Merry Christmas les_w
thank you same for you
ok. Relaxing mostly. Also pondering how to keep exponential business growth exponetial.
clone yourself every week
also have some real estate dealings to do.
remember, sales are the key.
buying another swamp?
Well family has some extra land here. I am going to try to move it. I think I can get a million or so. We'll see.
million is not so much anymore.
million==middle class stuff now.
land is always a good investment. they stopped making it AGES ago.
I am more concerned with good investments for company profits.
rainy day money
ive been looking over the regulations for building this nitrogen generator
real interest rate of most investments is negative
* Jymmm refers robin_sz to Active Volcano Realty
robin, call it an oxygen generator. Regs will be easier
man, if you want to build apressure vessel bigger than a coke can, you need so many inspections ...
* fenn refers Jymmm to floatinc.com
nah, its ANY pressure vessel bigger than 100 bar litres ...
only a few atmospheres right?
not much to that here
so Im going to use some 4mm wall, 4000 psi capable proper pipe
some 10mm thick end flanges and caps
yeah...it's nothing compared to typical hydraulics
I'll water test it to 30 bar or so
if its fine at 500 psi, im happy.
If the reaction is reasonably adiabatic and fast couldn't you make some kind of diesel engine thing to do it?
I suspect you possibly could ...
hmm.. 100 bar-liters is 10kilojoules
that's a good deal of energy
I wish we had diesels in our cars like you do
maybe the 10:1 compression requirememtn wouldnt leave much room for carbon
we waste a lot of fuel in our cars
my tractor is 24:1 I think
fenn: yes, but its not exactly a difficult engineering task ... 10 bar is not high pressure. allsorts of badly welded, half coroded tanks work as 10 bar receivers every day ...
done right, a simple 1.5m long, 200mm dia pipe with end flanges will be super reliable
I had a leak in a shop air tank...made a nice weep valve...then it corroded so much it sealed.
most 10 bar recievers are 3mm or less, a 10mm plate is going to be overkill
i'm just saying there's a reason for the regs
oh, i know
not saying that they're well implemented
I agree with them too ...
except actually complying is impossible for one-off jobs
send in samples, have a random visit during manufacure ...
oh, you're too late
come back next time
yeah, but I bet the expense is man thousands
so stuff em.
who checks this stuff out anyway?
doit right, test it hard.
i mean who actually comes and says "hey that air tank isn't certified!"
well, you are supposed to appoint an authorised test house .. theres a list in the regs
err, the local factory inspector I think ...
hum.. never owned a factory
can any old ASTM certified welder just say "she's good to go"
I'll make sure its inside a big box marked "experimental"
of course we all use chinese pressure vessels that are basically exempt
danger do not store faulty air tank in remaining building
fenn: saldy not ... its ASTM certified welder .. PLUS all the other crap
stuff em, I'll just build it to regs for 3 times the pressure, test it way high and pray.
I over engineer everything pretty much by 2 or 3 times
I cant see any way to do "one offs" in the regs .. no wonder innovation is stuffed over here
les_w is that repetition or magnitude =)
one or the other
fenn that was pretty good =)
usually both, by the end
* Jymmm ROTFLMAO
fenn's law: everything will take 5 times longer than you plan for it
fenn's corollary: don't take fenn's law into account when planning schedules
wow, you plan things?
I just fight fires
plan to fight fires.
I went out to the factory today, its SO clean after the christmas tidy up
you'll have to fix that
my shop is too clean as well
went out to bring back bits of the spare laser for rework :)
kitchen table is covered in "things" now
I worked a little today too
yeah .. gentle fun stuff
The corporate managers are still using up their accrued holiday time
sio I get a break
gant charts start back up jan 2
yeah. pretty much. only I get a lot more money for it now.
Owning a business is good.
no wait ...
owning a profitable business ..
Well keeping that exponential growth without tripping over your ass in the process
that's what I think about.
the local competitors have felt the pain ... taken a good chunk out of a couple of them
I just think less engineer...more salesman
yeah. sales is everyting
the factory just has to deliver the goods ...
thats the EASY bit
too many people get it arse about face ...
think having the gear is enough ...
fine with me...cause I won't
prob 80% of my time is sell....20% engineer
doesn't that strike you as odd?
yep. its a bit like a swan ... you just present the customer with good service, super quick, great product. underneath, the factory is thrashing all hell together to make it happen, but they never see it.
heh it's called "object oriented design".. separation of interface from implementation
fenn: not odd for me. I have engineered a very long time. Most of what I see has been done by me or someone else. I just have to sell it.
we even sell stuf we can't make ... just buy it in underneath
yeek google understands acronyms now
one day, when volumes are up, we will make in house
I'm just still looking for cars. Need to make the statement when I drive up to a client:
"I have a shitload of money and my fees are very high."
heh just say it to them and drive whatever you want
drive up in a beater and it says "failure"
same with hairdresser car.
what's wrong with hairdresser's car?
"follow my advice and you'll do as well as me"
== inherited wealthand useless
heh that describes me pretty well
(drives a hairdresser car too)
when selling product
arriving in a merc is BAD
if you are trying to seel on "good value"
All cars I have looked at have been declared hairdresser cars by robin
I am getting worried...
been out in the woods too long?
les_w: get a mini-pickup like robin's
nice 911, S class merc, the Bentley ...
911 isn't a hairdresser's car?
Looking at infiniti q45 now as you know
the boxster is.
big v8 sled
ahh, get a Bentley :)
I'm too poor.
those infinitis are real oldman cars
new bentley looks just like a cadillac
too many gadgets
gps is ok
but I don't really want to talk to it
chilled seats are too far
active suspension I could do without too
the bentley has 550 bhp ...
les get a rocket sled
6 litre, W12
infiniti q I AM LOOKING AT HAS 340
too much HP
130 or so would be plenty
[02:19:59] <robin_sz> http://www.europeancarweb.com/firstlook/0402ec_bentley/
looks nice to me
now, that, is a statement. it says "you probably cannot afford me, but feel free to try"
these days I want a car that basically fells like a rolling ingot
I am often dissapointed by handling as compared to my old BMW
well, beemers handle.
older lightweights especially so
thats the homologation special .. all allloy body
Well I have a 325i.
it's very good.
yeah, good car
ah well ;)
actually, it is very good
had one many years ago
Bought it new...but it is very old now.
