I refused to sell the thing with mach2/stepper
les , fine, just sell the license to someone else is all
That machine needs servos and emc.
anonimasu: how often does the bug appear?
here it shows up after about 120-150 mode changes
Has emc *ALWAYS* been gpl?
Jymmm: emc was released under PD first
emc2 is GPL
alex_joni ok so emc2 is GPL, but emc1 is PD ?
BDI == gpl or pd ?
anonimasu: strange.. all I can get is tkemc to lock up
Jymmm: emc1 is partly GPL too now
you can check out some very old version if you want PD
BDI is GPL afaik
cvs up -d "01.01.2000"
* alex_joni goes to bed
what's left of it
anonimasu: can you say how often the bug shows on your setup?
alex_joni: I just did.. very very often
after 1 mode change? after 10?
do you change extremely fast from Manual to MDI to Auto and back
on the lower period setting..
how about higher?
can you check smthg?
it appears less on higher period settings
1. open a console, and do a tail -f /var/log/messages
2. open an editor and edit emc2/src/emc/motion/motion.c
This guy has one machine in the states. Sales are 6-8 weeks lead....yeah sure. Won't fly.
2.a find the function int init_module(void)
and uncomment this line:
recompiling it now..
3. cd emc2/src/emc/ && make
now.. after the lockup check if the motion understands the commands
e.g. the command number increases
and there should be messages like: FREE, TELEOP, COORD
err TELEOP only for nontrivkins, so probably you won't see it
but FREE & COORD
and when you're switching from MANUAL to MDI it should say COORD
switch back to MANUAL, it should say a list of ABORTS then FREE
_sem.c: rcs_sem_wait(): Invalid argument
Aug 12 00:04:10 localhost kernel: 50880: CMD 234, code 15 COORD
what messages show after it fails
all of them?
after it's locked up
hit F3 for Manual
Aug 12 00:05:05 localhost kernel: 105933: CMD 239, code 14 FREE
still the same message?
so motion gets the right command
I had to force it a bit to make it lock up now..
tracing where emcsh actually gets the idea of the current mode
since I have the default period now
* alex_joni digged for that bug
wot you broken now ?
report it at least
where are you looking ?
one question for you
motion seems to be getting the command
and interpret it ok
what bugs me..where the heck does emcsh get the mode from?
I'm with paul_c.... What did yo ubreak now alex_joni!?!?!?!
I can't find where that gets written to
alex_joni you broke that too?!?!?!?!
how do you mean "get written to" ?
like where does that data come from?
does it map to a motion struct?
it gets passed up from motion via emcStatus
any idea where that happens?
lemme pull up a LyX doc of the status data.
can't find that in motion
all I found was emcstatus.motion_state
emcStatus shouldn't be in motion afaik
emcmotStatus is however
somewhere there should be a mapping between the two I think
from what I'm reading now, there's a function update_status() that should take care of reporting the current status up higher
there might be a way to investigate it further
did you ever use usrmot ?
try running emc the normal way, and use tkemc to lock up the mode change
then run usrmot
and enter the command "show", it should prin the current emcmotStatus
hm, no idea..
yes I am running emc2 the normal way
alex_joni: Start with grepping for determineMode()
tkemc dosent lock up
I mean the mode change should lock up
hm, no idea
motmod is what I am running
motmod is the core module
usrmot is a userinterface to talk to it
but not quite a GUI
anonimasu: one more question. does it always lock on the same mode?
example: won't change off MANUAL
I think it always locks on manual
mode is updated via emcTrajUpdate
gawd I hate such programming :
motionFlag & EMCMOT_MOTION_TELEOP_BIT ? EMC_TRAJ_MODE_TELEOP
motionFlag & EMCMOT_MOTION_COORD_BIT ? EMC_TRAJ_MODE_COORD :
That's the section.
Now you need to look at the RT motion that updates motionFlag
my guess is the update function(s) is buggered.
motionFlag doesn't update anymore after the lockup
That's what I said in the notes in the bug report - The command is not being honoured at the RT motion level.
well the command is beeing honoured
but smthg is messed up in the process
So look for [SET|GET]_MOTION_[COORD|TELEOP]_FLAG
already doing that
* alex_joni hunts COORD
based on :
for those are the macros that handle the setting of the flags
Aug 12 00:50:32 localhost kernel: Motion flag 0007 -> 0001
I think I know what happens
the INPOS flag gets set to 0 somehow
and that leads to bad things
as the next mode change won't get honoured
/* not in position-- don't honor mode change */
emcmotDebug->coordinating = 1;
no, it's supposed to do that
during a move you cannot change modes
as bad things will happen
looked at motion_state....
and INPOS reflects that (or should)
that's not really used from what I see
suggestion: Leave it for JMK to fix.
