I don't see much need for this page, anymore... does anyone disagree?  Two posts, or so, this year.  I'd like to recycle the IPv6 address- which are in short supply. 

One of the things I really didn't like about the way the DSP is initially programmed was the touch off for Z height using the macro function (holding the two buttons to trigger it)... for most projects it wasn't 'that' crucial--- a few thou of variance cutting a profile/contour meant only the difference of not cutting through material or getting a little bit into your spoilboard.... both not a big deal and leading to just adding a few thou to cut depths (such as putting depth of cut in vetrics to z+.01, for instance)...

that doesn't cut it when cutting a 2.5D project, or cutting pockets that are to be seen where one bit clears and the other cleans the corners- that depth is crucial, then.

so... i programmatically altered the function for slowing the auto-z down some, and double tapping the touch plate. 

if anyone is curious as to how this is done and would like to follow suit, you've read the right thread. 

you'll need to fire up your machine and home it- and then insert a clean thumbdrive... do to machine setup in your menu, then to 'backup data' and select teh thumbdrive, and save it.... then, open your thumbdrive on your computer using NOTEPAD (notepad doesn't insert weird characters that other editors will- you MUST use notepad).  Save your file, which is titled data.bak as ORIGINAL_data.bak and put it in a safe place never to be altered... save the open file as data.bak... the machine will only read the one titled data.bak and it MUST be saved as 'all files' instead of a .txt file.

that aside, and most of you guys know that by now, you're gonna want to open that file and find several lines of code to change.  do it like this by copying the part in 'find' using 'find' in notepad edit menu:

FIND:
AUTO_TOOLSETTING_SPEED=10.000000Change to:
AUTO_TOOLSETTING_SPEED=5.000000
^that slows the Z descend to half of what it was...

then:

FIND:
TOOLSETTING_REPEAT_COUNT=1Change to:
TOOLSETTING_REPEAT_COUNT=2
^That 'double taps' the probe to ensure your Z is collected accurately.

and while we're at it... something else that bothers me about the distributed data file is the 'spindle acceleration' of seven seconds in countdown before the axis take motion... this is fine if your first cut is somewhere away from the datum position (origin), but i've noticed the spindle not having time to fully accelerate while using higher spindle speeds if the first cut is close to that origin... so... i've changed mine to 10 seconds instead of 7 to ensure it's at full speed before plunging and cutting... if you want to change yours, it's in milliseconds and it's:

FIND:
SPINDLE_ON_DELAY_TIME=7000Change to:
SPINDLE_ON_DELAY_TIME=10000
^that's the simple difference between seven seconds to speed up to ten seconds.

If you guys are using the touch plate that came with your machine, chances are you're using this one:


If so, congratulations- you have one of both the worst and the best touch off plates made.

that's weird to say, huh?

If you want to disassemble your touch plate (aka CAD) go right ahead.  there is nothing difficult about it or restoring it... if it works 'good enough' for your purposes, well, that's good too... you can leave it be.... however, if you want to turn that thing from a consistency of maybe 95% and accuracy to maybe .05 to 99.9% and .003, read on.

the design is pretty dang good on those things- except for the anodizing.  that anodizing can cost you connectivity and sometimes a bit... the wire lead that inserts into the casing is held with two set screws... this is another problem... the spring rides in an insulated/isolated sleeve and the final coil takes a right at 10 o'clock and shoots toward 2 o'clock... this is an issue... the screw that attaches the magnet extends through the casing and maybe .25" into the spring basin, and when that coil takes that right?  it's 'hoping' to make contact with that screw.... and that's how this thing works.  out of the box? it's not a very good device.  I stopped using mine about a year ago altogether- opting for manual z zeroing and limiting my accuracy to .5mm which is the difference in lowering between z- button punches...

it worked okay... but at some point my spoil was going to get cut up to a degree it would have to be replaced.

