Enhancement
Opened this issue · 21 comments
loucksg commented
Hi Bar,
I could not see how to lodge an enhancement rather than an issue, i think this selected enhancement, the following is Enhancement and may be achieved by G-Codes or another question:
Manual set / adjust of the chain lengths / home position to calibrate the 'centre' of the cutting area, this could also be extended to the 'corners' so you can set your calibration rectangle, or derive it from there.
Alternate, the ability to run 'calibration' sub-steps as well, you can sorta do this via start, then skip....
BarbourSmith commented
You did it perfectly. I just needed to add the enhancement label (for some reason Git Hub won't let anyone else add labels 🙄 )
The good news is if I'm understanding right this feature is already in place although a little hidden.
You can choose how much chain is extended during the "extend chains" step in the advanced settings menu:
That setting will adjust how much chain is extended. The "Center" position is computed automatically from the distance between the motors and the vertical height found in the calibration process along with the bed size settings here:
Are those the features needed, or is there more functionality also?
davidelang commented
On Tue, 29 May 2018, BarbourSmith wrote:
You did it perfectly. I just needed to add the enhancement label (for some
reason Git Hub won't let anyone else add labels 🙄 )
you need to give other people full commit access to the repo, then they can act
as additional admins and do tagging)
k
loucksg commented
Hi Bar,
I’ll look at this on the weekend, this may well do it, I dropped a note to Hannah, explaining the ‘stall’ scenario, the critical one I’m running into currently is the motors driving into ‘stall’. The length will ‘provide room’ what I’m looking at now is so the chain does not get too-short.
So you have - chain length, the length you’ve fed the chain, the size of the sled, however, you can ‘never’ get to the minimum chain length. Especially once you add the sled! When you run the chain length to stall to measure the distance between the motors, even with the ‘sled’ offset I think a ‘safety’ floor for the total chain length may be required?
As an EE, remember Catenary wires / transmission towers, a*cosh(x/a)? I think normally people don’t run into this, unfortunately, I made my motor mounts too small :-o, trying to hang on wall under beams in the shed, some welding will fix on weekend :-p!
Now knowing about those advanced settings, I’ll take a look, is there a way to run a ‘single’ calibration’ step, or must you always go to the beginning and click skip, skip, skip?
Great work, your code looks good, it’s much better than the base code I was working with when I built my 3d printer a few years back, Delta style ;-)!
I think my request here I was looking at ways to ‘easily’ set some of the setting points other than ‘home’, I’ll go back and have a read as well as looking at this, care to move this to the ‘close pending’ and I’ll explore on the weekend?
Cheers,
Guy
… On 30 May 2018, at 4:10 am, BarbourSmith ***@***.***> wrote:
You did it perfectly. I just needed to add the enhancement label (for some reason Git Hub won't let anyone else add labels 🙄 )
The good news is if I'm understanding right this feature is already in place although a little hidden.
You can choose how much chain is extended during the "extend chains" step in the advanced settings menu:
<https://user-images.githubusercontent.com/9359447/40676796-a93e3a42-6330-11e8-9dcc-902d551967a2.png>
That setting will adjust how much chain is extended. The "Center" position is computed automatically from the distance between the motors and the vertical height found in the calibration process along with the bed size settings here:
<https://user-images.githubusercontent.com/9359447/40676875-e042c828-6330-11e8-9072-f831dafafa66.png>
Are those the features needed, or is there more functionality also?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub <#728 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/Ak-sfq-pMYZJRmSoy9omBQqw5dPhaSfQks5t3Y8jgaJpZM4UPJx8>.
BarbourSmith commented
Thanks 😀
I added the "question" tag and we'll close it once we've got you a good answer.
Which step in the calibration process are you looking to do without skipping? Is it the extend the chains step? If so you can run that one by clicking Actions -> Set chain lengths automatic
davidelang commented
On Wed, 30 May 2018, loucksg wrote:
I’ll look at this on the weekend, this may well do it, I dropped a note to
Hannah, explaining the ‘stall’ scenario, the critical one I’m running into
currently is the motors driving into ‘stall’. The length will ‘provide room’
what I’m looking at now is so the chain does not get too-short.