I'm not going to spend any more time on it... (sorry anonimasu ).
alex_joni: If I were you, update the bug report with your findings
anonimasu: try restarting emc
and then go to bed.
paul_c: I dont bother more, I'll be running emc1 with the USC
as for now..
until I get time to fiddle with a driver for emc2
if I ever..
anonimasu: can you test if it works?
heh.. maybe someone else will
ok.. try again please ;)
only 1-2 more times
anonimasu: still there?
one more place where it can happen ;)
tell me when I should run it again
did you break it?
I cant lock it up now..
it doesn't seem to lock up anymore ;)
you broke it.
BASE_PERIOD = 0.000025
but it's only an ungly hack to see what causes the lockup
don't be too happy
you can try any period you like
oh it's the period I use
I don't expect it to work properly under any conditions
like: change modes during a jog
hm, I usually dont :)
I guess now
ok. going to update the info on the bugtracker
anonimasu: enjoy your emc2 ;)
alex_joni: thanks :9
going to bed now
anyone home ?
I have several machining centers, but emc seems cool
it is :D
how to handle tool changers and that I wondering.
well.. I would chose emc2 for that
there is a software-PLC component for that
classicladder it's called
so you can do PLC programming for the toolchanging
looks like its still under development though
a bit ;)
I mean.. it always is
toolchanger, pallet changer and specific feedrate and spindle pots would be important.
can you describe your needs for a toolchanging?
the rest is basically fluff.
as a sequence, yes.
like the exact steps the machine needs to take to change the tool
move to unload position, etc.
* alex_joni lacks the knowledge of actual machining
at each point if theres a failure then the prox sensor doesnt satisfy, yes i can describe that.
I only do some software ;)
and build my own toy to use with emc :D
Im mostly a machinist.
[09:04:33] <alex_joni> http://dsplabs.utt.ro/~juve/emc/mytoy/
3 servos, interesting concept
open loop generally, or used to be.
You know Jon Elson--he frequents a newsgroup I read sometimes ?
JonE does the USC
and other products, that are supported by emc
no driver for emc2 though.. yet
Anything that will do a single axis ?
feed bar stock into a saw ?
dumb question I know.
would like to set like 4in.
feeds 4in at each cycle, then shuts off.
but you won't need emc for that
there is a new thingie, called HAL
Ya but you already said your a programmer <G>
Hardware Abstraction Layer
it is used in EMC2 to wrap around hardware
hal then it is.
feed 12 ft bars into milling machine.
but in HAL you have frequency generators
so you could set up a freq gen to output pulses for 4ipm
and let it run for 10 secs, or so
or an exact number of steps from the encoder
you get the picture
We are sawing barstock, clamping into cnc machining center one at a time.
there is some software to be written though
maybe even a GUI for the operator
to let him select how much stock you need
would rather place complete bars.
at what speed to move the bars
doesnt have to be very fast.
a pair of vices,
shove the finished part into a catch conveyor
a handfull of io
Ther is 4th axis, but it repeats in absolute every rotation.
coding could be cumbersome using g code for an axis that was meant to be rotary.
these are Fadal machining centers.
pisses me off, their eprom only supports pallet changes on a 20 x 40 inch machine mode in the firmware.
well I think the tasks need to be split up
I have specific 'm' functions i can call
could you drop an e-mail explaining the toolchanging stuff you need?
extras on the board.
best to the emc-developers list on sourceforge.net
That is down the road some, best for me to mostly lurk here, IMO
as the toolchangers I have work fine as is.
well, do it IBM style then
'never touch a running system'
Getting late here.
getting early here ;)
[11:28] <HacKjoB> Getting late here.
I have spare outputs, in case of a failure in a sequence I could simple initiate feed hold to the main controller.
you usually find folks from over the pond too
lurking in here, I mean
oh IRC knows no time boundry, thats for sure
and it's best like that
1:30 am here, west coast US
also if you're afraid you missed smthg, there's a logger in here
NO time to read logs really.
lol.. ok ;)
Ill take my chances <G>
I read back some though.
well there are a few folks around here that might prove a nice company :)
looks like a diverse buncha folks.
we got guys from the US, UK, sweden, Italy
I mainly just have some motion control issues I wanna work out, not necessarily will it be emc turns out to be the solution.
