damn sony !
[00:07:26] <Jacky^> http://esupport.sony.com/perl/news-item.pl?mdl=DSCV1&news_id=95
Jacky^: That sounds pretty good of Sony to do that. They even cover the shipping costs. And usually it'll be replaced with a better unit.
Jymmm: bought a camera 2 week ago :(����������
Jacky^ Ok, so return it
and get another one
yes, need to understand if work or not before
.. but may be accelerated if they are exposed to hot and humid environments
what they mean ?
SWPadnos: not stepgen running on my system here.
Jacky^L Take it into a hot steam shower running in the bathroom
want to know how many time of warranties it have..
rayh_emc2: not stepgen? are you using a USC or a UnivPWM?
ah. sorry swp I guess there wouldn't be one
no - jmk called them ppmc.0.stepgen<n>
USC is all I've got running here at the moment.
I'm seeing problems as well, but they may be AXIS related - I'll switch to tkemc and see if they persist
Jymmm: other interested models PowerShot A60, PowerShot A70, PowerShot A75, PowerShot A300, PowerShot A310, Digital IXUS V3, Digital IXUS II, Digital IXUS IIs di Canon; FinePix S2 Pro Digital, FinePix A303, F410 e F700 di Fijifilm; Dimage 7i, 7Hi, A1, Xi, Xt, X20, S414 e F300 di Konica Minolta; Coolpix SQ, Coolpix 3100 e Coolpix 5700 di Nikon, nonch� le F717, U10, U20, U30, U60, DSC-V1 e la serie P di Sony.
a lot of brands other
and video - it's not just still cameras
do not bought that gift for christmas
I wonder if it could be a problem with my ini. Shouldn't though.
not sure - I'm not positive that everything is correct there (at least in the cvs ini/hal files)
last commit message Added an enable pin to the PPMC step generators, and updated the .hal file to us...
yep - they're there - I see it with halcmd
(btw - halcmd -kf gives you an interactive mode - type commands, get answers, doesn't bomb out on error)
Right. I used that in a tickle script but forgot to shut it down before closing emc.
It left all the kernel modules
well - I'm not getting feedback either, but if I change to encoder input, it works fine
I'll see if any steps are being output
hmmm - hold on a sec - that reminded me of something
ok - let me try with my test setup - I can send those files to you if you like
I can see the enable signals change when tkemc comes out of estop and to machine on.
yes - that happens
there are some scale variables that may be wrong though
Need to drive SO to town for a few. Yes send the files.
ok - can I dcc them to you now (they're short)?
oops, sorry - reading about dcc and had something in the buffer.
uhm founded ice on mars
SWPadnos_: I don't think I've got dcc setup right on my net.
ok - I decided to fiddle a bit anyway
I get the message that someone is trying to send me files
but I don't seem to be able to respond.
I never tried actually
ok - I'll email in a little bit then
something may have gotten broken in the last commit, or my board has gone south
I do see the velocity commands in ppmc.0.stepgen ramp up when it tries to move.
I'm not seeing any step output at the moment
then again, I am seeing odd noise on the power supply line, so I need to do some hardware testing
axis.0.motor-pos-cmd ==> Xpos-cmd changes with jog but pos does not.
shouldn't Xpos-cmd be connected to ppmc rather than pid
yeah - the stepgens aren't outputting anything - I'm trying to figure out why
no - pid, these are servo-like
oops. I'm reading the wrong signal
float -2.50000e-04 Xpos-fb
ok - that's one count (probably) on the encoder
I keep getting a following error that causes estop
if you look at param ppmc.0.encoder.0.count, you should see -1
oops - encoder.00
oh so much better - plugging the scope into the same power strip as the USC and emc computer ;)
05 float -W -2.50000e-04 ppmc.0.encoder.00.position ==> Xpos-fb
05 s32 -W -1 ppmc.0.encoder.01.count
yep - as I thought - one count of startup glitch
OK - I see that the step outputs are going - now to transfer some of the stuff from my test .hal file to an actual emc hal file
Jacky^ is now known as Jacky^afk
catch you later.
I played with emc2 + axis tonight. No problems (still getting used to the interface). One comment is that I think the cone should stay the same size on the screen. when you zoom in the cone gets bigger and bigger. Just a comment - axis is awsome.
hey rayh_emc2, is the default jog speed is 1 IPM or 1 IPS?
skunkworks: we've considered a change like that. Right now, the cone's size is relative to the size of the g-code loaded.
