http://ef.algebra.com/e/300417172682
http://www.lowcostcncretrofits.com/
This kit goes along the lines of your "go nuclear" option.
This would look like it is a one day job to install. Considering how
many things can go wrong with a DIY conversion, tthis system seems
very attractive.
Selling off removed parts can recoup a part of the cost.
I want to avoid the morass like my [ultimately successful] welder
microcontroller project. It sapped all my energy for six months,
eventually worked exactly as intended, welded great, but was
inconvenient to use and cost a lot of money in trinkets, parts that I
ended up not using, etc.
Any thoughts?
i
I can't comment on whether this particular vendor's offer is good. There are
many offers like this around. I've not used any nor investigated them.
Your concern about spending a lot of energy is well founded. The first
project is often a struggle, particularly if you don't have similar
experience. A well engineered quick build kit should go in a couple days.
The offer you showed, and most others use stepper motors. Quite a step down
from servos with position feedback under EMC. Many hobbyists are satisfied
with stepper systems - depends on your needs.
Another option to consider is hiring install help. In this economy $2K will
buy a lot, particularly if your money has a picture of Ben on it.
Right now, I'm consulting on a bit more involved bridgy refit with
toolchanger. The guy does all the wrenching but I direct and problem solve.
He'll end up paying me about $1K.
Karl
Are there any kits that you have investigated?
> Your concern about spending a lot of energy is well founded. The first
> project is often a struggle, particularly if you don't have similar
> experience. A well engineered quick build kit should go in a couple days.
Yep
> The offer you showed, and most others use stepper motors. Quite a step down
> from servos with position feedback under EMC. Many hobbyists are satisfied
> with stepper systems - depends on your needs.
>
> Another option to consider is hiring install help. In this economy $2K will
> buy a lot, particularly if your money has a picture of Ben on it.
>
> Right now, I'm consulting on a bit more involved bridgy refit with
> toolchanger. The guy does all the wrenching but I direct and problem solve.
> He'll end up paying me about $1K.
I would pay you in a heartbeat, if you were local.
i
Bob Swinney
"Ignoramus18864" <ignoram...@NOSPAM.18864.invalid> wrote in message
news:-5ednU9A94Cix1vW...@giganews.com...
It doesn't matter if he's local Iggy, you certainly have the skills to
do the physical work, and you have phone, email and digital cameras to
remote collaborate with.
No, steppers ABSOLUTELY do not meet my needs. I changed my servo Bandit knee
mill over to an Ahha control with steppers and was sorry. That was about
1995. I strictly scrounge auctions and eBay. Doing this, a commercial
control worth 15K can be done for 2K or even less. After doing Ahha, Mach,
Centroid, and (very briefly) EMC; I've settled on only Camsoft for my
control software.
...
>> Another option to consider is hiring install help. In this economy $2K
>> will
>> buy a lot, particularly if your money has a picture of Ben on it.
>>
>> Right now, I'm consulting on a bit more involved bridgy refit with
>> toolchanger. The guy does all the wrenching but I direct and problem
>> solve.
>> He'll end up paying me about $1K.
>
> I would pay you in a heartbeat, if you were local.
I bet there's several hundred fully qualified people to do this work in
Chicago land.
Yep, still owe you one. Lots of other folks too. Trouble is I don't deliver.
karl
I mean, servo kits, are there any servo retrofit kits?
> I strictly scrounge auctions and eBay. Doing this, a commercial
> control worth 15K can be done for 2K or even less. After doing Ahha,
> Mach, Centroid, and (very briefly) EMC; I've settled on only Camsoft
> for my control software.
Cool, I will look./
> ...
>>> Another option to consider is hiring install help. In this economy $2K
>>> will
>>> buy a lot, particularly if your money has a picture of Ben on it.
>>>
>>> Right now, I'm consulting on a bit more involved bridgy refit with
>>> toolchanger. The guy does all the wrenching but I direct and problem
>>> solve.
>>> He'll end up paying me about $1K.
>>
>> I would pay you in a heartbeat, if you were local.
>
> I bet there's several hundred fully qualified people to do this work in
> Chicago land.
>
>
I think so too, I had one person recommended.
i
Iggy,
Did you find out what is good and what is bad on your mill? You may be able
to use another control, such as PC with EMC2, with your existing drives and
servos. It takes a little to get into it but it starts making sense pretty
quick. I even added speed feedback to my lathe, it corrects the spindle
speed according to the encoder, took a bit to get stable.
RogerN
You already saw some of them available from MachMotion.
Why don't you look up Jon Elson? He's an EMC guru. I know EMC has a real
active IRC channel if you're into that.
To answer your question, there's got to be a company that offers step-servo
drivers with a breakout box to Mach. just connect wires and go. I've never
worked with this, but the control still doesn't get position feedback (open
loop). I don't see how this is much better than steppers.