So you have - chain length, the length you’ve fed the chain, the size of the
sled, however, you can ‘never’ get to the minimum chain length. Especially
once you add the sled! When you run the chain length to stall to measure the
distance between the motors, even with the ‘sled’ offset I think a ‘safety’
floor for the total chain length may be required?
what are you meaning when you talk about "running the motor to stall"?
the measurement between the motors is not done with the sled on the chains, it's
running just the chain between the motors.
As an EE, remember Catenary wires / transmission towers, a*cosh(x/a)? I think
normally people don’t run into this, unfortunately, I made my motor mounts too
small :-o, trying to hang on wall under beams in the shed, some welding will
fix on weekend :-p!
we do take the chain sag into account.
Great work, your code looks good, it’s much better than the base code I was
working with when I built my 3d printer a few years back, Delta style ;-)!
it's gone through a lot of revision and cleanup from the community (and more
help is welcome :-) )
I think my request here I was looking at ways to ‘easily’ set some of the
setting points other than ‘home’, I’ll go back and have a read as well as
looking at this, care to move this to the ‘close pending’ and I’ll explore on
the weekend?
what needs to be done is to add support for the G54 and similar codes
David Lang
loucksg commented
Hi,
The test cuts after the chain length and sled / cutting edge sets.
Is there a min chain length for safety, ulterior ask, I’ve a nice triangular set is like to avoid pulling apart...
Guy
Guy's mobile
… On 31 May 2018, at 2:20 am, BarbourSmith ***@***.***> wrote:
Thanks 😀
I added the "question" tag and we'll close it once we've got you a good answer.
Which step in the calibration process are you looking to do without skipping? Is it the extend the chains step? If so you can run that one by clicking Actions -> Set chain lengths automatic
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub, or mute the thread.
davidelang commented
On Wed, 30 May 2018, loucksg wrote:
The test cuts after the chain length and sled / cutting edge sets.
If you have changed anything, you need to do the cuts.
If you don't change anything, you can have a link in the chains marked and reset
to a known position using this.
Is there a min chain length for safety, ulterior ask, I’ve a nice triangular set is like to avoid pulling apart...
avoiding pulling the linkage apart isn't a matter of avoiding a minimum length.
The distance from the corner of the workpiece to the motor is FAR larger than
the rotation radius (typically something > 3x, even with the very large top
mount kit I make)
the highest stresses are in the top center, even further away from the motors.
loucksg commented
Hi David,
So in the setup (calibration):
chain length measure - motor distance - ok
chain length measure out to set length - ok, NB: it did wind the one chain backward however could be picked up and place in ‘right’ direction
sled measure - ok,…
Then doing the ‘trial’ cuts it went to move the sled ‘up’ to start the cut, went all the way to the top of the board, and then pulled until stall. chucked error. It actually started to distort the 2*4, I had to put on my nice thick welding gloves and safety glasses ‘lifted’ the tensioned chain off the cogs!!! A lot of energy stored there!
G54 ;-) yes that would do it!
Once I get cutting, I’ll dig through and look at the code further. I saw a comment on the ‘chain’ variability, that’ll be an interesting one, it’s also related to the fact chains ‘stretch’ over time and temperature!
Guy
… On 31 May 2018, at 1:44 pm, David Lang ***@***.***> wrote:
On Wed, 30 May 2018, loucksg wrote:
> I’ll look at this on the weekend, this may well do it, I dropped a note to
> Hannah, explaining the ‘stall’ scenario, the critical one I’m running into
> currently is the motors driving into ‘stall’. The length will ‘provide room’
> what I’m looking at now is so the chain does not get too-short.
>
> So you have - chain length, the length you’ve fed the chain, the size of the
> sled, however, you can ‘never’ get to the minimum chain length. Especially
> once you add the sled! When you run the chain length to stall to measure the
> distance between the motors, even with the ‘sled’ offset I think a ‘safety’
> floor for the total chain length may be required?
what are you meaning when you talk about "running the motor to stall"?
the measurement between the motors is not done with the sled on the chains, it's
running just the chain between the motors.