SWPadnos_: 1 IPM I think
well, it comes from the inifile
ok - I'll look
though for all I know emc2 is different
jog_speed = float(inifile.find("TRAJ", "DEFAULT_VELOCITY"))
this is what AXIS uses to find the value to put in the jog message
Right and the user should set it higher if they want reasonable manual jogs
ok - I think I'm hitting a rapid bug in emc2, because I can do a g1Z1F1 without trouble, but a g0z1 gets following errors
maybe related to the jogging bug
SWPadnos: is default accel somewhat less than max accel?
yes - somewhat less
but I'm just using a USC with step outputs looped back to encoder inputs - no machine hardware
I have not had any following errors sinse I have stopped using the jog at high speeds - Thanks axis ;)
interesting - default accel is 15, max accel is 20
hmmm - hold on
and I keep the max velocity less than ((1/base_period)/2)/input_scale <- think that is right
rayh_emc2 is now known as rayh-away
Hi All...can somebody send me a file to see if the changes I just made have effected anything ?
seemed to work
OK...I m back to where I was...Can you accept a file ?
from some people ;)
go ahead and try
Did you see a request yet ?
yes, and I accepted
Ok..but the IP address was a 172.X.X.X ?
it's showing 0% complete though
yes - (172.16.1.33:1850)
Ok...still messed up...but I have upgraded my firewall/dsl box software...which I read is neccassary to eventually get DCC working...
ok - it shows it as failed now
Just did some more firewall adjustment...
Still not working...
you can try me
but faster this time
I have the DCC ports setup in the firewall now...now all I need is a client that will go out and fetch the real IP address and send the packets with it..
Which according to what I read, mIRC does that or something similar...
I was getting addresses that look real - not common NAT (192.168.x.y) addresses
Oh ? What was the address ?
and I note that the address in the send request and the acceptance are different
[DCC]Got DCC File Transfer offer from “wb9mjn” (172.16.1.33:2029) of “dm6nov05.jpg” (689 KiB) [Accept] [Decline].
[DCC]Accepted DCC File Transfer of “dm6nov05.jpg” from “wb9mjn” (188.8.131.52:2029).
No...that 172 is my LAN block here ...
Its not a real address...
So, it still has that problem...
ok, but the 184.108.40.206 was different
The address I m using right now is 220.127.116.11
I wonder where the 216 addr came from
That s probably the freenet node...
Once the transfer is accepted, its probably out of their...
Or that is your address ?
wait - yep ;)
Anyway....I have mIRC downloaded...I guess I ll try it...but after a month it costs money...hi..
you'd think that I would actually remember the address, since I'm paying extra for static
back in a bit...
I m back...
how does mirc get your outside address when it changes?
Gotta figure how to do a transfer now...
And also how to turn on the localhost ? feature..
I think it might be dfaulted on...Cradek, ready for a file ?
Yep it worked...That is a picture of my Y drive!
what am I, chopped lilver? ;)
ooooohhh - shiny
Its trying to you now...
I see some 80/20 extrusion there as well
Wierd,,,I never got confirmation of the send ?
Did you get it SWPadnos ?
I never got anything from you - only from cradek
you don't love me any more
I sent it to him
I m trying to send to SWPadnos...
Yea, the plate was raw bar stock....
or swpadnos_ ?
So, I fly cut it flat...as it was convex as it was made by DM...!...
is that stainless flex conduit?
Then I put in the additional counter bore and idler bearing...
Nope, its just Tyco pastic waterproof conduit..
a friend has some nice stainless flex sheathing, that I think is liquid tight
I've been looking all over for it
Yea, on the 8020...I bought about $300 bucks worth of it...Used some for the lathe table, and the rest on the
The pulleys are original from the DM, but the motor just happened to have a pilot bushing that fit
it's cool stuff. I have about 50-75 feet or so of the 1010 (used to be a trade show display)
the belt slot in the DM plate....
That is 2020 there...
yep - I see the two slots ;)
Want to see the whole thing ?
sure - if cradek doesn't mind being a go-between ;)
It actually cut meta for the first time tonite!...
why can't you send to one another?
NAT -> NAT ?
should be fine...
Takes a long time for some reason..but think we can now...
you're both fast for me
14 secs to receive from wb, 4 to send to SWP
yeah - I was just going to ask you what your connection is (but I think I already did, some time ago)
I m using an ADS 3M/.5M connection...
DS3 or some such lustful thing
of course upload would be slower - it is *A* DSL
I'm on 2.0/0.5 here
cool - you even have a chip shield
what's the work envelope? It looks like you have a fair amount of Z travel
I actually knew what ADSL meant some time ago, as I helped calibrate Westell's T1 Test Sets when they made some of
heh - no offense intended :)
the first ADSL equipment for - get this - Video on Demand....Telco execs were braindead back then (1994)...
still are, I think
The traves are 7.6, 6.1 and 7.5 (xyz)...