Still runs ok
maintain em,. they run forever
have a 318 touring for the family bus, had it .. ages.
I just feel I need something else for 2000 km drives
wow that bentley is 5300 lbs
will probably keep the 325
I would have preferred the 318 if I had to pay thr petrol prices you do
325 gets 25 MPG at 100 mph
the q45 I am looking at only gets 18
but our gas prices are still really cheap
we bitch about it... but the cost is really low
les_w: I bet no car gets 25 mpg at 100mph
100 kph, sure!
ok, bed time
I think it does...gets about 35 at 70
usually mileage goes up with speed in sports cars
les_w: on a road or on a treadmill?
to a point
low drag coefficient...high compression ratio
damn, no hot chicks ... it will have ot be the wife .. again, sigh.
ok, night ...
robin_sz: don't say that too loud
robin_sz: asking for trouble
* robin_sz juggles with fire
I will admit the mileage meter might be wrong
but it is awfully good
sure is nice to have good mileage
lower drag and friction enough and why not?
with my "new" car I can usually get 300 miles on a tank which is new for me
I only get 23 in my daily driving but I can't complain, it's a very powerful car
basically an audi A3
wind drag goes up with the square of the speed, right?
Roughly but not exactly
it's no wonder we had 55mph speed limits in the 70s
with the 3 speeds everyone drove then I bet it made a huge difference
Well, as a pilot I know that wind drag (induced drag) can be very low compared to rolling friction
cars with very low drag coeficients (below 0.3) can do very well at hight speeds
german cars usually do
I can watch my mpg go down as I go over 60mph
new gti right?
car has what seems to be a very accurate trip computer (I tested it over an entire tank)
vr-6 engine with the 6 speed
I want to look at the body shape
kind of a boxy hatchback
03cradek 07simple_tp * 10emc2/src/emc/kinematics/ (tc.c tc.h tp.c tp.h):
add some documentation and remove most of the irritating
pseudo-object-oriented stuff that obscures the code.
[02:42:09] <les_w> http://www.canadiandriver.com/articles/rp/03gtivr6.htm
cute isn't it?
totally a hairdresser's car :)
* cradek makes a rude gesture at fenn
just stating the obvious, since robin's not here
ok. makes sense. The hatchback shape is a bit utility over extreme autobahn speeds. It makes sense that high speed mileage should drop...especiallty in the US
City 11.1 L/100km (25 mpg) Hwy 7.3 l/100 km (39 mpg)
now that's interesting
the liter/km measures are about what I expect
but the mpg conversions are way off
I think my "good deal" on a rackmount PC wasn't a good deal after all
might be an error...didn't check them
I wonder if europe/us have different standards for measuring that
it's an industrial SBC in there
I know our procedures are totally bogus
that was custom ordered with no vga or ps2 ports
but it's got three PCI slots
on a backplane
and it has header connectors for ps2 mouse and keyboard
so I've been hacking connectors to be able to plug in a normal mouse and keyboard
which finally worked
now the BIOS has a passwd on it
but otherwise it'll be perfect for EMC!
I just bought radio sack out of there serial mice ;) (still have ome old computers in service)
djb_rh: there's typically a jumper to clear the bios.
it didn't work
how long should it sit with the jumper off?
yeah what if someone broke into your server room
djb_rh: just a minute
I left it off like 15 seconds
or turn it on with the jumper off
just dink with it
or remove the battery
I did turn it on with the jumper off
is the jumper marked?
took me forever to find a manual for it, but I found one
I found a password cracker for the Award BIOS
I just have to find something to write a floppy!
seems none of my computers have that ability any more
though this one does have a floppy drive
just need a pci video adaptor and a usb card ;)
it has a floppy but no mouse or keyboard?
so I'm hopeful it'll boot a DOS floppy with a DOS password cracker
les_w: considering how much road grime gets pulled up onto the window I'm not surprised to hear that the aerodynamics are probably a compromise
fenn: stupid, huh?
it was set up for serial console (including the BIOS, though I couldn't get the BIOS to come up on the serial console)
I was getting serial output at the BIOS level
what's stupid is these systems came with a VGA card and serial console port, but no keyboard port at all (including no USB port)
and with only three PCI slots, I need one for VGA, one for Parallel (no parallel port) and one for network
which leaves none for USB
it does have a USB header connector on it with two ports that appears to exist, but I didn't have anything to plug into a connector like that to give me USB and figured hacking the ps2 headers was easier (and it was)
this was some special box that was supposed to just be a web content caching server
Cradek: cars are always a compromise. but looking at the review the gti seems a nice ride. Expensive though.
les_w: of course I didn't pay anything like that! I'm not rich.
I am doing the same thing...couple year old CPO car. It's the best deal.
les_w: it's really fun and practical too. four medium-sized people can ride in it (for short trips) and the hatchback is surprisingly utile (I hadn't had one before)
I had no idea it retailed at $30k
I think that's crazy
that is canadian
still, it is expensive
but nice ride.
it's a lot of car in a small package, which is what I like
I wasn't at all happy with my lumbering mustang
can be modded up yo an outrageous race car too.
it already spins the tires in the first three gears... I don't know what you could possibly add
it's a trick to drive it. I'm used to being able to punch it to go faster, but it doesn't work that way in this car
(the traction control kicks in)
Friend had a GTI...It wasn't slow
it's kind of a sleeper I think
it can be mistaken for an econobox if you don't know what it is
I'm too old (and smart) to race on the streets much, but a couple people have made that mistake
I'm in a different market segment now...I need old man car.
old men don't drive hatchbacks, as a rule
fenn: only hairdressers?
heh i drive a hatchback
I say drive whatever seems fun and practical
I don't give two shits what anyone else thinks
think a hatchback is the best all around compromise
I need heavy car. Poor handling. Cushy seats. V-8.
fenn: I sort of miss having a trunk where I can throw stuff and it's out of sight
no not a caddy
cradek: spare wheel well
les_w: full-size ford bronco
les_w: i was talking about the jag
fenn: it's got a computer in it already (homemade mp3 jukebox)
les_w: radial engine, like old airplanes
You see, I am too old to have fun. I don't deserve a BMW really. Need a Crown Vic or something.
damn... out of memory again *sigh*
but I get this I think...