> As an EE, remember Catenary wires / transmission towers, a*cosh(x/a)? I think
> normally people don’t run into this, unfortunately, I made my motor mounts too
> small :-o, trying to hang on wall under beams in the shed, some welding will
> fix on weekend :-p!
we do take the chain sag into account.
> Great work, your code looks good, it’s much better than the base code I was
> working with when I built my 3d printer a few years back, Delta style ;-)!
it's gone through a lot of revision and cleanup from the community (and more
help is welcome :-) )
> I think my request here I was looking at ways to ‘easily’ set some of the
> setting points other than ‘home’, I’ll go back and have a read as well as
> looking at this, care to move this to the ‘close pending’ and I’ll explore on
> the weekend?
what needs to be done is to add support for the G54 and similar codes
David Lang
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub <#728 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/Ak-sfgWVQp5slZVIHsqT9Fc3G9Av8su1ks5t32cmgaJpZM4UPJx8>.
blurfl commented
chain length measure out to set length - ok, NB: it did wind the one chain backward however could be picked up and place in ‘right’ direction
The calibration steps, when done in sequence, track the amount of chain extended since the sprockets were zeroed. When the motor separation has been measured, and the end of the chain lifted free but the middle of the chain still engaged on its sprocket, the next step will run the chain 'backward' until 1650mm of chain remains beyond that sprocket. The other sprocket will turn to extend 1650mm of chain.
That same chain calibration can be run alone and then both sprockets will turn as expected to move the chain ends.
A tip, once both chains have come to rest during the chain calibration, but before clicking 'Move to Center' or 'Next', mark the links on the top of the sprockets with some paint. These 'king links' are at a set point 1650mm from the end. Then go ahead and click the 'Next' or 'Return to Center' button to go on with what you were about. The next time you need to calibrate the chains, use the first step of Actions/SetChainLength-Automatic to zero the sprockets, then 'Quit' that without moving the chains. Put the 'king links' on the top tooth of the sprockets and use Actions/Advanced/SetChainLength-Manual to tell the machine that the chains are extended to the set point. If you have a way of supporting/hanging the sled to remove tension from the chains, this manual recalibration can be done without needing to disconnect the sled. Finally, use Actions/ReturnToCenter to move the sled to 0,0.
BarbourSmith commented
The issue is that the software has no safe check to see if the motors are a safe distance above the board. You can build a frame with the motors mounted right at the top of the sheet but then when the sled tries to move to the top of the sheet it will cause issues.
I don't want to restrict the way people can build their machines, but would a warning work? That way if someone entered a vertical offset measurement which was too small they would see a pop-up warning that it could cause problems
davidelang commented
On Thu, 31 May 2018, loucksg wrote:
Hi David,
So in the setup (calibration):
chain length measure - motor distance - ok
chain length measure out to set length - ok, NB: it did wind the one chain backward however could be picked up and place in ‘right’ direction
sled measure - ok,…
Then doing the ‘trial’ cuts it went to move the sled ‘up’ to start the cut, went all the way to the top of the board, and then pulled until stall. chucked error. It actually started to distort the 2*4, I had to put on my nice thick welding gloves and safety glasses ‘lifted’ the tensioned chain off the cogs!!! A lot of energy stored there!
In that case, something is wrong with the settings of the machine, this should
not be happening.
loucksg commented
A warning would be perfect, however can there also be a simple limit check as well?
Guy
Guy's mobile
… On 1 Jun 2018, at 3:15 am, BarbourSmith ***@***.***> wrote:
The issue is that the software has no safe check to see if the motors are a safe distance above the board. You can build a frame with the motors mounted right at the top of the sheet but then when the sled tries to move to the top of the sheet it will cause issues.
I don't want to restrict the way people can build their machines, but would a warning work? That way if someone entered a vertical offset measurement which was too small they would see a pop-up warning that it could cause problems
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub, or mute the thread.
davidelang commented
On Thu, 31 May 2018, loucksg wrote:
A warning would be perfect, however can there also be a simple limit check as well?
not until after calibration. until then you don't know what the rotation radius
of the linkage kit is (and even then, you need some unknown amount of extra
distance to be safe, so it's still a guess.