I recall wanting 'net bandwidth already in 1994
ok - I thought the Z looked disproportionately large
but most people had never heard of it
If I would have had a work table integrated into the motion table, it would have be 10 inches on the Z ...
although 33.6k dialup was still pretty fast
yeah - the only broadband was 128k ISDN, or a T1 (or higher)
or fractional T, for almost as much as a full T
The Tslot (work) table is a precision european thing, with DIN 508 tslots in the x and y...all hardened...
our whole office was on a leased-line 28.8k fill-time modem connection
AND WE GOT USENET OVER IT
I had counter bores added into the corner bolt holes, to be able to bolt it to the motion table...
talk about fill-time ;)
I was on 9600 baud Ham Radio back then, hi !
It worked ... but that was SLOWWW...
I think I was still getting news and email over uucp at home...
not sure about the exact years here
Yea, and Westell was cranking out these ADSL units back then trying to horn in on Cable TV...
that sure hasn't worked out
kind athe opposite, in fact
Well,,,it kinda has...allot of the work that went into it is resurfacing as mesh WIFI companies...
Just because of the restrictive FCC rules, nobody could take a risk on Ham Radio with it...
all I want is the same thing they get in Japan - 100Mbit @ $35-50/month
I'm thrilled with my service here
I get ~ 6Mb for ~ $50
The packet satellites are quite unique in their capabilities too...Anywhere on the world, need a medium sized file..just stick
up a quadfilar helix and wait a day...
One of our ham packet guys ended up as a cofounder of one of the major cable tv internet access companies...
Except he left before the big bucks flowed in...
I hate when that happens
Went back to Tellabs...and had the startup put there cable access buisness out of buisness...
He is really kickin himeself...
Allot of the protocolls for home access over wired networks (PPP is one) were written for Ham packet radio originally...
Anyway...the milling machine is all servos...does about 240 in/min...can do XYZ full travel simultaneously in about 3 seconds...
that's pretty quick
Wonder if Les is up ?
He was the one who wanted to see this...hi...
interesting to note that you have more Z travel than a Bridgeport (assuming the quill is CNC'd)
Would like to have more, but that s going to be it...would like more X, but those linear bearing tables only came in
Its sticking the motor out the enclosure already , too...
When the guy had this on Ebay he misled everybody saying it had a 9 by 6 X and Y ...
the machine operating envelope goes way up when you increase the range
that's the table size, right? not motion envelope
It actually only had 6 by 4 with the original cheapo asian Tslot table, with horribly modified NEMA 34 steppers...
The original stepper mounting was totally poor enginnering...
SO, there was allot of competition, as people were thinking it had the same table travels as the $20K benchtop machines out there...
Anyway, I gave up on the original tables...
it's not too far off, but still wrong
Well, area wise it was being sold 225 percent of what it was !
Anyway...the stepper drive was a R/L drive...with big massive heat sinks in the back...
the table doesn't have to be as big as the travel - a Bridgeport has more Y travel than the table depth
with monster power resistors mounted to them...
ugh - now that's cheap design
And it was all just screwed into the blue box...no sub-panel...
So, first thing I did was pull all that out but the power transformer, and add in a sub panel...
Then put the SCR spindle drive back in, an Opto 22 board, and 4 Gallil servo amps...
Then wired it all to the feed-thru panel...
Need a bigger Z servo motor...it gets a little warm...
Z and Y are fine...
X and Y,,,that is...
Then about 1/2 way through the project, Abdul came out withthe MotenC-100...so went with that...
Its been working well...
heh - you're creeping toward that $20k benchtop here ;)
No..more like $12 K....the mechanicals are bolted together, not cast like in the $20 K machines...
Probably have spent close to $8K...
Half of it being the original purchase...
including tooling and stuff, or just the machine
Really got zapped by bad description....The original column did have a nice 5/8 ths ball scree though...
Just the machine...
Only tooling I have are a few cutters, a drill chuck, and some R8 collets...
hm - I think I've got about the same into my Bridgeport, and I'm pretty sure I have all the parts I need (just not the time)
Yea...I noticed if I would have bought something that weighed a ton (litterally) it would have been cheaper !
heh - all but the delivery (that was a PITA)
This machine only wieghs about 300 pounds...
Including the blue box, and the steel stand with the 8020 reinforcement....
But, its been an adventure, hi...
that's about what my table weighs
Yep...the tool table wieghs about 30 pounds, and the motion x/y table also about 30 pounds (aluminum)...
The column is cast iron about 50 pounds, and the foot is steel also about 50 pounds...
I think my X ballscrew is around 50 pounds
hm - maybe 35
All three ballscrews are on the machine are 5/8 ths...with double nuts...
that should be plenty for that length
At least I got the 6 inch Y travel...and enough Z to play with .... only the X was under the goal I wanted...