Link to our products :
[09:33:22] <HacKjoB> http://www.thesmallshop.com/products.htm
One more beer an its closin time.
doesn't seem to work (the website), at least not from here :(
Kill enough brain cells an I can become a neo-con like the rest the zombies we got here, eh ?
hmm thats odd, loads here just fine.
right.. seems there's smthg wrong here :)
hmm I better look into that very soon then.
[09:40:13] <HacKjoB> http://www.thesmallshop.com/
do me a favor try this one too then :
[09:41:03] <HacKjoB> http://www.kalamaprecision.com/
both hosted from same server.
smthg's busted on my setup
the server works ok, tried from another host
Okay had me going for a bit.
weird eh ?
they customers use it to bend tiny bits of brass.
so you can see where it would be advantagous for me to feed barstock now.
instead of sawing it first.
I guess :D
not woken up completely yet
was up till 4am chasing a bug
emc2 bug, not a live one :))
didja kill it ?
I trapped it
not fully dead yet
I'll let it starve
it's a bit more sadistic, but it'll warn other bugs to stay away
see what effect there is in quaranineing it
I got a linux box here my son set up about 3 years ago went dead recently.
4 disc drives, in a raid 0 array
one went dead
I got it spinning again, replaced the top board on the drive.
now gotta hound him to get it runniong again
okay Im gone
Thx for the info
and, it looks like the Softools license sillyness is solved too ...
it looks like it, but they _really_ should rewrite that part
robin_sz: check this out
[10:20:11] <alex_joni> http://www.frars.org.uk/images/contentimages/1145.jpg
[10:41:07] <alex_joni> http://www.stealthmicrowave.com/2_4.htm
what happening ?
bad weather here..
hi anonimasu :)
heh, someone else not fully woken up
alex_joni: ah, I've got caffine.
* anonimasu is going to take the mill for a real spin this afternoon
and make some real parts :)
anonimasu: 10 hp ?
but I have motor + pump for a huydralic spindle.. that'd give me several kw ;)
but I'd need to rebuild the column to use it
how much weight ?
less then the electric motor I have there now
but to get it more rigid..
wich material will machine ?
no need coolant for aluminum right ?
well, any aluminium + steel
you NEED coolant..
I use wd40 usually
or dry machine..
with just air..
but it's not enough
air helps lots, but it dosent compare to flood coolant ;)
but mist isnt too good if you havent got good ventilation
but is inexpensive
mist = crap
anonimasu: in german that is ;)
* anonimasu tried running mist it cut well..
but I had a thick fog in the shop
it started to hurt in the eyes ;)
i'm going to check and double check controllers on my little machine
i think PS may is not enough ..
anonimasu: tried emc2?
after the hack?
alex_joni: could you tell me the rilevant advantages of emc2 ?
i never tried..
Jacky^: easy to install :D
compiles on most RT platforms
HAL is _very_ configurable
works on 2.6
will have a STG driver *g*
* alex_joni still works on that
according to anonimasu it works a bit better than emc1
it has i18n support
you could do an italian translation of TkEmc
even if you don
i'm not interesting in translation
maybe someone else is ;)
i'm tryng to learn english :P
you could do it :D
::msgcat::mcset ro "File" "Fisier"
::msgcat::mcset ro "Open..." "Deschide"
::msgcat::mcset ro "Edit..." "Editeaza"
::msgcat::mcset ro "Reset" "Reseteaza"
::msgcat::mcset ro "Exit" "Iesire"
::msgcat::mcset ro "View" "Vizualizare"
smthg like this
leave the text on the left in english, translate the one on the right to italian
i could run it on debian i suppose..
compiling a new kernel right ?
you don't necessarly need to compile a new kernel
if you've got a working BDI, then you're set
ok, thanks :)
[11:28:26] <alex_joni> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?BDI-4_Compile_EMC2
to compile emc2 you need to issue:
and that's it ;)
emc1 is way more complicated than that
i will try
you got to compile rcslib
both with 2 PLAT's
alex_joni is now known as alex_joni_away
Y axis on my machine have input_scale double respect to XZ
is it normal ?
it have same motor, same leadscrew and pitch
leadscrew is 6mm pitch 1 mm , stepper is 200/rev.
so , input scale is 200.000
for XZ it work fine
not for Y
is the Y axis working in half step ?
ic's are L297-298
pin 19 of L297 seem are used to switch half/full
it is in state low
that should be right, i suppose..
the usc works but it ferrors as soon as I move a axis..
it works with the default config and "
seems like input_scale is fscked
paul_c: are you away of anything regarding input_scale on the USC?
it seems like it's scaled differently then when using freqmod..
