at the CNC PC screen we were paying between 120-350units at speed controller). Whole run for both plates took about 40-60 mins.

a tool with two cutters

yep, it's that one.

plate is clamped really good

yes, it was placed on sandwiched plywood, of about 20mm thickness.

i can't find any G42, G40s... therefore i post original solidcam CNC txt outputs if you can debug them :-)

http://rapidshare.com/files/73443641/karto_CNC.zip.html

what's interesting is that we can see correct dimensions at a3 and b4 files , the final frames of plates.

[LaRssA]

Hi!

I haven't been doing any solidcam before, sorry... Only EdgeCAM and Solidworks for milling in 3D.

Anyhow - I checked out your code, and it seems to me like it's missing indeed. It also looks like your milling machine

are using a different control-system then what I'm used to. I have just about only worked with FANUC, and there are some

things I'm not quite sure what does in your code

(B4):

%

G90 G71 G99

T2

S100 M3              <--The code suggests 100rpm, even though your video shows more. 3mm tool in aluminium can be run at 5000 rpm,

                              if the machine can handle it. You can use A LOT of speed here, and a lot of coolant. (Did I mention coolant?)

M7

G0 X54.483 Y54.053

Z10.

  Z2.

G1 Z-0.5 E330                                <---I'm not sure what the E does, that's why I don't think this is the FANUC-standard I know.

  X217.932 F1000

G2 X219.407 Y52.578 I0. J-1.475

G1 Y0.

G2 X217.932 Y-1.475 I-1.475 J0.

G1 X0.

G2 X-1.475 Y0. I0. J1.475

G1 Y52.578

G2 X0. Y54.053 I1.475 J0.

G1 X54.483

G0 Z10.

G0 Z50.

G0 X0. Y0.

M5

M9

M30

%

If you can find out this for me, I can check around for proper coding...

On FANUC, you insert the tool radius as a value (either on the machine or in the CAM-software), and just include G41 X-value Y-value as the final line. It will then move up until the X and Y, compensating for the tool radius, and you can use the finished values as your X and Y without thinking of the tool radius.

Lars

[smashtv]

Wow your coolant is so fresh I can't even see it!  ;)

<---I'm not sure what the E does, that's why I don't think this is the FANUC-standard I know.

E is a counter for macro functions where you are cutting the same shape but shifting it's position or other variables.

Apparently dreamed up by the perforated metal industry (;D) it can really break tool width compensation on a lot of parsers.  Any toolpath optimizers that use E or macros should be avoided, and the developer hunted down and macro'd.  ;)

Macros are throwback to when NC ops had to feed paper tape to the machine for commands, so compact job code file was king.....Now that you can blast your job straight to the mill via network IMHO verbose human readable Gcode rules.

Best

Smash

[LaRssA]

ah, so that's it... I'm used to work with G-codes for the different macros..

same shit, different wrapping ;)

[Kartoshka]

the CNC machine itself is called SERVO.

 

speed field...

but what about the dimensions in these files? ain't they should be either less by 3mm bigger by 3mm ???

like, as i understand, my code gives the machine a path for tool... then maybe on a operating station it should be finally configured and it wasn't?..... hmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm

i want to solve scaling issue  :'(

and... we spilled plenty of coolant, you only saw the beginning of cutting. Every tool-down on the plate rised coolant from every hole made.  :)

[huck]

if you do everything in absolute you can avoid scaling.

[LaRssA]

That's true, Huck...

Anyway, since I don't know that system, I'm suggesting that you compensate for all the cut-outs manually.

Just fool the machine to make it 3mm bigger in each direction so the final result will be correct. You won't have

to think about this for the 3mm holes, since they are being drilled.

Lars

[Kartoshka]

thought about it too, this could be nifty solution.

should check with them (incl. redoing solidcam again).

[Echopraxia]

Hey just wanted to throw out an offer to buy any ones faulty MB SID V2 front panel from the bulk order. PM if you want to get rid of yours for resonable price in the US. Thanks

[Kartoshka]

hmmmm the person doesn't get back to me... i fucking feel like disabled when i can't fucking do solidcam by myself..

it's not a hard task but you have to know how to do it, and sorry it's not my cup of fucking tea..

:-[ hate to be waiting for no reason (or a good reason and i don't know it)

i'll try to contact my ind.design friend this week again, have to fucking insist of doing it RIGHT!!!!!!!!!!!!  :-X

[supraanimo]

Hi, all if you are looking to get rid of a couple of your deffective panels PM me, I might take a few of them

[illiac]

I as well will take a defective off somebody' hands. US please.