[03:01:59] <les_w> http://www.burtmanindustries.com/q45/q45.html
way old man.
I can tool around at 25 mph and irritate people.
that middle part of the dash board looks like one saggy ... forget it
oh, it melted.
it was a NURBS accident.
what's with the cheesy wood grain stuff? I hope that's an option
The cheesy wood grean stuff is called birds-eye maple.
It's from trees...
trees?? is that a new kind of plastic?
it's an old fart thing...we like tree stuff
i like the new compressed bamboo stuff
it's kinda expensive though
ok....seriously nurbed out dash....old hippie genuine organic wood....lazyboy style seats....
just like a real 18k gold watch looks cheesy after seeing so many cheap gold-plated ripoffs, I think real wood in a car looks cheesy after seeing so much corny plastic "wood"
03rayhenry * 10emc2/directory.map: added config subdirectories.
I did say old man car did I not?
(it never crossed my mind that it might actually be wood)
the bad photos don't help either I guess
Actually as a woodworking and automotive oem type I can tell you that real wood has to be modified a good bit to survive auto interior extremes
anyway it is reliable. it is heavy. It is fairly unpopular. It gets crummy mileage. Perfect cpo buy.
But I still love Audis...
* Jymmm doesn't know wether to shoot les or the audi and put us out of our misery =)
I really do research stuff like this too much
les_w Hey, it's a big purchase, I should have when I bought my SUV
I think I would have bought a pickup instead.
I am a bit ana....I mean nit picking.
I like it, EXCEPT for the milages
Well it is a lot of money. Those rides ain't cheap.
Well, yeah. but I never expected the gas prices to raise as much as they have.
I just wanted a car I fit into
Gas I cheap. It's the money. It ain't worth nothin.
Jymmm: the new VW beetle has more head and legroom than any SUV - it's amazing
i.e. $2 gas is as cheap as water.
Jymmm: too bad the car is stupid looking
I had a 67 bettle.
if they had heat, I'd probably have one today
cradek heh, too bad.
they had heat...kinda
cradek they had "heat"
I rode to high school with a friend and helped him keep the inside of the windows scraped
that's not my idea of heat
ideally, the interior of the car would get above freezing
Mine heated fine. I lived in Florida.
that must be the secret
then what did those two levers actually do?
Reminds me of the heat in small aircraft
if you pumped them up and down enough, your arm would get a little warmer
but there are limits
03rayhenry * 10documents/lyx/Hal_Introduction.lyx: added emc modules first draft
well maybe his was just broken - it was a real pile
Funny I could get a 60 F temp rise out of a horizontally opposed air cooled Lycoming IO 360 180 HP
in an aircraft
which was good in chicago in the winter
but bug engines was in the back, not the front like a plane.
sometimes in NE a 60F temp rise doesn't put you above freezing...
(but not often)
I have had some cold rides in the airplane for sure
the cold air is more like water
plane handles well
feels more like a car
I desperately want a non-rt sim for emc2
I thought all the sim was non rt
emc1 I guess
but I want to work in emc2
well, I will too later
when I can
what do you mean by non real time simulator
skunkworks: all of emc should run in userspace
skunkworks: otherwise, you can't attach the debugger, you can barely get a printf out (has to go through syslog!)
hmmm - little over my head.
well, bedtime for me...I herebye declare tommorow a work day. Enough vacation. I'm bored.
does anyone know what triggers the cvs emails and irc message?
where is it?
[04:02:19] <Jymmm> http://cia.navi.cx/stats/project/gaim
[04:03:08] <Jymmm> http://cia.navi.cx/doc
cradek is that what you were asking about?
well, that's my entire knowledge of it =)
I was just curious myself one day, and got nosey
* Jymmm needs a computer with 8GB ram, and 8MB L2 cache
cradek: Now I have a question... what is PID ? (the drain bamage version)
Jymmm: you know how when you want to accel your car to exactly 60mph, you push the gas pretty far, then when you get up close to 60, then you start letting off, and you end up right at 60?
isnt' PID an acronym?
it stands for proportional integral derivative
proportional = how far you push the gas when you want to get moving fast
Oh, I thought it was three things that "combined" together =)
yes, it is
each one is multiplied by a constant and then they are added to get the output
that's the tuning
you should ask someone else who's an expert on this
and this is a stepper or servo thing?
(or any thing =)
used mostly for servo, since they take an analog voltage (like how hard you push the gas)
imagine you're as good as possible at getting your car up to exactly 60mph and holding it there
within the car's constraints (max accel)
you're a perfectly tuned pid loop
so as fast as you can in the curves without crashing =)
ever ride with a new driver and they can't keep a constant speed, so the alternately speed up and slow down?
no, no steering involved here
just the gas and brake
I'm talking NASCAR here =)
a badly tuned pid loop will get up to speed/position slowly, or oscillate around the desired speed/position etc
so PID has nothing todo with going around a corner, just pure straight line
well it has to do with following the desired position/vel/accel
you can sure use that to go around a corner
but it's not about choosing the desired path
Yeah, but that's another thing that deals with that moreso than PID itself.
I don't understand the question
TP is about choosing the desired path
PID is about following it with motors
yeah, ok. that answered my question =)
PID is purely performance
if you're interested in the math in pid, ask google or les/ray/whoever
it's pretty fascinating
oh no, I just wanted to comprehend PID
used in all sorts of closed-loop feedback circuits - for example probably your car's cruise control
* Jymmm is still tryign to figure out what LOGIN: is for =)
oh, ty =)
well, cracked my BIOS problem
what was the secret?
got a new mobo?
needed a bios cracker
www.11a.nu has one
cracker for what?
it brute force found the BIOS password
why would you need that? just remove the battery
left it disconnected for 15 minutes, anyway
There is not one mobo I have ever came across that removing the battery and/or moving the jumper that didn't clear the CMOS
that's nice to know
did you unplug the system?
but I did that and it didn't work on this one. :)
did you change the jumper?
and the correct jumper?
tried powering up with the jumper out
with the power removed?
well, tried removing power, tried adding power and powering up with jumper out
tried all sorts of things
but in the end, the cracker was pretty easy
you tried powering up with the power removed?
powering up with the humper removed
you tried powering up with the power removed and the jumper in the RESET position?
yeah, some boards need that
what system/mobo is this?
some crazy industrial PC
03cradek 07simple_tp * 10emc2/src/emc/kinematics/ (tc.c tc.h tp.c tp.h): more cleaning
one of those SBCs that you put in a backplane
dude, I've build a few thousand PCs in my day
and it was removed from the backplane?
I tried the tricks
no, but there was no power on the backplane
the password cracker just runs from a DOS boot disk
took about 15 seconds
at least your on your way
I need to find one of those little USB dongle things made to plug into a header plug on a mobo with two ports on it
I've seen them on some systems, but I don't have any hanging around...all the mobos I have have the USB ports directly on the PCB
hopefully tomorrow I'll install BDI on it
I had an IPCop install CD laying around, which installed fine
03cradek * 10CVSROOT/loginfo: unified diff format
03paul_c * 10emc2-auto/wiki/ (9 files in 6 dirs): "Auto update wiki from a cron job. Wed Dec 28 05:30:02 GMT 2005 "
odd time of year for thunder
On vacation today, tomorrow...
I might have melted my Z motor....
Going to see if it overheats quickly sometime today...it got extremely hot yesterday....
Really need to offset some of the milling head weight...
Got the Y axis retuned...its now holding a few tenths, like the x ...
Brushed DC motors have lots a torque but no stamina....thats going to be a problem on the Z axis aparently...
Cannot use them to counterbalance stuff, just move stuff....
It was well above water boiling temp yesterday, holding at 0 for a few hours while I was working on something...started to smell....cooled it down with
took over an hour to be cool enough to touch...
many brushed dc motors have a higher cont current rating if an external fan is used
Suprized the encoder housing did not melt...but that appears OK...
The ones I used to sell did
I probably should do a heat sink...ends of motor were quite hot...where the heat transfers fromt he armature....
I was deciding yeaterday whether to wire the themistor on the new spindle motor
but ap engineering for colombo said the motor would be fried by the time the sensor went opn
I moved my spindle motor up on top of the milling head from inside...so far so good on that...but have not given it a long run yet...
so I will just detect fan blade motion somehow
I think a box fan would work wonders for you
The metal cutting I did with it , worked out allot better than before, however....ventalation holes can breath now....and its well out of the way of any swarf/dust...
Could easily tap off the brake power to run a 24 v fan...
When the Brake is released, the fan would run....
Need a bigger motor too...they made one bigger in this series too, if I can find one...
yeah the new spindle uses a 240v single phase box fan
must have verified fan function in the estop chain
The spindle here is a 90 VDC SCR drive...with internal fan...
Do you have an switch in your spindle drive for fan pressure ?
or rotation ?
well, this is a $2200 motor. It would get fried if the fan failed. And yeah, have plenty of logic stuff to detect that
some kind of vane switch
or a hall switch and a tiny bit of metal on fan blades
The problem is when the axis is static...when its runing it gets warm, but never really hot...
Holding the weight against gravity though, when the machine is turned off the axis will drop about .003 inches, is the problem...a brushless DC motor could probably do that
right even running slowly gets some airflow between rotor and stator
all day, as the windings are heat sinked to the motor housing, but a brushed DC motor the windings are insulated by air inside on the rotor...
well pehaps a spring is in order to cunteract gravity
What do you know about the gas springs ? Are they constant force?
A metal spring has the reverse function I need...low force at low compression...
what is your z travel?
About 7 inches...
you can make variable sprig constant by tying different springs together
From a non-kludge enginering approach, I think something like a clock-work constant force spring on the screw would be neet, but still leaves extra load on the nut...
doe not have to be exactly constant
just take some load off
Well constant force springs are a vailable
but typically have low lifetime
negator they are called
I think somebody on here mentioned using an air-cylinder with a pressure regulator...
lots of friction though
But continuous air pressure supply is something I am trying to avoid...
what is the mass of the Z axis?
I saw a high end bench top machine with a weight and pulley....
Would need uranium for the weight here though, so as to be small enough....
well weights work, but add mass.
gearing etc can minimize thar
But adds travel...
difference between first and second moments
big weight with block and tacke
Welp...gotta think on what I m going to do....
in the limit high ratios would give force with zero apparent mass
Could you explain that some more ?
yeah...I have been off work, but am getting very bored.
the weight thing?
High ratios ?
Do you mean approximatly zero mass ?
imagine a see saw
equal lengths, equal mass
our big nc mill has hydrolic cylinders to counter-balance the head
then imagine a see saw still balanced but with a very heavy weight on a short arm on one side
moment of inertia is weight times distance squared
but see saw torque is just weight times distance
so a heavy weight near the fulcrum supplies force to the other side with little apparent inertia
pretty cool les.
well way back when I designed audio gear it was a big issue in things like recordplayer tone arms
But, to do that the mass would be bigger than the head, althought it might be implemented in the space that is available...
we would use lead, tungsten catbide, and uranium as close as possible to the arm fulcrum
for High end stuff
Uranium is startingly more heavy than lead....
Its quite an experience the first time one lifts a piece...
I keep hearing scotty "you can't defy the laws of physics" ;0
Well consider a 4:1 block and tackle and a lead weight 4 times the mass of the head
That is outside my weight limits here...a 200 lb counterwieght.....
well even 2:1 would work
Need a spring with a negative K...hi..
snap disks have neg k
How about thermally controlling a snap disk ?
well a whole line of motor safety switches use a thermal snap disk
but you just need a constant force
have room for a weather ballon? ha
That s a good one...
A negative and a positive could be turned to a constant...maybe...
That would be a good project for Smalley....
but the constant is zero
Not if the force is in the same direction....
A coil spring pushing a stack of snap disks...
Force could be made constant, but the displacement would be wierd...
kinda like a current source...
how big is that servo?
Or a coil compression spring on one side of the head, and the snap ring stack on the other side...forces are now in parallel...
Its about 5 inches, but 2 inches in diameter...
Pretty standard small servo for the 1970's...
and it's a 50 lb axis weight?
Yep, with a .2 pitch screw, and a 2:1 pulley ...
It has a 360 oz in peak....
With like a 40 oz in continuous...
Not sure o nthe continuous...
ok sounds a little low
what is the max rpm?
Let me go get the specs...