AHA!! I think I found the stupid following error problem
Have some more gaurding to do and I m reworking the X table limits so they do not stick out as far...
If somebody can get one of these DM 6 s for under 1.5 K, its good hobby rebuild project...So far tht s my opinion..
But for $4 to 6 K one should be able to get better (brigeman XT ?) ...
woohoo - 180 IPM rapids!!! no following errors in G0 mode!!!
What was the problem ?
(no machine attached, of course)
SWPadnos: Now, turn on the power supply to the drivers.
the stupid PID blocks had their output limited to 1
I changed that to [AXIS_n]MAX_VELOCITY
* Jymmm actually can get 100IPM
hell, I can get 200 IPM
I'll be able to get 170 if I want, but I thikn that wouldn't be the greatest on a Bridgeport
Want to see a picture of my 240 ipm machine lymm ?
wb9mjn sure, why the fuck not
if I switch to Rutex drives, I'd be able to go *really* fast
Its smaller than a brigeport...
* Jymmm hates fucking computers - damn POS
all of em
I m trying to DCC to u...not working....
Or is it ?
Its a D and M 6...vocational ed machine from the 70 s...rebuilt it with servos, and a new X/Y linear bearing table..
wb9mjn go get an acnt on flickr.com
wb9mjn if you already have a yahoo acnt, you can just login
Not tonite...time for me to go to bed...
But another day...spent allot of time today figuring out DCC through DSL !
I like the idea of the servo loop going through the computer SWPadnos...makes it easier to work on...
That s why I chose the MotenC-100 when it came out over the Rutex...
emc is the only control software (available to hobbyists) I know of that does that
all the other stuff, including everything for Windows, is open loop - even if it's using servos and encoders
Do you need features not in EMC ?
there are some convenience features in DeskCNC and Mach
things like loading an image, and being to engrave it
that can be done on emc, but it's several "manual" steps
There is software to convert images into 3d engrave G codes, no ?
yes, but not "integrated" - no big deal for me, but less convenient
also, DeskCNC comes with a STL-> G-code converter
Anyway...time for bed...night all...
load up a model, select some machining methods, and start cutting
03paul_c * 10emc2-auto/wiki/ (30 files in 20 dirs): "Auto update wiki from a cron job. Thu Dec 1 05:30:01 GMT 2005 "
03swpadnos * 10emc2/configs/ (ppmc_io.hal ppmc_motion.hal): Changed a bunch of parameters to be loaded from appropriate [AXIS_n] sections of emc.ini. Added some comments on how to use the PPMC hal files.
Anybody awake ?
* alex_joni is barely awake
Morning Alex_joni....I m actually an hour late waking up...on vacation...
Aren t you in Europe ?
it's 14:30 over here, but I am kinda sick
so not fully awake :(
Sorry to hear that....I get a flu shot each year...but even so, had the flu last spring...
Its 06:30 here....
Figured out how to do DCC here...so can send pictures now !
what kind of pictures ? :)
I have a picture of the Ydrive on my machine, and the machine itself....which is no longer a virgin, it cut metal
for the first time last night!
put some on frappr ;)
Can you recieve DCC pictures ? Want to see them ?
[11:41:51] <alex_joni> http://www.frappr.com/emctheenhancedmachinecontroller
Not really done far enough for Frapper yet...
ok ;) dcc them over
Here comes the Ydrive...
That was the original DandM 6 drive plate for the NEMA 34 on a cheap asian table...I had to flatten it, they used
bar stock which was somewhat convex (bulged out on both sides)...
Luckily, the slot in the bar stock was the same size as the pilot boss on the servo motor...So just had to put in throuh
holes to secure the motor....
On the top is the 5K encoder...on a comon shaft as the timing pulley...Made that part from SDP timing pulley bar stock...
Turned down one end to 3/8 ths , and counter bored the other end for the table drive shaft, and left the pulley teeth
in the middle...That was the trickiest part of it...
In the end, the runout on the tip of the shaft was less than .001 ...
To tension the belt, I added the bearing pocket and idler bearing...
Here comes the whole machine...
The recent acquisition was that European (DIN 508) Tslot table..its hardened, and I could not get bolts in from
below...So, I had to have the threaded holes in each corner EDM counter bored, so I could drop in bolts from
Luckily, found a local small EDM shop whould could do it....West of Ohare Airport there are all sorts of small
machine shop companies....
The Tslots are 12 mm standard...
The X/Y table are linear bearing tables, same size and kind for each...Rather than mount in the usual way,
I bolted the two travellers together to make a saddle...
That was tricky, as it required removing the travellers, then reinstalling them..In order to get them back on I
had to use a piece of tubing to keep the first few bearing balls from poping out due to imperfect hand-fed initial
Each table was so close in allignment to its traveller, that the other traveller would not fit....