i'm tired of replacing spoilboard all the time- like, once every two weeks or so... furthermore, i'm tired of dialing the toolpath to skim .02 and watching it cut over a 16th".... so.... I fixed both issues.

first off, the CAD/Puck/touch off plate... instructions follow:

• disassemble it.
• remove the set screws holding the wire lead and carefully pull the wire out.
• using wire strippers, remove some shielding from the end of that wire (the end that goes into the CAD housing)
• twist that wire back into a fine tight configuration.
• break out your drill- select a bit that fits the hole the wire inserts into and leaves just a touch of margin around the perimeter.
• start warming up your soldering iron.
• get your solder ready to go.
• drill through the casing between the end of the hole the wire fits into and the spring's cavity/tunnel.  through the 'wall' if you would.
• run your wire through the hole it's intended to use, but all the way into the spring's tunnel (it'll be at the bottom if you drilled nice and level)

here is where the approaches vary... y'all are smart folks, I'll let you decide how you want to proceed... if one way I'll offer is better than the other, I can't figure out where or why...


way one:

• solder the wire to the base of the casing/bottom of the tunnel making sure the bead line comes into contact with the base of the spring.

way two:

• introduce a washer- one small enough to fit snugly at the base of the tunnel and with space enough the magnet screw can pass through.  solder to wire to the washer.

way three:

• expose enough wire that it can be pulled through the top and a loop formed that can be tucked back into the tunnel and fit snuggly against the base and the walls of the tunnel.

way four:

• (what I did) loop around the base of the spring with the wire, snugly, and drop a bit of solder on it.

then:

• reassemble the puck/cad/touch plate.

now, your accuracy just tightened up like a ducks backside when it plops on the pond.

then..... there is another thing. 

the button on your CAD is likely scarred from touching off with sharp tools... V-Bits with points instead of flats are notorious for this.  that should be smoothed down with fine grit sandpaper- try not to take anything off and for the love of all that's hold don't use a grinder... you want a flat and even surface on that button... afterwards, measure the stack height of your CAD... they're supposed to be 43mm, but they can surely vary some... get a good solid measurement of it.  repeat it and get the same. 

there is, yet, two more things... you've got to confirm your stack height for the puck/cad in your system.  You can find it under machine set-up, or you can back-up your profile/settings using 'system setup' and then 'backup settings' (select to save it to your thumb drive)... either way, the setting you're looking for in the file or using the pendent is "feeler gauge height"... if accessing using the file method, it's easier to just do a search for "43" in notepad (USE NOTEPAD, NOT WORDPAD AND CERTAINLY NOT WORD OR ANY OTHER WORD PROCESSOR!!!).... change that 43.000000 to whatever the measured height of your CAD is...

now, IF YOU WANT, there is something else you can do- it adds a few seconds to job setup, but, it further increases the accuracy and keeps you from making new divots on your CAD's button.... ask me how and I'll share... I don't know how much interest there would be so i'll save it for another post.. you can see it in action in the video below.

Newer official released post processors will likely have this feature, but for now it's just a little something you can add to your existing vetrics post processor IF, y'all hear me? IF you have massive files being sent to cut and the machine chokes on them. 

a little of explanation: 

have you loaded what you know to be a sizable file on your machine, be it a T, an S or running DSP, or a 4x4 running with an HD100 controller?  You'll notice when the machine scans the file it shows a percentage of loading... what the machine is doing other than the obvious is it's "looking ahead" .... it will not let you run that file if that file is going to make it crash.  you may have seen this before with an error message such as "Y is out of soft limits"... because of this function, and this function is crucial, the machine can only chew on so many lines at one time. 

each controller is different and it's not something that can be altered. 

we're talking a LOT of lines, but, there may come a time where you encounter this.  Displayvets found it when he was trying to cut a massive single cut file that neared 1M lines.... that, y'all, is a LOT of lines.

so IF (there is that IF again) you are a user that finds themselves cutting huge (most often 3/2.5D) files, you may want to consider this little adjustment to your post processor. 

