two independent X at the same time.
08 Nov 2010 08:49 - 08 Nov 2010 13:53 #5154
by Gelar
Replied by Gelar on topic Re:two independent X at the same time.
finally, yes, everything works with no FW. it was my childish inattentiveness and haste, i apologize for them. don't know how to thank you.
yep ) program runs okay, but now we have a driver (L297) error. maybe not right pins.
yep ) program runs okay, but now we have a driver (L297) error. maybe not right pins.
Last edit: 08 Nov 2010 13:53 by Gelar. Reason: problems aren't finished/
Please Log in or Create an account to join the conversation.
- step4linux
- Offline
- Premium Member
-
Less
More
- Posts: 115
- Thank you received: 5
08 Nov 2010 13:48 #5162
by step4linux
Replied by step4linux on topic Re:two independent X at the same time.
no reason to apologize, nobody is perfect.
Let us know about progress of your work.
Gerd
Let us know about progress of your work.
Gerd
Please Log in or Create an account to join the conversation.
08 Nov 2010 15:33 #5163
by andypugh
Replied by andypugh on topic Re:two independent X at the same time.
Gelar wrote:
I think that the error must be being triggered somewhere other than in the L297 as (as far as I can see) that chip doesn't have any fault outputs.
but now we have a driver (L297) error. maybe not right pins.
I think that the error must be being triggered somewhere other than in the L297 as (as far as I can see) that chip doesn't have any fault outputs.
Please Log in or Create an account to join the conversation.
08 Nov 2010 21:49 - 09 Nov 2010 07:23 #5168
by Gelar
Replied by Gelar on topic Re:two independent X at the same time.
yep, scope said "you've mixed up step and dir pins" for new pair of axes. although it should not happen. l297 wasn't bought, it was made by ourselves.
Definetely the reason was pins. Now the machine works as it is supposed to. All the axes are moving.
i tried simple action g91 g1 f800 x100 y100 u50 v50 and it made it in the same time although manual says that all velocities are equal so UV had to go 50mm faster than XY 100mm (by time). but it is not. it is good for me, but again i don't understand or misunderstood manual. or... this is what manual means by phrase UV velocity will depend on XY velocity (as a master).
Definetely the reason was pins. Now the machine works as it is supposed to. All the axes are moving.
i tried simple action g91 g1 f800 x100 y100 u50 v50 and it made it in the same time although manual says that all velocities are equal so UV had to go 50mm faster than XY 100mm (by time). but it is not. it is good for me, but again i don't understand or misunderstood manual. or... this is what manual means by phrase UV velocity will depend on XY velocity (as a master).
Last edit: 09 Nov 2010 07:23 by Gelar.
Please Log in or Create an account to join the conversation.
09 Nov 2010 22:03 #5187
by Gelar
Replied by Gelar on topic Re:two independent X at the same time.
oh, no. although linear movements work good, arcs have problems. IJ can be used only once per line, thus i can't do tasks i was making it all for. are there any tricks for that? manual does not have any.
Please Log in or Create an account to join the conversation.
09 Nov 2010 22:20 #5189
by andypugh
Replied by andypugh on topic Re:two independent X at the same time.
What it means is that the UV move will finish at exactly the same time as the XY move.
I would expect that XY would slow down if UV can't keep up, but have never tested it.
As for your G2 G3 Arc issue, it is unfortunate but yes, you can only do linear moves if you are moving both XY and UV. It is not easy to see a way round that problem either, as there are no spare letters available to use for UVW centres.
It ought to be possible to do two arcs at the same radius (sharing the R word) but that is not necessarily what is wanted either.
You will need some CAM software (or simple homebrew scripts) to convert the curves to short straight lines. The good news is that the EMC2 "naive CAM detector" will smooth out the lines into curves.
I would expect that XY would slow down if UV can't keep up, but have never tested it.
As for your G2 G3 Arc issue, it is unfortunate but yes, you can only do linear moves if you are moving both XY and UV. It is not easy to see a way round that problem either, as there are no spare letters available to use for UVW centres.
It ought to be possible to do two arcs at the same radius (sharing the R word) but that is not necessarily what is wanted either.
You will need some CAM software (or simple homebrew scripts) to convert the curves to short straight lines. The good news is that the EMC2 "naive CAM detector" will smooth out the lines into curves.
Please Log in or Create an account to join the conversation.
11 Nov 2010 05:53 #5208
by Gelar
Replied by Gelar on topic Re:two independent X at the same time.
yesterday i have cut different cones (different radii from each side). everything is okay. the only one thing i'd like to have is a 3D plot, just to see what's happening to not reckon only on a spacial imagination. but emc doesn't know distances between machine levers.
Please Log in or Create an account to join the conversation.
- step4linux
- Offline
- Premium Member
-
Less
More
- Posts: 115
- Thank you received: 5
11 Nov 2010 06:21 #5209
by step4linux
Replied by step4linux on topic Re:two independent X at the same time.
When you cut cones with the small side at XY, the UV side moves faster and you must try to calculate the required speed at XY to avoid too high speed at UV.
With pure circular cones, it is not a problem to swap XY and UV in gcode to move the big side at XY. With non-circular cones (with an arbitrary profile) this will produce a mirrored shape. I made some hal files to swap XY and UV at the machine too. As a result the cone is mirrored twice and back at its original shape.
Additionally in some cases you need two equal but mirrored parts (e.g. two sides of a wing for a model aircraft). I can cut these from one g-code, just mirroring the machine.
Gerd
With pure circular cones, it is not a problem to swap XY and UV in gcode to move the big side at XY. With non-circular cones (with an arbitrary profile) this will produce a mirrored shape. I made some hal files to swap XY and UV at the machine too. As a result the cone is mirrored twice and back at its original shape.
Additionally in some cases you need two equal but mirrored parts (e.g. two sides of a wing for a model aircraft). I can cut these from one g-code, just mirroring the machine.
Gerd
Please Log in or Create an account to join the conversation.
11 Nov 2010 09:11 #5214
by Gelar
Replied by Gelar on topic Re:two independent X at the same time.
faster? nope. from UV side i cut circle with smaller radius. what really needs to be calculated is a "real" radius that will be cut because the length of the NiCr string (from its holder with spring) is not equal to that of a foam block.
why mirroring axes if you can just set appropriate coordinates for UV?
why mirroring axes if you can just set appropriate coordinates for UV?
Please Log in or Create an account to join the conversation.
- step4linux
- Offline
- Premium Member
-
Less
More
- Posts: 115
- Thank you received: 5
11 Nov 2010 10:08 #5217
by step4linux
Replied by step4linux on topic Re:two independent X at the same time.
as long as your cone has a circular cross section its ok.
But when the cone has a non-circular, unsymmetric cross section, you cannot just swap XY / UV. This would make a different part.
But when the cone has a non-circular, unsymmetric cross section, you cannot just swap XY / UV. This would make a different part.
Please Log in or Create an account to join the conversation.
Time to create page: 0.089 seconds