David Lang
BarbourSmith commented
simple limit check as well?
What about a warning that shows up if the limit check doesn't pass? My concern with a hard limit check which won't let you proceed if the distance is too small is that could lock out some folks from building other designs. For example a hanging plotter with a very light weight pen could operate much closer to the top than a heavy router with weights.
I've added a limit check which spawns a warning popup if the motors are too close to the top of the sheet. You can ignore it, but without clicking "Continue" you won't be able to do anything which seems like a good middle of the road solution to me. Someone building a different type of machine has the option to still use the software, but for most folks the warning will let them know to move the motors up.
What do we think?
davidelang commented
I would only alert if the diagonal distance between the top corner of the
workplace is < (rotation radius + a few sprocket diameters)
I say "a few sprocket diameters" as that is something likely to scale with the
size of other components, and you need some safety margin to keep you from
sucking the linkage into the sprocket.
And the problem isn't that you are too close to the top and the motors may not
have the power, it's that the motors are too close to the workpiece and trying
to go into the top corner will cause the linkage to collide with the sprocket.
The case where things were mounted incorrectly, so the sled moved to the top and
kept pulling was due to incorrect chain lenghs compared to what the system
thought they were no sanity check on the values will help you if the physical
connections don't match the settings.
David Lang
BarbourSmith commented
And the problem isn't that you are too close to the top and the motors may not
have the power, it's that the motors are too close to the workpiece and trying
to go into the top corner will cause the linkage to collide with the sprocket.
I don't think this is true, right @loucksg? The issue you ran into wasn't the sled crashing into one of the motors, it was the tension in the chain being too large at the top of the work area?
loucksg commented
Perfect, that gives balance!
Guy
… On 2 Jun 2018, at 3:28 am, BarbourSmith ***@***.***> wrote:
simple limit check as well?
What about a warning that shows up if the limit check doesn't pass? My concern with a hard limit check which won't let you proceed if the distance is too small is that could lock out some folks from building other designs. For example a hanging plotter with a very light weight pen could operate much closer to the top than a heavy router with weights.
I've added a limit check which spawns a warning popup if the motors are too close to the top of the sheet. You can ignore it, but without clicking "Continue" you won't be able to do anything which seems like a good middle of the road solution to me. Someone building a different type of machine has the option to still use the software, but for most folks the warning will let them know to move the motors up.
<https://user-images.githubusercontent.com/9359447/40854490-874322ce-6586-11e8-8755-5189bc008b30.png>
What do we think?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub <#728 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/Ak-sfvJcSN3YY-6l9esRBQE2ypev9eflks5t4XnHgaJpZM4UPJx8>.
loucksg commented
Unfortunately no, it was literally trying to pull the sled apart.
Sled was in the middle of the sheet, trying to be drawn up closer to the top, and pulled motors to stall. Both motors were pulling against each other trying to raise the sled up above the minimum length, the problem is the motors are too close, I’m welding up some mounts to locate them up higher now. I think the chain was too short on my length estimate.
Guy
… On 2 Jun 2018, at 8:10 am, BarbourSmith ***@***.***> wrote:
And the problem isn't that you are too close to the top and the motors may not
have the power, it's that the motors are too close to the workpiece and trying
to go into the top corner will cause the linkage to collide with the sprocket.
I don't think this is true, right @loucksg <https://github.com/loucksg>? The issue you ran into wasn't the sled crashing into one of the motors, it was the tension in the chain being too large at the top of the work area?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub <#728 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/Ak-sflXbi3Vbod1L6PSzUjQhThZVS_Ajks5t4bu4gaJpZM4UPJx8>.
davidelang commented
On Sat, 2 Jun 2018, loucksg wrote:
Sled was in the middle of the sheet, trying to be drawn up closer to the top, and pulled motors to stall. Both motors were pulling against each other trying to raise the sled up above the minimum length, the problem is the motors are too close, I’m welding up some mounts to locate them up higher now. I think the chain was too short on my length estimate.
two things.