Why don't you spend a couple evernings cruising CNCzone.com Its a newsgroup
so you've got the usual signal/noise problem but lots of good info. All
these questions have been asked before.
karl
It's a half closed loop. The servo drives flag an error to stop the
control if the servo error exceeds a threshold like 128 encoder counts.
With steppers the control blindly continues when the stepper misses
steps or stalls, with the step/dir servos, the control stops if the
servos can't hold position, just like the fully closed loop systems.
>
> Why don't you spend a couple evernings cruising CNCzone.com Its a newsgroup
> so you've got the usual signal/noise problem but lots of good info. All
> these questions have been asked before.
Really, he should be going to the Digital Machinist Workshop in June
where he can talk to a lot of folks who have done conversions, look at a
lot of the products available and talk to the vendors.
They're worth the wait!
>
>>> I can't comment on whether this particular vendor's offer is good. There
>>> are
>>> many offers like this around. I've not used any nor investigated them.
>>
>> Are there any kits that you have investigated?
>
>No, steppers ABSOLUTELY do not meet my needs.
Do they leave a stepped finish, or what? Why the horror/caps?
>I changed my servo Bandit knee
>mill over to an Ahha control with steppers and was sorry. That was about
>1995. I strictly scrounge auctions and eBay. Doing this, a commercial
>control worth 15K can be done for 2K or even less. After doing Ahha, Mach,
>Centroid, and (very briefly) EMC; I've settled on only Camsoft for my
>control software.
What did you get with Camsoft that the others didn't have, Karl?
LJ, in learning mode.
--
STOP THE SLAUGHTER! Boycott Baby Oil!
Open loop control, i.e. you have to supervise the machine closely to
stop it if the steppers start missing steps due to high loading since
the control can't detect it. Otherwise, as long as they are operating
within their limits they work just fine and are less expensive.
--
"Steamboat Ed" Haas : Blue Cross socks us
Hacking the Trailing Edge! : $23,000/yr!! ...
www.nmpproducts.com
---Decks a-wash in a sea of words---
I spoke to Jon today and it was a great conversation. It was nice to
talk to him after knowing him for so many years. I will bring the mill
home from my friend's warehouse, in 2 weeks, and I am sure that Jon will
hear from me again!
i
Just a thought for you. I'd WAY rather refit at my place. Bet anyone you
find feels the same. Its just handy and faster to use your own shop.
Everytime you travel it seems something is missing and you have to wait and
come back. Wasted time costs money you didn't need to spend.
Karl
I feel the same way. But the mill is taller than my garage door and it
needs to go in tilted. I will weld a special carriage for it starting
this weekend.
i
Just think of your mill as a heavy lego mindstorm :-)
You can find out how it works and in the process you can learn to add
functionality to it. It will also be a benefit to you when you need to make
a repair also. If you learn the signal from the encoder to the control, and
the command signal from the control to the servo drives, you will be able to
determine what part of the system is not working correctly.
I'm just trying to encourage you to learn to do the retrofit yourself unless
you are pressed for time, such as need to get it running right away for
making money producing parts. The system consists of a motor, amplifier,
control, and position feedback. None of what you need to know is very
difficult, you can get it working 1 step at a time. Get the encoders
working so that when you crank the table you can read the position in the
control. Turn the gain settings way down in the program, connect the
controls signal and get ready on the E-Stop, if it takes off the wrong way
E-stop it and reverse the signal wires. After you get the feedback in the
right direction, and the control signal in the right direction, then you
adjust the gain settings to get a responsive system without oscillation.
Each step has a little more detail than I went into but that's the overview
of the process, it isn't very difficult but is rewarding to know how it
works. "I think you can, I think you can..."
RogerN
My bandit was a decent enough mill with servos. Got tired of the whopping
120K program size limit. When i switched to steppers with PC control, I
screwed up several parts. So, I spent nearly 3K on large steppers and large
stepper drives. Still screwed up parts. Keep in mind it only takes a split
second on one move for the rest of the program to be off. All in all this
change cost me thousands and untold wasted time. A servo can go 100% over
power for a split second when it gets behind. There's no way to tell a
stepper - no feedback.
A person making tiny parts at very slow speeds won't have issues. Not what I
do.
>
>
>>I changed my servo Bandit knee
>>mill over to an Ahha control with steppers and was sorry. That was about
>>1995. I strictly scrounge auctions and eBay. Doing this, a commercial
>>control worth 15K can be done for 2K or even less. After doing Ahha, Mach,
>>Centroid, and (very briefly) EMC; I've settled on only Camsoft for my
>>control software.
>
> What did you get with Camsoft that the others didn't have, Karl?
Camsoft is open source. use almost any vendor's hardware. It also uses a
Galil motion card for real time motion control - top of the line. it is also
completely programable. If you have the ability there is nothing it can't
do. The one downside, its not cheap.