Anyway...its been a long project...I think I started in fall of 2003 ...
probably I'm too sick to get what you're talking about :)
but I'm happy for you that it works :P
and that you seem happy with it
Oh...the linear bearing tables have a part that moves, the traveller, which is driven by the ball screw...
yeah I know..
but I'm too dizzy to follow it more than that :)
Normally, these tables are mounted so that the traveller bolts to the table above....
don't waste your time trying to explain ;) it's not bad explaining..
But, in normal X/Y milling machine service the two travellers make a saddle...
it's just I've got a major headache, and my head is refusing to think ..
between the Y table, and the X table...ok...
Anyway, the original machine was all NEMA 34 steppers with R/L single step drives...I tore that all out and
replaced it with a servo system...Using Abdul's MotenC-100...
The R/L drive had a 2 foot by 6 inch heat sink ! The Gallil servo amps only need to be bolted to the panel for
Anyway...let you go get some recuperation there.....
and an early one at that
I mean, hi
Hi Les...want to see some pictures ?
Running mIRC now, and have my firewall set to let DCC through...
good day to make limit switch brackets...hi..
yeah. Snowing huh? may have to be up your way in a bit. the tools works have summoned.
Corp headquarters is in glenview, but I would prob stay in Schaumburg near where I lived
Are you getting that DCC request ?
Hmm...oh well...Its been working with several other guys on here...
I may have a problem on this end since I replaced the router
Mirc has a way of finding out the real IP address and using it...
i'm using that
Think the problem is there, now...supprised you did not at least get the request as previously..
Anyway...it cut metal for the first time last night...
Try to send me something les...
skunk, want to see some pictures of my machine ?
I may not have ports open in my router - what would I need open?
I accepted your send request Les...
Connection failed less...Looks like the problem is in the me to you direction...
1024 to 5000 is what they recommend skunkworks..
yeah. checking some things here
What is your internet connection there skunkworks ?
Ok...welp, i m on DSL, and had to switch to mIRC for DCC to work...it has the capability of ferreting out the
those ports should be open to my machine now - anything I need to do to mirc?
real IP address....
my mirc found mine ok
Ok...mIRC was defaulted in the way it needed to be...here goes...
everything looks fine
trying you again Les....
I m not familiar how cable does the IP translation....
Skunk timed out,,,trying it again...
nothing here yet
Les timed out...resending...
Cradek, alex_joni, SWPadnos_ and lymm all got these pictures...
oh sure - blame it on us ;)
There are lots of web resources out there to figure out DCC...just do a google skunk...
les - let me try to send you a picture
time for a new keyboard...or glasses
skunk, try to send me one too..sometimes the problem is unidirectional...
Apparently, identd protocoll needs to work too...as well as the ports for DCC...
Nothing happening on this end...
I had dcc server unchecked
trying to send
Got the request Les...
It bombed out...lost connection...
i'm googling this
tried to send wb9mjn a picture
going to restart here
wb9mjn - could you try to send me a picture one last time
rayh-away is now known as rayh
was reading up on the mirc site...it says dcc problems are usually caused by the sender
trying some stuff
the send request is done on the normal irc channel....that's why we see the request box
but if it cannot find both ip addresses or one is wrong it times out
this coulb be an issue for who use ipv6, maybe..
ipv4 should work
what we don't have is someone that has a dcc system known to work
so we can find who has the problem
i think who use a tunnel broker could have some issue
[13:05:20] <les_w> http://www.mirc.com/faq6.html#section6-9
anyone know ports used for dcc ?
Hi Jacky...1024 to 5000 ...
Attempting a send skunkworks...
Yes jacky...I think so...
going to portfoward these ports on my internal pc ..
There may be ports needed for identd, but I did not have to open them here...
Morning Ray..want to see some pictures ?
You bet but remember my dialup is s l o w.
Ok..send you the small one first...
I do need to test getting dcc past my firewall.
Did you get the request ray ?
les_w: try to take this file from me
Timed out ray..
I never even saw the offer.
15:04 DCC SEND request sent to les_w: tux.jpeg
Ok..probably the firewall zapped it...
I did see offers from swp yesterday
hit my ingnore list jacky
let me fix that
of course we are all registered to the server I suppose
you have a point there...Do not remember setting the password into mIRC..but it worked last night...
hmm I am set to accept *.jpg
your file was .jpg jacky?
jpeg, but ive jpg also
wait a sec
try the .jpg
try to send one to me
wb9mjn: I don't see the note "is identified to services"
timed out jacky
Might be the problem this morning...
Looking into it...ray...
les_w: strange ive not received any request
others are getting the request
that I send
The other software logged in automatically...this software is not set up right, or does not automattically login..
actualy ive 1024 to 5000 tcp ports forwarded to my local IP
wb9mjn: could you please try to send a file to me ?
Im logging out and back in...will try then...
Ok, Im back now...
so I might be ok
Tried all three of you there...
But nothing happening so far but the request...
looking at irssi config ..
Ive autoaccept dcc = false
send to me again wb9
but i should be able to manually accept using /dcc get nickname ..
There you go Les...
arg nothing....no accept request....but jacky's was fine
Still had a problem with the password...
why don't you guys use e-mail?
Will try again...
Les's file got throug...
SRI about that...new software is wierd, and has poor help file...
That is the Y drive Les...here comes the whole machine..
wb9mjn: connection refused. I am seeing the offer though.
other way now
The trick to the Y drive was making a custom pulley with the shaft for the encoder and a counter bore for the table
drive, all in one piece...from SDP timing belt pulley bar stock...You can see the tip of the part
what kind of mill is that?
coming out of the encoder....down inside it has pulley teeth...
It is a 1970's vocational ed machine...with allot of rework...hi..
Called a DandM 6 ...
Yes...but a very crappy implimentation...Nema34 steppers, R/L drive...
send timed out.
Very poor X/Y table and mounting to it...The bar in the Y drive was originally screwed to the Y axis with two 1/4x20 cap screws...
spaced about 1 inch appart...
15:24 DCC aborted receiving file Ydrive.jpg from wb9mjn
And it was convex by 1/8 an inch...So, i face milled it, so it would fit flat on the linear bearing table end cap..
websys: can retry ?
I did a few times Jacky...want me to start again ?
damn tab completation ;P
wb9mjn: yes, if you can
The servo motor pilot fit in the original belt slot....
ok ok, work
So, just screwed it on...used timing belt bar stock for the top pulley, the same size as the original pulley...
the prob is another.. 15:26 DCC can't create file /home/james/.irssi/dcc/Ydrive.jpg: No such
file or directory
ive not created that dir :)
Needed to add the idler bearing, as the belt would skip...
The encoder is 5000 line, or 20000 steps, directly connected to the table drive screw...
All screws are .2 inch pitch double nut 5/8 inch diameter ...
that is 10 times the encoder resolution I am using
The X and Y work well...need to do some work on the Z to get the encoder in the screw shaft, counter balance the wieght and
maybe a bigger motor...
I just use a bias voltage rather than counterbalance on z
The motor runs too hot for that....
when the machine is off the servo is shorted...that plus friction keeps it from falling
So, either counter balance, or get a bigger motor...haven't decided yet...
It has a brake for falling...it needs it on occaision too...
tiny amount of bias current
100 ma or so
ah I can send to wb9mjn
Because of the friction, it does not quite make it to nominal..and the current will ramp up..
Got some relay ladder logic here...
so this is step-servo?
I can receive from wb9mjn
All servos, with MotenC-100 / EMC implimentation...
ray let me try a send to you
15:34 DCC received file Ydrive.jpg [89kB] from wb9mjn in 00:00:06
Uses Galil amps...48 v, 10 amps, motors are 36 volt, amps regulate down by gain setting...
What box are you using? And what is your servo update period?
and what emc ;)
servo update period is 1 ms...with each axis set to a quarter ms...
Oops...Ray really does have a slow wire...
The PC is a 700 MHz Pentium...
Yup! And the wires are cold today.
have you tried to see how fast you can update the servos?
That makes those electrons flow fast Ray...metal has less resistance when its colder...
wb9mjn: what servo card are you using?
well no need to portforward any port then , I portforward 1024-5000 but dcc used port 57666
MotenC-100, which is a PCI card, its working well...
I switched my firestarter to permissive on outgoing. That seemed to do the job.
Abdul had is introductory pricing just about the time I was deciding if I was going to do a Rutex or EMC servo ..
The X table also has a custom pulley/encoder shaft on it...in the direction away from the picture view..
The encoders have the flex mounting...so, no shaft coupler needed..the custom puleys are accurate within .001 ... thanks to the
schaublin lathe I did them on...
so CYCLE_TIME is set to .00025 for each axis?
There is one shaft coupler between the X motor and and screw...also made custom mounting NEMA 23 adapters for the
That s right Les...
tried any faster? when does it lock up/go nuts?
This machine does not have allot of momentum...and PC is handling that speed...talked to a ham guy who works
for Bosch/Rexroth/whichever company they bought this week, who is local..and he said the whole machine should have
less than a msec sampling for the 3D ...
Nope...it does go nuts about 1 in 20 times when it jumps doing a home...
wb9mjn: You need a small bag of blue swarf to throw on that bed before pics.
I think the axis cycle times are non cumulative. If all are set to .00025 the whole 3 axes are done in that time
I always carry one for shows.
The galil amps do not shut off as cleanly as I would like when a limit is hit, and jump again when its released..
The .00025 is so I can do a rotary axis eventually...
The jump on powerup may be a command from EMC. If the shutdown rolled past commanded.
Need data on servo rates with other than an stg card
LOL, Ray...it actually has some aluminum swarf on it this morning...
To do the home, i run the axis into the limit, reset, then hit home....
no index pulse?
The problem is limit related...I do not have the limits wired into EMC, just the amps...If I push a limit when the axis is
stationary and in the middle of its travel, it will jump...
about .1 inches...
Yes, all the axiis have indexes...
To hit the same index each time, i start from the limit stop...
Does the limit kill power to the amp or do they have limit inputs?
I use a separate home switch that enables index detection
It homes on the index, reversing off the limit...The amps have limit inputs...
The Estop kills power to the amps, though...
yeah, my home switch is about 0.1 in before the limit switch
Wanted to bounce off the limits automatically, and find the index, but the jump makes that problematic...
_____ __ __ ____ ____
| ____| \/ |/ ___|___ \
| _| | |\/| | | __) |
| |___| | | | |___ / __/
as you see that ?
hmm well if you can bounce the switch...it isn't really a hard limit?
may mirc dont show ascii art good at all
The motor jumps when the limit is hit...then stops...Yea, and on occaision, it will oscillate..but that is rare..
heh...yeah. What is it?
It says EMC2...
For industrial stuff the limit switches have to be wired to the separate estop. Computer is not involved.
[14:02:09] <Jacky^> http://www.sp-plus.nl/home/re240284/penguin.htm
`''<<:: o. `\
correct showing depend on client I think
Computer is not involved in this...just amps and limit switches...But not an estop...
I think Galil did that to put some counter torque into the motors to help brake the axis when the limit is hit..
yeah. A little different here. Limit switch disables the amps, disconnects and shorts the servos thru a 10 ohm resistor
but my machine is so small it actually reverses the axis...
Yep, I remember Les...do not have a 500 lbs travelling down the ways here...
neither do I.
I have 1000!
Time to get some stuff done here....going into standbye...
yeah me too
THE ITALIAN WHO WENT TO NEW YORK
One day Ima go to New York to a bigga hotel. I go down to eat some breakfast. I tella the waitress I wanna two piss toast. She branga me only wona piss. I tella her I wanna two piss, she say go to the toilet. I say you no understand, I wanna two piss on my plate. She say you better not piss on the plate you sonna ma bitch! I don't even know the lady and she calla me sonna ma bitch. Later, I go to eat soma lunch at Emma's Restaurant, the waitress
Oops...just ignore me...
going to reboot for a test.
leodict translates via dict.leo.org
this could be a nice script
03swpadnos * 10emc2/configs/core_servo.hal: Changed PID limits to be [AXIS_n]MAX_VELOCITY, and Pgain to 100 - this should fix many of the following errors people were getting with jogs / g0 in emc2.
That is what I was looking for
it may not help, but then again it might ;)
are you using core_servo?
18:52 | <Q� #emc
18:52 | From: english
18:52 | To : italian
anyone can speak english please :P
Jacky^: a few of us can
thats a neat idea
Jacky^: not making much sense..
WWW::Babelfish: Language "english" is not available
how is possible ? O_O
that's a NICE error message :D
it was loaded .. uhm
or is it spelled "'merkin"?
that's the british spelling
uhmm i think under all, some key is needed..
like for google api
jepler: im testing some script http://scripts.irssi.org/
found someone very funny :P
Very simple script that sends incoming text to freetranslation.com for english->norwegian translation. May be modified to translate other languages.
norwegian and finland are lucky .. plate ready
this is interesting too Adds words from IRC to your tab-completion list,plus fixes typos
already work tab comp. for nick and dir/files
well, going to eat some pizza and beer for the birthday of my daughter
Jacky^ is now known as Jacky^party
hey anders, long time no see
in the mean time there's a ppmc driver for HAL ;)
Hi Jymmm, whats up ?