INPUT_SCALE = number of steps per unit
OUTPUT_SCALE = 1.000
paul_c: that is strange
input_scale = 640
output_scale = 1
I mailed jon and asked him it it were the same
what's the ferror(s) set to ?
I might have too few pulses/divider on the usc set wrong..
ferror = 1.0
min_ferror = 0.01
it works with the inch config
with input_scale set to -65000
or whatever it is
increase ferror to 5 and min_ferror to 0.1 (for starters)
You may find the default PID gains are too high for metric
does it use PID?
oh, well I guess I have my error there :)
* anonimasu will go outside in a couple of hours again
outside is over rated
hm, outside to the shop
alex_joni_away is now known as alex_joni
heh.. Jymmm is up
paul_c: did the stuff I posted on the mode change bug make sense?
I mean.. was it inteligible?
[14:59:56] <Jymmm> http://www.partsexpress.com/pe/showdetl.cfm?&DID=7&Partnumber=122-655
Are they saying two 24V sec @ 115v pri?
whic would be two 12v sec @ 230V pri?
no. It would still be two 24V sec @ 230V
how is that?
* alex_joni goes home...
Jymmm: start the day with coffee
plenty of it
oh, two 115 in series?
so kind rev from typical transformers
* Jymmm thinks he'll take up alex' suggestion =)
ok, coffee brewing...
Alright, if I wire this up for 230V pri, but only apply 115V instead of 230v, I'll get half the sec voltage, but the same current rating?
way too early to be thinking =(
sounds about right ...
typically transformers run on lower voltages are OK, its when you run them on higher voltages tha intended that it all starts to go tits up
hey,, my computer survived ...
i thought it might be damaged
I came >< this close to beating it to death with a hammer
sodding Mozilla thunderbird
* Jymmm gives robin_sz a cookie
* Jacky^ Jacky^ gives robin_sz pizza
hello Jymmm :P
i'm moving optical switches of my machine in correct position
i extended X axis of 10 cm :D
I has a Ni-MH battery leak in my camera's flash. Anyone know of anything I can put on a qtip and neutralize it?
brb, playing with the mill..
going to try and get the PID stuff sane
I usually use baking soda for such things, but I can't get to the battery terminals
no luck with what?
getting the USC to work.
any reply back from Jon by chance?
Just got off the phone with Energizer... They lifetime warranty *ALL* their products, even if discontinued.
Jymmm: why did you call them?
Due to the batteries and charger I have
it's a matter of tuning the PID loop
ROBIN you here?
robin is a toroid guy
I want to confirm the atx power supply ones are Size T94 mix 26 powdered iron
you want the ATX specs?
Near the bottom --> http://www.intel.com/design/motherbd/atx.htm
I need a little inductance to cancel the shunt capacitance of these piezos an want to just hand wind a little toroid inductor
so I pulled one from a popped atx supply
it's just the right frequency range
les : I had a Ni-MH battery leak in my camera's flash. Know of anything I can put on a qtip and neutralize it?
I think the electrolyte is sodium hydroxide
so something acid
I use to use baking soda and water, but I'm chicken to use water in the flash
vinegar sounds good.
well baking soda, like sodium hydroxide, is alkaline
it works on cleaning car batteries.
yeah...car batteries are sufuric acid
I thuought baking soda was a base?
it is. That is another word for alkaline
ah =) I was getting stoned during that class =)
hmm I have no magnet wire
trip to rat shack?
take apart a fan
20" box fan == $9
RS is too small.... I have some
They may not have any
but vinager WILL or SHOULD work?
I just want to test that inductance on the piezo today
vinegar....hmmm will neutralize and make sodeum acetate salt
yeah I think it's safe...rinse with water
max feedrate i can get with my steppers is 360, too slow ?
les: ok, thanks. Trying to save my $400 camera flash from an early death
I just googled nickel metal hydride vinegar and got a zillion hits
so I think it's ok
jacky 360 is awfully slow for wood
steppers are 2A, 200 step /rev. controllers are using L297-L298
Try using a hand router at that speed....
