Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: TonyP on November 21, 2007, 04:34:50 PM
-
I've been playing with the turn wizards and have come across a problem with the OD arcs. When I set up say, a hemisphere, on the end of a bar using OD Arc, the code generated is a G03. This produces a toolpath which goes all the way round through the work. If I manually change all the G03's to G02's all is ok. The same applies to the OD Arc2. I haven't tried any of the other arc wizards yet.
This is on V 2.52, which is on my lathe PC. I have the latest version V 2.6 on another PC, just to play with, but I find that if I go into the wizards they won't post code and I can't get out of the wizard program.
Is it just me?
Tony
-
Had the same problem, I can't remember exactly what it was but I think it was related to me having the option on the ports and pins, turn options page for reversed arcs checked. I have a rear toolpost so it could be that you need it checked. Have a look in any case and either check or uncheck it and see what happens.
Hood
-
Hood,
thanks for the reply. Its certainly to do with the reversed arcs & rear toolpost. If I check reversed arcs & specify a front tool all is well. I haven't found a combination yet that works for rear tools.
I found the cause of my problem with the wizards locking out on V2.60 was that I hadn't re-loaded the wizards when doing the upgrade. They work just the same as the old version now.
Tony
-
It suddenly ocurred to me that I don't know what the covention of 'front' & rear is on a slant bed lathe. On my TCL125 the x axis goes more +ve as the slide moves up. I mostly use a tool above the work, so an infeed is going negative. Do I call that a front or a rear toolpost? I've assumed it to be rear but only in relation to normal lathes. This may be my problem with the arcs.
Tony
-
Tony
I am waiting on more tooling for my lathe so I have not done too much testing yet. I have only used the rear turret so far, hopefully next week I should be getting into using the front post so will let you know what I find out.
Have you made any progress/conclusions?
Hood
-
Hood
I've found that if I check 'reversed arcs' and specify a front tool, the code generated by the wizard runs fine. Any other combination gives silly results. This makes me think that what looks like a rear tool on the TCL is really a front tool, and I have the convention wrong. There is a bit in the Mach turn manual about 'looking at a slant bed lathe while lying underneath' which sort of supports that theory.
Still on front & rear toolposts, I'm convinced that there is something wrong with the application of offsets in the toolsetting menu. I can't yet describe it properly, but its possible to corrupt offsets in one tool by setting another in the opposite toolpost.
My son and I playing for an hour the other day, trying to make sense of it, with 3 carefully labelled tools on a toolplate. We set a master tool & jogged back and forth trying to achieve 2 front tools & one back. Anything set as front works correctly when switching tools, but as soon as a rear tool is introduced everything goes wrong. The distance between front & rear toolposts also comes into play, just to confuse matters.
I know the manual does not recommend switching toolposts dynamicly, but I think it just doesn't work. It is some time since I heard of any work being done on the Turn program - certainly there's been nothing in the release notes.
Tony
-
Hood
Further to my previous post- I spent another frustrating hour or so playing offline with offsets etc.
The first observation is that the x offset dro on the main tools screen seems to read diameter while the offset in the table is halved, for radial movement I suppose. ( machine is in diameter mode).
If I set tool 1 as master and zero everything ( machine co-ords etc.), do a move to say X50, Z50 switch to tool 2, touch x & z & all makes sense. I get a table offset of x -25 z -50. I can switch backward & forwards between tools, do moves, return to tool1 g0 x0 z0 and still have all readings at zero.
The difficulty is then with tool3 which is say, a parting tool in a rear toolpost. So I declare tool 3 as a rear tool & set the distance between toolposts to -100mm, say. I move to x00 z00 in tool 1, change to tool 3 & touch x & z.I rather expected the axis coordinates to go to -100, but no I get -150! Bear in mind that nothing has actually moved here, I'm just changing tools backwards & forwards. Now for the fun bit - I change back to tool1. Do I get back to Zero? No it's -50. The table offsets have remained at zero for all but tool 2 in all this.There seems to be some interaction between tool2 & 3 so it's possible to get all sorts of combinations in the DRO. You can even catch it changing it's mind now & then.
I find all of this extremely confusing (probably you do too). The answer is that you really can't switch toolposts in the current version, but it would be really useful to be able to do so.
Tony
-
I remember reading recently on the Yahoo group that Art said that front and rear toolposts weren't working correctly but I talked with Brian a bit later and he seemed to think they were. We didnt have time to discuss things and havent spoken about it since so I am not sure what the situation is although your testings seem to suggest its not working.
Hood
-
When you set a tool in turn, make sure you set it as the true diameter of a reference diameter, for example, if you have an exact 1.000" diameter ground pin in the chuck, set the tool as 1" in diameter and don't set that as zero. A lathe always uses X zero as the center of the part. I saw a post a few weeks ago where somebody was setting X zero on a reference part diameter and set that to zero. This way you would certainly have problems with X axis offsets between front and rear toolposts.
DaveA.
-
Yes I can see that would be a problem, but in this case I was effectively just locating tool tips, without any other distractions. The way I understand offsets is that they are the numbers to be added to the master tool location to place the new tool tip in the same location. Obviously you don't actually want to place the tips of a front & rear tool in the same position for any other condition but zero diameter, as they will normally cut from opposite sides of the stock.