here is how it's done:

go to your vetrics file and find the post processor you use.  open it in notepad (only notepad or notepad+ as anything microsoft will add character you don't want in that file)... then, find the block of information that reads:


search for this or the block that looks really close to it- it's near the top of the post processor file:
LINE_NUMBER_START     = 0
LINE_NUMBER_INCREMENT = 10
LINE_NUMBER_MAXIMUM = 999999

and change it to FOR DSP CONTROLLED MACHINES: (copy this and paste it to replace the above block)
TAPE_SPLITTING = 250000 100 "%s_%d.phs-v" 1 "YES"
LINE_NUMBER_START     = 1
LINE_NUMBER_INCREMENT = 1
LINE_NUMBER_MAXIMUM = 9999999

or change it to FOR HD100 CONTROLLED MACHINES:
TAPE_SPLITTING = 250000 100 "%s_%d.NC" 1 "YES"
LINE_NUMBER_START     = 1
LINE_NUMBER_INCREMENT = 1
LINE_NUMBER_MAXIMUM = 9999999

or change it to FOR T-Series MACHINES:
TAPE_SPLITTING = 250000 100 "%s_%d.nc" 1 "YES"
LINE_NUMBER_START     = 1
LINE_NUMBER_INCREMENT = 1
LINE_NUMBER_MAXIMUM = 9999999

a real quick run down of what this does:

TAPE_SPLITTING is a command to tell vetrics to only post X number of lines in your tool path, and in this case the number of lines is 250,000. 
100 is telling vetrics to look backward of line at 250000 and see if there is a retract command; if there is to stop there instead of at 250k.
"%s_%d.nc/phs-v" tells vetrics to write another file, starting at 250k of the original file (or within 100 lines of if there is a retract command) and write a new file, and to name it with an incremental number.  i.e if your file was called 'cut-this.phs-v' and the original file was 1M lines long, you'd have saved four files instead of one... one named 1-cut-this.phs-v, 2-cut-this.phs-v, 3-cut-this.phs-v, and 4-cut-this.phs-v... or with file extension .nc, depending on what controller you're using.

the other alterations are small- such as start counting lines at 1 and increment them by 1, i.e lines 1,2,3,4,5 instead of lines 10,20,30,40,50 being in immediate succession.  why did I do this? because adding a single byte (that zero to every line) adds up quick when you're talking about 250k lines. 

once you've made this change, click file-save... make sure it's saved as an 'all file type' and ends in .pp ..... then, fire up vetrics and go along your merry way.

adding this to a post processor should cause you no harm whatsoever even if you never approach that 250k number of lines... what it is, is a fail safe so you don't have to painstakingly and manually break the code apart.

For SCV44-Series Machines running HD100 controllers and with Vetrics software up to version 11.5.

This one is solid as a rock and for SCR-Series Machines running DSP controllers and with Vetrics software up to version 11.5.

I've long been a supporter of Amana almost exclusively.  ...and for one reason- their advertised dimensions matched the product.

until recently. 

i worked through a shelix geo down cut three flute and replaced it- same part number and the diameter is smaller... the difference between .46 and .5... same with a chip blaster.  but more concerning is replacing bits i use all the time.. i do it in pairs, usually, so i have a spare if i need one between ordering and receiving.  we're talking 1/8" oflutes, 1/4" oflutes, 1/8" and 1/4" DC and UC endmills, and 1/8, 1/4, 3/8, and 1/2 compressions bits most often... at this point, though, this should include the few v-bits that i use (mostly for chamfering runs)...

every amana i bought over the last month has broken- the same speeds and feeds as the previous bit with the same part number- no hesitation, just snapped... i wrote amana and they don't seem to care about it at all- so- i don't care about the brand loyalty anymore either..

bitsbitscompany - help!  anyone else with advice or specific brand please pipe in!

S-Series DSP machines Fusion Post Processor: