Thread: Pull Stud Broke - Brother Spindle Trashed

Originally Posted by Finegrain

When I go visit other shops in the area and watch their 40-taper machines do tool changes, I want to lay down and take a nap .

Regards.

Mike

I do the same thing when trying to surface with our vm2 over 100 ipm.

Originally Posted by plastikdreams

I do the same thing when trying to surface with our vm2 over 100 ipm.

Funny thing is: I can't cut an accurate chamfer in my Brother at more than 20ipm

Originally Posted by wheelieking71

Funny thing is: I can't cut an accurate chamfer in my Brother at more than 20ipm

That is a you thing. You just need to turn on the high accuracy modes and program accordingly.

Originally Posted by wheelieking71

Funny thing is: I can't cut an accurate chamfer in my Brother at more than 20ipm

M280

I chamfer with an 8-flute Harvey double-angle cutter @ 192 IPM.

Regards.

Mike

Originally Posted by Finegrain

M280

I chamfer with an 8-flute Harvey double-angle cutter @ 192 IPM.

Regards.

Mike

Plastiks Haas can chamfer at 193...

Originally Posted by 2outof3

That is a you thing. You just need to turn on the high accuracy modes and program accordingly.

Yea, I know its on me (still annoying)

Originally Posted by barbter

Plastiks Haas can chamfer at 193...

I dunno it may, in a straight line.

Originally Posted by wheelieking71

Yea, I know its on me (still annoying)

You should get the V10 firmware, plug in some parameters, and the Speedio now acts like a Haas with regard to high accuracy. With the new system, you can set the machine to automatically run all code in a default accuracy mode- no M codes in the program required.

I’m a little surprised Brother isn’t turning this on from the factory yet. All Speedios should have the new system turned on, dog-leg rapids turned off, and auto thermal turned on when they are delivered.

Originally Posted by gkoenig

You should get the V10 firmware, plug in some parameters, and the Speedio now acts like a Haas with regard to high accuracy. With the new system, you can set the machine to automatically run all code in a default accuracy mode- no M codes in the program required.

I’m a little surprised Brother isn’t turning this on from the factory yet. All Speedios should have the new system turned on, dog-leg rapids turned off, and auto thermal turned on when they are delivered.

My V8.something has background accuracy codes. I don't use them because it's so easy to tell it to go from grandma to racecar to grandma, but yeah for simple work it's nice to not need to go into the code and add M codes.

I haven't turned thermal on. I don't remember how to get into that. I'd like to get that going. My machine does grow a few thou in Z when it's been running for a while. I don't remember if we disabled dog-leg. Mine hasn't exploded yet.

Originally Posted by Finegrain

When I go visit other shops in the area and watch their 40-taper machines do tool changes, I want to lay down and take a nap .

Regards.

Mike

On what, their 1985 YAM's...? Haas toolchange (with a sidemount carousel) is around 3 seconds, a speedio or robodrill'ish 1-1.5 secs. Gee I can almost think what a "oh tools have changed in both machines"...

Don't know Brothers, but ran a robo and yes it is damn fast at 100% rapids 2200ipm. One thing is though, not sure on accell/decel, but those machines are also traveling about half the distance in X/Y at extremes than a "regular" 40 taper is moving (thinking 30-40" in X and 20" in Y). So ya, if you are moving 4" in X and 2" in Y 2200ipm is lightning fast. If you are moving 20" in X and 10" in Y, the comparable (more modern?) 40 taper at 1000 or 1400ipm seems slow....

Originally Posted by wheelieking71

Funny thing is: I can't cut an accurate chamfer in my Brother at more than 20ipm

See this is something that should be brought up more often IMO (thread is way OT now so.. ). You have to either -

1) adjust parameters in these little 30 tapers to get them to cut right
2) slow them down if you don't want to do #1
3) Have to re-think probably everything in programming (more stock on roughing, etc)

Not saying that is necessarily bad, but it should be #2 on salesman list of things to discuss. "Hey we know you run finishing at 50ipm in your xxx machine, but to do that in a Brother you need to program an Mxxx code for accuracy"

So it's great to have so many people thinking "what is a better pull stud design/materiel" , but I still think the pull stud was not the problem. I think it was a machine crash (plain and simple) that broke the pull stud. Has OP been back to discuss further?