1. the sled should be strong enough to handle the two motors pulling at full
strength
2. the only reason it would do this unexpectedly is if you don't have the
machine in the configuration the software expects it to be in.
In your case, you talked about the chain moving in an unexpected direction, so
you manually moved the chain. I'll bet that left the machine in an unexpected
position and caused the unexpected movement.
David Lang
loucksg commented
1. Agreed, the SLED is actually made of 3/4” marine ply scarp from a shipping container, so it won’t be damaged, the nice little triangulation linkage kit is another story, however if I build a replica of it out of 360 Stainless, let say bring it on baby. However in the first order getting everything setup right will be much better!
2. yes, and the reason for this was the LHS cog chain length feed went in the wrong direction for measuring out the chains even though in the first step it was selected that the chain was fed from the top, the motor wanted the chain on the RHS of the LHS motor - BACK to Front!, RHS on RHS motor correct Feed into sled, requiring some manual juggling to measure the chain, now with the sizes this should be ok, too cold a day to go to the shed though.
LHS Motor —> SLED <— RHS Motor
LHS Chain start at top rotate Clockwise-Cog —> Feed length to sled
RHS Chain start at top, rotate Counter-Clockwise —> Feed length to sled
Problem statement: The LHS motor fed from the RHS of the COG not the LHS, and rotated anti-clockwise!!! As oppose to from the LHS rotating clockwise, the correct position.
The source of the issue was understanding the machine, and the fact that the software drove the chain length on the LHS motor in the wrong direction during calibration, I see there’s new code, so I’ll take a look, now however as I know the exact chain length, these chains are going to get marked, and as I need to raise the motor’s I’ll be welding up some arm extensions from SHS, as hollow section is quite strong, compared to those toque little motors that is ;-), great punch for little 12v units!
I’m going to make sure everything is correct before attacking the linkage kit, as I don’t want this unit ever to come near stall with that on.
Cheers,
Guy
… On 3 Jun 2018, at 10:21 am, David Lang ***@***.***> wrote:
On Sat, 2 Jun 2018, loucksg wrote:
> Sled was in the middle of the sheet, trying to be drawn up closer to the top, and pulled motors to stall. Both motors were pulling against each other trying to raise the sled up above the minimum length, the problem is the motors are too close, I’m welding up some mounts to locate them up higher now. I think the chain was too short on my length estimate.
two things.
1. the sled should be strong enough to handle the two motors pulling at full
strength
2. the only reason it would do this unexpectedly is if you don't have the
machine in the configuration the software expects it to be in.
In your case, you talked about the chain moving in an unexpected direction, so
you manually moved the chain. I'll bet that left the machine in an unexpected
position and caused the unexpected movement.
David Lang
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub <#728 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/Ak-sftgs_RwfCbMLi4YLLfP6KjyYxJUeks5t4ywcgaJpZM4UPJx8>.
davidelang commented
On Sat, 2 Jun 2018, loucksg wrote:
1. Agreed, the SLED is actually made of 3/4” marine ply scarp from a shipping
container, so it won’t be damaged, the nice little triangulation linkage kit
is another story, however if I build a replica of it out of 360 Stainless, let
say bring it on baby. However in the first order getting everything setup
right will be much better!
which linkage kit do you have? both of them are designed to handle the full
strength of both motors
2. yes, and the reason for this was the LHS cog chain length feed went in the
wrong direction for measuring out the chains even though in the first step it
was selected that the chain was fed from the top, the motor wanted the chain
on the RHS of the LHS motor - BACK to Front!, RHS on RHS motor correct Feed
into sled, requiring some manual juggling to measure the chain, now with the
sizes this should be ok, too cold a day to go to the shed though.
LHS Motor —> SLED <— RHS Motor
LHS Chain start at top rotate Clockwise-Cog —> Feed length to sled
RHS Chain start at top, rotate Counter-Clockwise —> Feed length to sled
Problem statement: The LHS motor fed from the RHS of the COG not the LHS, and
rotated anti-clockwise!!! As oppose to from the LHS rotating clockwise, the
correct position.
that's because it knew that you had already fed out chain to measure the
distance, so it's faster to retract the extra than to remove and reset the chain
and feed it out again.