>
>>>No, steppers ABSOLUTELY do not meet my needs.
>>
>> Do they leave a stepped finish, or what? Why the horror/caps?
>
>My bandit was a decent enough mill with servos. Got tired of the whopping
>120K program size limit. When i switched to steppers with PC control, I
>screwed up several parts. So, I spent nearly 3K on large steppers and large
>stepper drives. Still screwed up parts. Keep in mind it only takes a split
>second on one move for the rest of the program to be off. All in all this
>change cost me thousands and untold wasted time. A servo can go 100% over
>power for a split second when it gets behind. There's no way to tell a
>stepper - no feedback.
>
>A person making tiny parts at very slow speeds won't have issues. Not what I
>do.
So the price of larger steppers/drives is nearly that of servos?
>>>I changed my servo Bandit knee
>>>mill over to an Ahha control with steppers and was sorry. That was about
>>>1995. I strictly scrounge auctions and eBay. Doing this, a commercial
>>>control worth 15K can be done for 2K or even less. After doing Ahha, Mach,
>>>Centroid, and (very briefly) EMC; I've settled on only Camsoft for my
>>>control software.
>>
>> What did you get with Camsoft that the others didn't have, Karl?
>
>Camsoft is open source. use almost any vendor's hardware. It also uses a
>Galil motion card for real time motion control - top of the line. it is also
>completely programable. If you have the ability there is nothing it can't
>do. The one downside, its not cheap.
Ouch, no, it isn't.
Thanks.
600 oz in steppers are common with the hobby market. You get into big bucks
going larger.
600/12/16 =3.1 ft lbs
You run steppers at low rpm for best results, say 200
torque*rpm/5252=hp
3.1*200/5252=.11 hp
A 1/4 hp. servo costs less and is rated for 1/2 hp for a few seconds AND
your control will let you know the second there is trouble.
Karl
To follow on to that, the next incremental step up from open-loop
stepper control, is DC brush servo control with step/dir drives such as
the Gecko drives. These provide closed-loop servo control and a fault
shutdown signal to the control (like Mach3) to stop motion instantly if
the servo error exceeds the threshold. Since it is servos, it has the
peak overload capability of servos, the speed of servos and since it has
the position encoders, it has the closed-loop of servos. The only thing
it lacks is the encoder position feedback to the control software, but
since it has the fault output, this lack is of little practical
difference for normal use. This setup eliminates the problems with
open-loop stepper control that Karl mentioned, without the expense of
industrial grade servos, drives and motion control cards.
Do you have knowledge of any fairly inexpensive systems on the market
built like this?
---
A book burrows into your life in a very profound way
because the experience of reading is not passive.
--Erica Jong
Software wise there is Mach3 that runs on a Windows system or EMC2 (free)
that runs on Linux (free). I went with EMC2 because with it I can use
hardware compatible with my existing servos and drives. My servo drives
take a +-10V signal and position is sensed by encoders. Last time I looked
I believe Mach3 only had step and direction outputs. EMC2 can be configured
for step and direction or analog signal outputs and they can be mixed and
matched. For example with my mill I plan to convert, I have 3 axis servo
motors and would like to motorize a dividing head with a stepper motor and
drive I already have. Plus if I add an encoder to my mill spindle then I
can do rigid taping.
Something I would like to do sometime is configure a stepper axis and use
position feedback, either DRO scales or a rotary encoder. I know with
steppers, if they slip and get behind, speeding up isn't usually the
solution, but at least I could fault on position error and not scrap the
part. I would like to figure out how to do a feed rate override and slow
down the feed if an axis got behind.
EMC2 is very configurable and that also means it has to be configured, but
it isn't that difficult because there are multiple configurations already
done, you just pick something close and modify it for your application. For
example for a stepper mill, the example may use a different number of steps
per inch (or mm)or may use limit switches or maybe not. For my system I
wired an encoder to my breakout board and played around with settings to
learn how things worked. Once you pick a similar configuration and look it
over, it starts to make sense and is easy enough to figure out.
RogerN
>I feel the same way. But the mill is taller than my garage door and it
>needs to go in tilted. I will weld a special carriage for it starting
>this weekend.
Iggy,
Are you involving your son in this? Gotta be more interesting then me watching/helping
grandpa with painting and plumbing as a child.
Wes
--
Interesting people had interested parents.
Look at the servo based systems there, they use the Gecko step/dir servo
drives.
Cool. Pricing not too bad. Mach3 license at half price, good!
Why are so many of these parallel port driven instead of USB?
Pretty simple - With a parallel port, the software on the PC, be it
Mach3, EMC, etc. can generate the step/dir pulse train directly, while a
USB interface required an external CPU to accept trajectory commands
from the PC host and generate the step/dir pulse train. It's simply a
matter of cost as there are a number of USB connected pulse generators
available for $100-$200 or so.
Good, clear answer. Thanks, Pete.