360 is ok for metal and some plastics
I think about the slowest you could go on typical wood is about 1000-1500
some industrial units cut at 20+ meters/min
i will never cut wood so..
well you can use very low rpm but finish will be rough
up to 360 i get no following error
but the motors lost steps
it seem to me..
you are full stepping right?
yeah. mid band resonance. Half stepping would be beter.
put a n absorptive coupling between the motor anf load
les: Can I rinse with denatured alcohol insted of water (after the vinegar)?
is it leadscrew?
les: leadscrew 6 mm pitch 1 mm
jymmm water is to disolve the sodium acetate. Alcohol would be great after to remove any trace of water left
les : ok, will do. thanks =)
Sometimes I just grab part of a leadscrew while it is turning (with oily rag) to make some friction to stop missed steps from mid band resonances
i also used some oil
but i think the problem is in the motors
here is the motor i used: http://www.micromed.it/Elettronica/images/C018.jpg
full step systems need some friction to prevent resonance slipping sometimes
it should be 2A
yes I use those for things.
is that one bipolar (4 wire)?
it is 6 wire
where i'm using 4
using much series resistance?
you mean inductance of motor ?
i've no idea :\
RESISTORS TO THE MOTOR WINDINGS FOR A MORE CONSTANT CURRENT
well, looks like you could use some more speed/power.
the cards are so expensive though.
it seem do not exist any schematics online
i searched a lot
do you have a picture of your machine?
[19:06:00] <Jacky^> http://www.roboitalia.com/modules.php?name=coppermine&file=thumbnails&album=75
ohm for motor are 1.1
controller have a trimmer do adjust current for motor
and also a test point
oh I see..you are using a dremel
not now ..
you need the variable speed one!
i changed it with a black and decker router 30000 rpm
ok...that must be one of the edge trimmer types
talking about the current ,
the strange thing is that if i set controller for 2A the motor dont run
seem too much current
it only work with low current
you know you can buy a speed control that will work with your present router
i will buy a bosh with electronic regulator
but I think on this machine i will not solve at all..
Yes...you need something a bit faster...even at min speed of 8000 rpm
anyway...biaw...must go to town to get some magnet wire
at the end of summer i will start to build a new machine
les: do you think a leadscrew of 8 mm could be better than 6 for my issue ?
rat shack actually had magnet wire
can hardly imagine
yep, three roll pack
Having to wind a little toroid here...must cancel non motional capacitance of this piezo
if you need more wire, take apart some old transformers... microwave oven is a good start!
well I was already desperate...using a T94 mix 26 core from burned atx supply
sounds like fun... I'm just building a constant current source for a possible servo drive
yellow white is supposed to be mix 26 anyway
hard to say in the chinese power supplies.... should be the right range.... easy to try with different # of turns to see if you are getting close
My copleys in the big machine are set up as voltage controlled current sources
oh well those atx switchers are 30 kHz I think
I need 40
yea... my idea is to build a cc supply and then use the pwm to select the current...
just doing a single quandrant now to see how the circuit behaves
Well torque mode works very well for me
all 5.6 kva of it!
wish me luck popping fets... this is a nice design only 1 fet for the actual switch
Mine seem pretty durable even though my estop scheme actually disconnects moving servos and shorts them
Copley eng said the freewheel diodes would protect...but I put some transorbs across them for a little extra
made a proto pcb on my mill (kind off.... only drilled a piece of formica and will handwire traces)
I need to make one of those toroid bobbin things
power electronics is fun.... easy to see where the smoke gets out!
I did some big audio amps
how many turns do you need?
designed one around APEX pa 05 devices
* LawrenceG looking for links
oh....just a rough calc around 3 millihenries
it changes depending on the air cavity tuning so I will just play it by ear
those things are wild!
www.apexmicrotech.com show service unavailable
hmmm I went to their site and they are down!
looks kinda like an out of business thing huh?
found a datasheet... looks like a servo amp!
[20:33:56] <LawrenceG> http://www.ortodoxism.ro/datasheets/ApexMicrotechnologyCorporation/mXrxyux.pdf
here it is also:
[20:34:42] <les> http://eportal.apexmicrotech.com/mainsite/products/pages/op_amps/pa05.asp
what kind of $$?
fre for me.
best price !
they are expensive. several hundred I think.
A cople of them make quite an audio amp
1000 wrms 8 ohm
a good bit more 4 ohm
6 kva peak for a pair
yep... that is quite a power op amp
I used it in conjuction with burr brown op 604 input stage
The devices are bolted on .75" thick case walls
they have very good dissipation
Right now that amp runs my...um...mid bass drivers.
nice 2us rise time and distortion levels are awesum
yes they are very good
Jymmm is now known as MrAsshole
MrAsshole is now known as Jymmm
I need more engineering samples!
thats the trouble when the keep inventing more channels on the stereo!
It is always scary to power up big linear amps like that for the first time
use a big variac
ground loops are tough too even with carefull star grounding
kind of hard on the things they are connected to as well.... failures tend to take the outputs to one of the power rails
well... if I am to get this thing running today, I better power up the soldering iron
On that to get the very lowest humI had to hook a groundup, wind one turn around the outside of the transformer, and hook it to another ground point
and I better wind toroids.
anyone know hot to minimize following errors using freqmod ?
several things can cause them
trying to go to fast for a given PERIOD setting
to go TOO fast
Period should be ok ..
paul gave me some help days ago..
he sayd i can reduce a lot following errors too
but i don't know wich variables to change
do you get a following error on every axis or just one?
i'm getting low speed on all axis: 360
i'm tryng to get axis working more fast
period is 0.000020 and seem work fine
is that 360 mm/min?
yeah.. i think so, i'm using mm as unit
I get 22ipm with PERIOD .000024
the value i read from gui is 360 (mini)
what is your SCALE?
wow, that should be very easy to get faster than 14ipm
is this emc1?
it doesnt :(
what is P,I,D?
P = 100.000 (suggested by paul)
I = 0.000
D = 0.000
that's very close to what I use
should be fine
is your MAX_ACCELERATION very slow?
what triggers the FE? jogging?
max_accelleration in [traj] section is 20
cradek: yeah, jogging
doesn't that mean it will take 18 seconds to accelerate to full speed?
cradek : I remembered! What Rx did you use in your nixie clock (such a tiny enclosure).
Jymmm: external receiver
cradek: no idea, but it doesnt seem to take 18 sec. to accellerate
jacky^^: my config is in inches so I'm not able to help you by comparing
cradek: ok, np ;)
my MAX_ACCELERATION is 128.0
I think that's in/s/s
so for you the equivalent would be 3251.2
servos or stepper ?
so your ACCEL might be set incorrectly very low
my steppers can accel from 0 to full speed in a blink
where full speed = 25in/min
(not very fast!)
not slow ..
as the mine
do you get the FE immediately when you start the jog, or after it gets up to speed?
after it gets up to speed
maybe you should ask on the mailing list
here's my ini file if someone want to take a look: http://rafb.net/paste/results/g1Xyz633.html
* paul_c runs like hell.
You'll never catch me Forest Gump.
the running part was the stupidest part of that movie
paul_c: could you tell me please the variables to check for minimize following error using freqmod ?
metric setup ?
Let's start from the top.... What is the period set to ?
i pasted my ini right now..
[22:24:14] <jacky^^> http://rafb.net/paste/results/g1Xyz633.html
if you remember pc was crashing
now work fine
the only problem is the speed on all axis
up to 360 i get following error
20uSec gives us a max step frequency of 25KHz....
You have 400 & 200 steps per millimetre on the axis....
ah.. then is the max speed i cang get ?
so abs. max would be 62.5mm/Sec & 125mm/Sec
divide by 4 to get a reasonable number...
say... 15 & 30 mm/Sec
i would like to cut some wood part
i've the issue on the bits
I run successfully at 3333 steps/sec with period .000024
all burned ..
after some job
speed router is to fast: 30.000 rpm
paul_c: do you think FERROR and MIN_FERROR are ok ?
4.000 and 1.000
are the default values ..
punching some numbers in...
mmm.. i think my steppers are craps
i will slow down the speed on my router in some way
OK... Under each AXIS_n section, velocity=8 & accel=20
gets us a top speed of 500mm/Min
use the same numbers for the TRAJ section..
Once you have a baseline that works, you can increase vel & accel along with PID...
ok, speed is 480 now, no following errors
thanks a lot