Imperator_: Nada, just working on some designs.
same here :-)
how's the mesa?
i study the FPGA beginners tutorial on the Xilinx side at teh moment
counter is running
and PWM stuff removed
maybe on the weekend i can test my 840D
imperator: you are modifying the mesa code ?
how fae have you got ?
encoders are working now
are you writing completely from scratch or based on the mesa code ?
that was not that difficult, because i had only to change the pinout
based on mesa code
i removed PWM and changed the counter pins
you want the torque command as 16-bit parallell for your DACs ?
torque or velosity
for teh first velosity, the rest is done by the amp
does the dac then just have a 16-bit mux for the different axes ?
so you need 2-3 bits for the mux also
two adress lines and 16 bit data
only 4-axis machining then...
but our hardware can expanded to 8
on the counter board, there are six inputs at the moment
but i think i remove the two optional inputs
we are using that lines for the limit and ref switches
you have rs422 decoder on the counter board ?
did you make your own optoisolator board also or are you using the mesa optoisolatorboard for I/O ?
we have a 16 Input and 8 output board
it is at the moment directly wired to one of the 50pin connectors. we want to use there a bus, so that you can use many of that io boards
yep, sounds good, again data+address format.
[19:29:29] <Imperator_> http://www.peters-cnc-ecke.de/forumupload/uploadFiles/3686_115677871897_16i8o.jpg
is that link working ?
thats the I/O card ?
thats the io board, but the Phoenix connectors are missing on that picture
[19:31:26] <Imperator_> http://www.peters-cnc-ecke.de/forumupload/uploadFiles/1298_112276751785_IMG_6418_klein.jpg
nice work. do you work at a university or school where you can do this ?
really nice photos :)
have worked in a university up to this summer
but that is done at home
do you have pics of the ad board and counter board also ?
ideal to place it into a web shop
that picture screams EBAY all over
on the picture with al that components, that are two boards, they are cutted now so that are four
too expensive for ebay
that only depends on your source for the material
I hope you are going to publish the plans !
when it is working
no one sees them for review ?
Nice looking stuff martin.
Hi rayh, thaks
have bought components for 400 EUR
are you making 20 boards ? :)
not enough for 20 Boards, the DAC is 50 EUR by itself
oh, sorry I did not realize that. sounds a bit pricey, much cheaper in volume I guess
maybe, but we want real industrial quality
if it is running well, then we have to think on how to test it !!
what machine(s) are you planning on driving ?
with a polymer concret frame
test the whole thing or just boards separately ?
just the boards
heh, DIY VMC
poly concrete frame cnc machine, or what were you saying ?
Poly concrete has better dampening characteristics than even cast iron...
[19:46:57] <Imperator_> http://www.peters-cnc-ecke.de/forumupload/uploadFiles/1298_110779439075_IMG_1642.JPG
[19:47:10] <Imperator_> http://www.peters-cnc-ecke.de/forumupload/uploadFiles/1298_110779439075_IMG_1668.JPG
[19:47:26] <Imperator_> http://www.peters-cnc-ecke.de/forumupload/uploadFiles/1298_110779439075_IMG_1695.JPG
Do they use iron rich granite for the agregate, like on railroad tracks ?
with a bit colour:
[19:48:02] <Imperator_> http://www.peters-cnc-ecke.de/forumupload/uploadFiles/1298_110779439075_IMG_1680.JPG
was not enough
what epoxy are you using ?
Is that high temp epoxy, to give enough open time to do the large casting ?
it is silicat
Yes, but what is the matrix ?
normal epoxi with a hardener that has 2h open time
Ok...looks red like West Systems stuff...
dont know the real english names for that stuff
how do you think the thermal properties will scale to something bigger ??
i dont whink that i will get some thermal problems, because there are only 8% epoxi in there
so you are actually going to build a machine from scratch using this stuff ??
im busy with the electronic at the moment
yah, gotta start somewhere
do you have motors and amps already ?
Siemens Simodrive, yes
and some Alsthom DC Servos
Alsthom Parvex servos
ok, my latest plan is to use the IR IRMCK201 to build an AC servodrive for the surpluscenter motors
thats a heavy thing
yeah but using the irmck takes care of the motor control algorithm
so I don't have to do it
motorcontroler != servocontroler
well, the servodrive would accept an analog torque command, use the iramx integrated powerstage, have the current feedback etc.
but the most complex things are done in the irmck
so hopefully it won't be that hard to get it to work
03alex_joni * 10emc2/tcl/bin/ (8 files): some efforts toemc2ify the tcl bins
03alex_joni * 10emc2/tcl/bin/genedit.tcl: changed that the editor pops up, only when run outside of tkemc, when run from tkemc the proper function gets called anyways
Jacky^party is now known as Jacky^
SWPadnos_ is now known as SWP_Away
come sta male mo
where is everybody?
03alex_joni * 10emc2/src/emc/iotask/ioControl.cc: fixed a stupid bug I just stumbled upon. a fp was closed that wasn't open
im watching TV
a political program :/ not much interesting
| |__ (_)
| '_ \ | |
| | | || |
_ \ \
(_) _____ | |
_ |_____|| |
(_) | |
scusa ma davvero non mi vedi in linea in msn'
___ | | __
/ _ \ | |/ /
| (_) || <