By the way, I haven't had a definitive answer to my question as to the 'front' toolpost of a slant bed lathe. Which is it?
Tony
-
I have my TCL set up with a toolplate and just set a nominal X zero Y zero as an origin by setting the Mach3 displays to zero. This could be done with any tool. This sets the origin like on graph paper. All other tools are then jogged to their start point and the values read off on the Mach3 screen, these values are then used in my G code. The code is a bit scruffy but it is quick to do. This method would not work if you needed to use wizards but it eliminates any confusion over front and rear toolposts. Not tried to do radii yet but assume it would just be a matter of establishing the start point on the "graph paper" from the mach3 screen display and deciding whether to do a concave or a convex cut. It should be possible to incorporate these values in the tool table as offsets and I may do this just to show it can be done.
The spindle always runs clockwise, so my "front" tool is mounted upside down, compared to the parting tool.
The code for my little video is on the Yahoo Mach group files as "Turret Video.txt"
Ian
-
Tony,
I forgot I had posted a reply to you in this thread :-
http://www.machsupport.com/forum/index.php/topic,4699.0.html
Repeating an old story!
Ian
-
Ian,
while we are in contact, I believe you have an ATC for the TCL. Would it be possible to have a few basic dimensions of the rotary tooplate? Also is it's centre hieght on the spindle centre? I'm in the process of trying to come up with a design for mine and it would save some time if I knew the limiting dimensions. I've never examined one closely, but from photos it look as though the tools are retained by a wedge action. Is this so?
Tony
-
Tony,
The turret has a new home now but I can get you any sizes that you may need. Centre height is the same as your current tools. Turret did have a wedge system for holding the turning tools, so that you could change tools without indexing but since you had to index to get at the grub screws for the drilling, it seemed an over complication.
I found it non user friendly and have gone over to a toolplate system. Have you seen my video of the hosetail? Using the turret it took about 60 seconds. With improvements to my TCL, this same job now takes only 35 seconds with a toolplate setup. I will try to do a video (not another video!!!!) in the next few days.
Ian
-
Thanks Ian, I don't want to put you to a lot of trouble, but overall diameter etc. would be a useful starting point. It would be nice to just build something that already works without designing everything from scratch. I don't want to use compressed air so it'll be a worm & ratchet or something similar.
From my limited experiments I find the toolplate rather restrictive and time consuming to set up. I think I would get on better with a set of tools that I can use for general purposes & quick setups. I don't envisage vast production runs of anything.
Tony
-
Tony,
I have sent you a PM.
Ian
-
Tony,
I found the picture of the Boxford ratchet!
Ian
-
Hello Ian,
Very nifty ratchet and pawl. Can you tell me the sequence of operation ?
Looks like a worm drive for the motor but what locks the turret against the pawl or in position after indexing ? Detents or a pin ?
Does the motor hold it ?
Thanks
-
Thanks for that, Ian. I imagine the stepper is a size 23, so I can probable scale the picture from that. If the outside pair of holes in the feet of the casting that was used for fixing then they are 50mm apart. Also the centre height will be 35mm. Its certainly quite a substantial ratchet. Presumably you haven't a similar picture of the other side?
I think there is probably a lot of interest in the group in toolchangers generally, and it might be worth trying to collect any available information together in one place. Not sure how we do that.
Tony
-
I'll add a couple of pictures of the boxford turret that I've found elswhere.
Tony
-
It won't let me attach two pix at once apparently
-
Tony,
Yes, stepper was size 23. The turret spindle was mounted in two ball bearings with the toolholder plate bolted onto a flange. Worm wheel ratio was 30:1.
Your picture shows an additional microswitch, presumably to generate a position signal for tool number 1.
On the Yahoo Mach group John Stevenson suggested using a rotary table with the ratchet mounted behind a toolplate. I though that it might be possible to use this idea,
but mount it directly to a stepper. The ratchet would take all the downward forces on the cutting tool and the side forces are very low on these little lathes. One of the Chinese 250 microstep drivers would give enough resolution with enough torque at low speed say, 12 rpm. Just an idea for discussion.
Ian
-
Overloaded,
Very simple really, the stepper indexes the plate to the next ratchet position then backs up and stalls. Step singals are stopped and position is held by motor detent torque. No other locking pins are used. I just wrote a subroutine call to index the turret.
Ian
-
Ian
I just tried a quick scaling exercise and it looks as though using the centre height for the scaling is incorrect. I can't get the same scale factor for the centre height & the stepper size. Looking at my turret1 pic I notice that the tool tips do not appear to lie on a line through the axis of the turret. Scaling from just the stepper gives me an axis height of 45mm above the cross slide & a possible toolplate dia of about 140 -160 mm. Does this sound about right?
Regarding the idea of using a rotary table, I think it would work well, but for a small lathe like the TCL it may be difficult to find one with a small enough centre distance. I've been thinking about the same sort of thing myself, but I'm sort of becoming resigned to having to make one from scratch.
Tony
-
Tony,
On the TCL 125, the turret toolplate is 132mm diameter and the centre height of the turning tools from the crosslide is 35.25mm. The TCL 240 (pictured) is a very much bigger machinne.
Ian
-
Ian,
ah! Thanks for that. I didn't know that there was a difference in capacity. I thought that they'd usedthe same basic mechanics with updated control layout & electronics. ( or was that just the 160 )
Tony