03rayhenry * 10emc2/configs/univstep/ (7 files): First commit of universal stepper config set.README
depending on rpm and rotor inertia more than 2:1 might be called for
03rayhenry * 10emc2/configs/univstep/README: First commit of universal stepper config set.
Ok, it can run up to 2700 RPM with the 36 volt output from the amp...
or 675 RPM at the screw, if the pulley was changed to 4:1 ...
or 168 in/min on the axis...
Is that what you were thinking ? Changing to 4:1 ?
I get 135
yeah you could change to a higher ratio IF the rotor inertia is low enough and if the max speed is enough for you
that s right...
have to be careful though
reflected rotor inertia i sthe square of the reduction ratio
gets out of hand quick
Right now, I run the max at 240 in/min, and the motor is capable of 270 in/min with the 2:1 gearing...
but I'll bet that servo has really low inertia due to the small diameter
rotor inertia is .0065 oz-in secsquared ...
well is 135 enough or too slow?
.0065 yeah that is rally low
let's see 4:1 reflected inertia would be 0.104 oz in sec^2
so for 5 tpi screws
that would be equivalent of about 200 kG
bit of a mismatch...motor would be the limiting factor in accel rather than head mass
depends on how much accel you need
what is your .ini set at?
Its set to 12.0 / 16.0 default/max in/sec^2 ...
that's only 0.04g
so 4:1 might work for you if the axix speed is enough
If I went up to 48 volts on the amp gain, that would get 3609 RPM, or about 180 in/min ...
Jl/Jm would be only about 0.1....the motor would not even see the mass of the rest of the machanism
it will work
3:1, 48 volt, would be back right at 240 .....
you really don't have to run at matched inertias
yeah 3:1 gives 9/16 the reflected inertia
but enough torque?
And the stall torque would be 3/2 improved, to 126 oz in at the screw, from 84 ....
Right now, its well over the 42 oz in at the motor apparently...
well 3 would help, 4 is about the limit
bracketed in nicely
just do a motioneering run to verify accel performance
Might be a plan...
yeah might be the easiest option of all
short of just a bigger motor
If I could get some counter force as well, that would do it ...
How would you do the motioneering ?
and remember little box fans....typically double the stall torque rating
motioneering is a mechanical modelling software
it's free too
I have more of these motors, so if a winding is blown, its OK...I can replace the motor...
timing belt reduction?
Just do not have the bigger motors...
I think you would be ok with 135 ipm...unless you cut a lot of balsa wood!
for any metal...that would be fine
can't cut at 2 inches a second anyway
seems fine for rapids
At 48 volts it would be 180, too...
It easilly drilled 7/8 s inch diameter plunges with a forstener bit in spruce the other day...
using 6 mm plunges down downto 30 mm ...
well if the amp can handle it go forthand find a 35v transformer or so
Forgot to set the plunge speed and it did not care...
ha I think I am the emc hole drilling winner though
That wood that sign makers use for sand-blasted signs....garbaged picked some...made a collet rack...
having drilled approx. 80,000
western red cedar
(the sign wood)
western red cedar? for?
Its not red though...
sandblasted wood signs. I't pretty much the standard but is getting horribly expensive
Motor did not even get hot doing that for an hour...
cause it was moving
while stalled at 0, it was probably up around 400 degrees F ...
chestnuts roasting over an smoking motor....
for 2 hours...yea, it was toasty ...
Its no wonder the world is changing over to brushless DC motors...
I still use brush in the router though
speaking of brushless DC motors...looked at a Prius yet, Les ??
I kinda like it
Its a cute car.....if you are into cute ...
I'm gonna be into a lot of drives from here to glenview ITW corp headquarters
Seat will have to fit ...and the Prius does not have a fully adjustable seat yet ...
I'm looking at big sled sedans
But, hell at 70 MPH it ll do better than 45 MPG ,, and those Indiana boys wont let you over 65 anyway...
On I 65 ...
heh...I had to do the run down to indy on 65 a lot to get to our powder coating division
I knew where the traps were
so I got away with 125 mph late at night in the Audi
trap at lafayette
and at...oh forgot...that show place near the lake
I got passed by a guy doing 120 or so in downtown Indy, once...had to nail the brakes down for 80, to give him room to fit between me and the car in front that was doing 65 , lane narrowing...
where the concerts are
Woke everybody in the car up instantly!
That and the aircraft engine noise swooshing by...
Later for the indy trip I would fly a little single engine plane when possible
That was after midnight too...
Well I do not drive at very high speeds anymore
but might like to go 85 or so if it's a safe road
I keep it under 80 during rush hour, and 75 when the cops are not busy, and 65 outside of the Chicagoland area where the cops are not interested in maximizing througput...
ha I kept it under about 5 in rush hour
I am about a 150,000 mile veteran of the eisenhower.
The Micro-Coax sales girl (from Pensylvania) could not get used to driving in Chicagoland due to the speed....
She moved back....
I'll tell you a worse city
They just reopened I 355 with 3 lanes of automatic tolling....
Been to Colorado...worse drivers I ve ever seen ....
So, what happens on I 355, left lane loafters get in the left lane and pace traffic to the right at 50 mph....with 1/2 mile empty lane in front of them ...
Never been to Atlanta...
Guess I'll see that soon....need to be up there. Living in an extended stay place I guess
Yep, that s been the big project this year...Get all the 2 lane automatic tolling switched over to 3 lanes...
Oh, and it cost twice as much for manual tolls now...
Yep, $1.00 toll manually...
so visitors are sol huh?
Pretty much, although they use the same transponders as many other states, and have cross-state billing agreements...
Why they did not build the system 3 lanes in the first place I do not know !
They call it ORT (open road tolling) ...
ran out of common sense
Yea, they spent a year and half building the 2 lane system, then another year of backups converting it to 3 ...
I don't have a transponder...we don't have toll roads
And this was all over, not just I 355 ....
Get your reciepts then, it might add up and you could bill ITW...
Really I need to just get out and show my face at many divisions
The real problem is not people driving slow, its people driving slow with 1/2 car length spacing...that causes more accidents !
to keep the development projects going
Right lane is almost always like that, 300 hard solid walls of cars...and the only way to get off the highway is through them....
I have decided it's time to grow up and get rich. Right now!
I'll prob stay in schaumburg
just used to the area
I 355 here you come...
how's construction right now? any bad spots?
still a mess around Gary?
I 88 west from Naperville is still bad....but otherwise they are supposed to have all the Chicagoland construction done for the year...
Not sure about Gary, do not get down there...
Cold and snowy this year too right?
Warmed up last several days,,,yesterday was mid 40 s....
I need to go up to miller too
Before that we had several days sub zero in the morning and not much over 20 during the day, since Thanksgiving...
If the weather is ok I might make day trips with a rented plane from Shaumburg regional, where I learned to fly.
appleton would be good
It s actually warmer in here during the cold weather, with the sun...yesterday was sunny though, and its nice again now...
Cold is good for flying...air is smooth and much much more lift
but airspeed slower
I'm used to landing on packed snow and ice too
Southwest Airlines will have to hire you then!
really on ice you just have to keep flying the plane while rolling
using adverse yaw
yoke left, turn right
and so on
aileron induced drag steers you
Never got into to flying...Was up in a glider once...
My dad was in the Airforce...rear seater in a the first jet fighter bomber in the late 50's ...
Well I was practically walking distance from the then Schaumburg air park
The one with the big wing tanks on the tips...
It was a fighter plane with a F designation...
two seater...preeeeee preeeee F 4 ...
late '50s jymmm
F 4 was around
they had planes in the 50's?!
F 80 shooting star had wing tanks and was big...but that was a "40s plane
Jymmm, I remember the '50s. Everything was in black and white, and everyone had flat tops.
They had TV in the 50's?
They had electricity in the 50's?
Next you'll say they had fire in the 50's too
and maybe even athe wheel
hey, I went to the first grade in...
no accounting for taste =)
really, things were weird then. It was kind of all this leave it to beaver crap.
really bizzare as I remember
it REALLY was like that?
it was awful
If anyone cares... disks at the bottom --> http://pascal.sources.ru/museum/bp7.htm
never wrote code in pascal
only fortran, basic, and c.
ah, bored her to death...
Ok...I think it was a P-80 ....
Well, might be nice to have 'just in case'
P 80 ok
big plane for a fighter
Yea, it was a fighter bomber, allweather...
gotta be a T33 variant...only 2 seat model...in production till 1959
They were stationed in Albany, GA , not too far from you there Les...
Turner AFB ...
love to fly one.
[17:29:15] <les_w> http://members.chello.nl/m.waterloo/t33-panel.html
I guess they originally were fighters, but the F86 replaced them...so they probably converted em into fighter bombers and trainers...
Anyway, lunch time...
yeah I will eat and then actually work on the cnc rewire a little
can linear scales be used as encoders?
with good mechanical linkages yes.
You can not use backlash compensation with linear scales
and if there is much backlash, motor windup will kill any stability.
davee did some testing of this stuff a few years ago.
moto windup ?
a servo builds up speed without feedback.
so you command it to move some short distance on the linear scale
with backlash, it will be moving fairly fast when it begins to see scale changes
so it overshoots then corrects but overshoots the other direction as well.
well, the scales would be absolute, wouldn't the TP do what the scales say?
assuming 0.004" backlash
greetings from the mountains :D
your net must work...
[18:41:13] <alex_joni> http://www.cncgear.com/EMC/BDI/Change.log
how is the camera working? Shouldn't we have pictures to view by know?
skunkworks: shitty weather
nothing nice to shoot :(
only snapshots so far
cradek: what is?
the bdi page
K4ts is now known as Jacky^
help ! sigh ..
now that's strange..
* K4ts is now known as Jacky^
Just arrived in naples :)
I see.. hello
im tryng to share dsl connection using windows xp and my laptop (linux ) seems hard
ive 2 wifi card, it work for the lan connection
I cant find the way to share dsl connection from PC dektop (xp)
the issue, is K4ts use dsl modem and the IP 192.168.0.1 is already busy by the modem ..
xp attomatically want to use this IP to share connection :(
* Jacky^ grr
Jacky^: I does want to use it, but you can set another one afterwards
and it'll work
I do it on 192.168.0.37
from PC desktop side ?
with xp ?
Hi Alex, perfekt timing :-)
Jacky^: did you get what I said?
[20:54] <alex_joni> first use the wizard to share the connection
[20:54] <alex_joni> and let it change your ip to 192.168.0.1
[20:54] <alex_joni> afterwards change it back to what you want
[20:54] <alex_joni> and it'll work
[20:55] * alex_joni thinks it's stupid.. but hey it's windoze
thanks.. now tryng
Imperator_: say quick I'm about to go away :D
its nothing, i was only surprised that you have joind at the same moment as I
heh.. I was here a bit earlier
but my connection crapped out
ok, I need to disconnect the modem before (it has 192.168.0.1) ..
* alex_joni will be back later
damn.. nothing to do
it take the adress but cant surf the net
maybe DNS is the problem
Jacky^: you use windoze as router to the net ?
the problem is the modem use 192.168.0.1, (I cant change it)
which one ?
XP side (pc desktop) i've an ethernet card DHCP with IP public static adress
D-Link G300 is the modem
type ipconfig /all
the modem can't have 192.168..... otherwise you are not connected
yes, the modem use this adress in lan (ethernet modem)
where the PC desktop have an ethernet card receiving the Public IP in dhcp by the modem
is that a modem or router ?
just a modem :(
not a router..
otherwise I would have to use a switch ..
has the modem more than one ethernet port ?
now, the ethernet card in PC destop have the public IP 22.214.171.124
the modem have a secondary port , but for console use
the wifi card (pc desktop side has 192.168.0.37 now)
I can ping mi laptop
and your Linux PC has 192.168.0.0 ore something like that
but cant surf from my laptop
the laptop (Linux) have 192.168.0.38
the lan work
you have added the ipadress of your windoze machine as standard gateway ?
on the laptop
I tried .. wont work
laptop is Linux ?
Linux and windows
having 2 partitions
I think default share connection in xp work only for 192.168.0.1 address
try this adress from your Laptop 126.96.36.199
and from desktop side ? 192.168.0.37 in the wifi card its ok ?
i have the 192.168.0.1 at the desktop, but i don't know if that is realy nessesary
microsoft says it is, but i think it is also possible with a other one, but Im not shure
Jacky^ What does the dlink modem look like? does it have an antenna?
hi Jymmm, no antenna, just an ethernet modem
oh, no idea i have cables
describe it.... color?
its not wifi
don't you have a normal ethernet connestor at your laptop
Jacky^: No, I asked you to DESCRIBE what it looks like.
ok grey, what else? perfectly square?
rectanguar box, gray colour, 4 green led
i have no d-Link stuff, i have two normal PCs with cheap noname ethernet cards
on the other side, ethernet port, console port, dsl port
and ps connector
Jacky^ no hub built in?
I could connect my laptop and share the connection with linux ..
but I would the opposite
to share the connection from xp
so, I could move in the house with the laptop
damn windows.. I think it share the connection only if 192.168.0.1 is free ..
If I remember well, theres some application (shareware) to solve this issue..
maybe I found ..http://www.nat32.com/
ghghg not for free :(
XP has built in NAT. As well as internet connection sharing.
does the xo box have two nics?
Jymmm, yes Desktop PC have ethernet card with public IP and wifi card with private IP
restriction in xp is the IP used to share the connection, xp want to use only 192.168.0.1
when the pc cant use it, already busy by the modem
Yes, 192 subnet on wifi
the modem is using NAT?
I tried to force it to use 192.168.0.37, as alex sayd, it take the IP but the connection isnt shared
192 is NOT a pubic IP address
you said it was
ok, then listen ..
what ip is the xp box?
Modem use 192.168.0.1
no not modem. XP
modem have onbuilt dhcp
so, the modem pass the Public IP to the pc desktop ethernet card
192 !+ Public IP
192 != Public IP
WHAT IS THE IP ADDRESS OF THE XP BOX?
well , I did ipconfig /all
I've: ethernet card 188.8.131.52 (thats public ip)
and the wifi card (192.168.0.37 - private)
paste the output of ipconfig /all here
I can ping 192.168.0.1 from dos, thats the modem
and how I can copy text from dos ? :/
fine... ipconfig /all > text.txt
fine... ipconfig /all > c:\text.txt
its long .. may is better to use a query ?
going to try nat32 ..
has the base unit in emc2 changed from millimeter to inches ?
my INPUT_SCALE need that factor
I don't quite understand what you mean by base unit?
If you set an axis UNITS = 0.03937007874016 than the INPUT_SCALE is pulses or encoder ticks per inch
If you set an axis UNITS = 1 then the INPUT_SCALE is pulses or encoder ticks per mm.
Jacky^ is now known as K4ts
K4ts is now known as Jacky^
rayh: sorry for the delay
rayh: the axis unit is 1.0
rayh: and i have 200 steps/rev steppers with 1mm/rev movement
rayh: its a simple 6mm screw axis
* Ephexis 00is Away, Reason: 14( 07nadando no sofa 14) 14|00 Since: 14( 07Wednesday00, 07December 2800, 07200500.07 1800:074300:0755 14)07 Xlack 00v072.1
rayh: so i expected an input scale of 400
sounds like a scale of 200
rayh: ah, the steppers are driven in half step mode
ah, then it is 400
yep, thats what i expected
* Ephexis 00is back 14(00 Away07 1 min 16 secs 14)
but to get the exact positions i need it to be 144
which is 400 / 2,54
minus 10 %
rcsu: I think that's a red herring
* rcsu runs a cvs co from yesterday evening
I don't think anything has changed since then...
i have no idea what parameter can cause this
the docs say what i set first
which emc are you using?
emc2 from cvs?
but the machine moves 2,54 mm per unit
emc2 from csv
have you tried g20 and g21?
nope, i thought metric is the default
lemme check that, takes a mmt
the default interp mode comes from the ini too
there's a startup g-code that is g20 or 21
but all of this will give you a 25.4 problem, not 2.54
where is that set ?
in the ini
my nc machine starts up right now
what ini did you start with? one of the distributed ones?
y, i copied the demo_step_cl
did you change [TRAJ]LINEAR_UNITS as well as [AXIS_*]UNITS
all to 1.0
everything is set to 1.0
just checked to be sure
after you run something like g0x10 does the gui say x=10.00?
yes, +/- a few hundrets
hmm, I'm not sure what's going on then
what do you need ?
i wanna help here
where is the calculation from units to steps in the source
well there's a lot of those
but if you're in mm, all of the units are the same
to verify that it has your user units right, you could set an offset and then examine the var file
but I'm sure it'll be right
for instane program g10 L2 P1 X10, then exit emc, then look in the .var file
you should see 10.0
another thing you could try is to look for differences between your ini and configs/stepper/stepper_mm.ini
man you learn another coordinate offset gcode every day
the .var 5221 is 10.0, but 5212 is -0.1632
wonder what 5212 is...
all else are 0
bah, 5220 is 1.0
looks like 5212 is G92's Y
yes, thats not aprob
ok so I think this means the interp is sane
the var 5220 is 1.0 must be a number between 1 and 10
do you need cl?
so that the interpreter knows which g54, g55 ... to start up using.
no, i commented it out completely
I know alex put together configs/stepper/stepper_mm.ini and it's known to work as mm
why don't you try running stepper_mm then
i checked against it
but i dont see any significant differences
the halfiles may be very different
they are, i adopted them to my needs
I guess there's no scaling in the hal file
mostly its commenting out, but no calculations in them
so the machine moves farther than you expect?
until i reduces the input scale
can you measure how much?
to be clear: 1 inch is 25.4 mm
ha, I know this
but you said it doesn't exactly move 2.54x too far
soryy, just to be sure to speak about the same :)
I live in inchland but I do know what a mm is :-)
are your accel and velocity conservative? Do the motors SOUND right?
the move @INPUT_SCALE == 400 was 2.79 mm per unit
that was exactly repeatable?
so i set it to 400/2.54 = 157
cradek: no hardware probs, a stepper sounds good or really bad :)
but that scale leads to a 11mm move when i said 10
and a SCALE of 144 leads to the real wanted move of 10 mm
I can't fathom why it would be off by such a strange amount
is your interface step/dir on parport?
1 mm pitch screws are very odd...Only machine I know like that is a Proxxon MF-70 (which I bought the first one sold in the USA) ...Are you sure you have 1 mm pitch screws ?
wb9mjn: my sherline has 1mm acme screws
really? the MF-70 uses trapezoidal screws....
rcsu: to be clear, I think your scale error of 2.7778 being close to 2.54 does not signify a units problem
wb9mjn: well I'm not sure it's acme, but I'm sure it's 1mm pitch
wb9mjn: I've turned the dials plenty of times to know that
I did too till i put steppers on it...
rcsu: emc barely knows about centimeters - any scaling problem due to units confusion would be 25.4x
What kind of machine is it Resu ?
The sherline is about twice the size of a proxon ...
What diameter are the screws on the Sherline, they are 6 mm on the MF-70...
cradek: 1st, my I/F is on the parport and 2nd that factor of 25.4 - 10% confuses me too
cradek: but i dont see any factors that influence that
rcsu: so if your scale is 400 and you command a move of 1mm your motor should make one turn exactly?
cradek: nope, it turns 2.3x mm
cradek: thats strange
cradek: err it turns 2.7x mm
On my MF-70 ini file I have input and output scales on each axis set to 2000...direct drive, 200 step per rev, X 10 microstepping...
rcsu: can you tar up your ini files and put them on the web somewhere, or email them to me?
With a 1.0 scale factor...
cradek: if you tell me your mail addy :)
in a mmt on the way
ok, I have to run
I will try to look at them later tonight, and I'll be on irc
Can you tell me about your machinec RCSU ?
Try setting the stepper drivers to full stepping and see if you get the same ratio with 200....If not, it would be time to actually count the pulses out of the parallel port, and see if
possibly there is some wierd driver action like multiple drivers steps intermittantly....like your on the edge of a false trigger with the pusle width and setup ...
wb9mjn: i can clearly see the unit differences
Counting the pulses will also confirm your system to the computer side...
wb9mjn: if you can circle clean, then ...
When I first tried full/half step drivers I had all sorts of problems with false triggers and stepper resonances...
I eventually went to micro-stepping drivers, which have a predistortion which tunes the stepper back into a stable zone...
wb9mjn: if your steppers sounds "round", then they drive such
the steppers wound quite smooth here.....
wb9mjn: and one point: keep is slow, not at the most speed
Do you have backlash set to 0 ?
ah, yes thats what i didnt changed
backlash is a feature that counter rotates the screw quickly after a movement reversal...
rcsu: any more problems?
alex_joni: yes, unit probs
Set it to 0 for initial configuration, to be sure the steppers are making the right number of turns...
alex_joni: all UNIT params set to 1.0
alex_joni: but the spindle moves 2.54 - 10%
alex_joni: i compensated it with INPUT_SCALE
alex_joni: but it cant be the truth
so start over :)
i've copied the demo_step_cl stuff to rcsu1
renamed all the files and all references to the files
then i commented out all the classicladder stuff, cause its just not needed here
but if i do some disatance measuring, i got strange results
i got to the result, that i had to divide the INPUT_SCALE by 2.54 and subtract 10% to get nearly that distance what i want
alex_joni_ is now known as alex_joni
that is odd
alex_joni: did you see what i typed ?
no my connection crapped out
what did y see last
[01:30] <alex_joni> rcsu: any more problems?
[01:34] <wb9mjn> the steppers wound quite smooth here.....
[01:35] * Disconnected
[00:31:23] <rcsu> alex_joni: all UNIT params set to 1.0
[00:31:44] <rcsu> alex_joni: but the spindle moves 2.54 - 10%
[00:32:06] <rcsu> alex_joni: i compensated it with INPUT_SCALE
[00:32:24] <rcsu> alex_joni: but it cant be the truth
[00:32:32] <rcsu> k,
[00:32:40] <rcsu> so start over :)
[00:33:27] <rcsu> i've copied the demo_step_cl stuff to rcsu1
[00:33:48] <rcsu> renamed all the files and all references to the files
[00:34:35] <rcsu> then i commented out all the classicladder stuff, cause its just not needed here
[00:35:06] <rcsu> but if i do some disatance measuring, i got strange results
[00:35:57] alex_joni_ (email@example.com) has joined channel #emc
[00:36:06] <rcsu> i got to the result, that i had to divide the INPUT_SCALE by 2.54 and subtract 10% to get nearly that distance what i want
what's your INPUT scale?
and how did you calculate it?
the first try was 400 0
with a 6mm spindle on a 200step/rnd stepper in halfstep mode
round beeing = ?
don't assume I know what your setup is
one turn of the stepper axis
are you using rack/pinion drive? or ballscrews?
or linear motors?
steppers directly connected to a "gewinde"
if it's a ballscrew (most likely) what's the tpi ?
rcsu: you can say it in german if it's easier
i know :)
wie sieht das Gewinde aus?
even in here, people won't mind
and if they do, I'll kick them *grin*
its "gewindestange" 6mm, mit einer Steigung von 1 mm / Drehung, DIN halt
das kann ich nicht verstanded.
ok.. got it
so it's a normal screw (not ballscrew) with 1mm / turn
6 mm diameter screw
are you very sure it's 1mm /turn?
yes! im really shure
set your INPUT_SCALE to 400 (if you have 200 pulses / turn, and running in halfstepping mode)
then try G1X1F10
and see if it turns exactly one turn
if it does then it's ok so far, and you need to see if it moved 1mm
if it doesn't then something might be wrong (either not 200 pulses/turn, or you're losing steps)
alex_joni: first, i tapped a plug on the axis
alex_joni: i wound more than one time around
hm, 48 steps *2 == 96 which isn nearly 144
Sounds like software setup problem, but could be driver false triggering...Try a measuring the displacement very accurately with a caliper, and see if its the same each time...
If it is, then keep looking at software setup...
If its a little different each time, then start looking at driver triggering issues...
set the scale to something like 100
and then try X1
wb9mjn: im quite sure that this is a software prob
it should give you exactly 100 pulses
been there, hi...
rcsu: that depends on what you see as a software problem
might be that your drives need a different timing
the measurement will quickly figure that out though...
or you have too high pulsing set..
alex_joni: either a setup prob or a program prob :)
If pulsing was too high the displacement would be short, but his is high...
rcsu: might, although highly unlikely as a lot of people are using it with success
Double triggering sometimes ?
wb9mjn: I suspect it's not halfstepping all the time
maybe you didn't pull that high/low as you should
left it floating?