Originally Posted by Mike1974

On what, their 1985 YAM's...? Haas toolchange (with a sidemount carousel) is around 3 seconds, a speedio or robodrill'ish 1-1.5 secs. Gee I can almost think what a "oh tools have changed in both machines"...

Don't know Brothers, but ran a robo and yes it is damn fast at 100% rapids 2200ipm. One thing is though, not sure on accell/decel, but those machines are also traveling about half the distance in X/Y at extremes than a "regular" 40 taper is moving (thinking 30-40" in X and 20" in Y). So ya, if you are moving 4" in X and 2" in Y 2200ipm is lightning fast. If you are moving 20" in X and 10" in Y, the comparable (more modern?) 40 taper at 1000 or 1400ipm seems slow....

I kinda doubt the Haas chip to chip is as close.

everyone focuses on rapid speed, but it is acceleration that makes the time disappear.

My DMG is a .5 G machine with a 2700 IPM X rapid, and machinists lose their minds when they watch it.

The dynamics of a multi peck countersink look like a ball bouncing.[think of a superball as it devolves to zero height]

I think Speedios are like 2 G. Add in that all it has to do is rapid up zMAX and change.

They are at full rapid in like .05 seconds or some silly crap, while my machine takes a turtle like quarter of a second

Another perspective, you know how an ooooold CNC looks like it stops at the top of a hole or the end of a non interpolated move?

It is not stopping, it is decelerating

Originally Posted by Mike1974

See this is something that should be brought up more often IMO (thread is way OT now so.. ). You have to either -

1) adjust parameters in these little 30 tapers to get them to cut right
2) slow them down if you don't want to do #1
3) Have to re-think probably everything in programming (more stock on roughing, etc)

Not saying that is necessarily bad, but it should be #2 on salesman list of things to discuss. "Hey we know you run finishing at 50ipm in your xxx machine, but to do that in a Brother you need to program an Mxxx code for accuracy"

EGGFUGGINZACTLY!!!!!!!!!!!!!!!!!!!!! I had no damn clue. All I was told was basically: "it will go as fast as you want"

Ahem, as delivered? Hell no.

Originally Posted by Mike1974

On what, their 1985 YAM's...? Haas toolchange (with a sidemount carousel) is around 3 seconds, a speedio or robodrill'ish 1-1.5 secs. Gee I can almost think what a "oh tools have changed in both machines"...

Don't know Brothers, but ran a robo and yes it is damn fast at 100% rapids 2200ipm. One thing is though, not sure on accell/decel, but those machines are also traveling about half the distance in X/Y at extremes than a "regular" 40 taper is moving (thinking 30-40" in X and 20" in Y). So ya, if you are moving 4" in X and 2" in Y 2200ipm is lightning fast. If you are moving 20" in X and 10" in Y, the comparable (more modern?) 40 taper at 1000 or 1400ipm seems slow....

I am running a Kitamura HX-500iG during most days, setting up a new production line/part. It is advertised as the fastest rapids in the world (IIRC for a box way machine?) and lightning fast tool changes. I want to lay down and take a nap when running that machine too. I don't give the slightest of shits about tool change speed. I care about chip to chip. If it takes 30 seconds to get back into the cut, who cares that the tool arm rotated in 0.2 seconds?

In the case of the Kitamura, it takes a while to decelerate the spindle. It has fast rapids, but it never achieves them because it is really sloooow to accelerate. Then it waits for a moment and thinks upon reaching the tool change position, then does the change, then it thinks for a moment and returns to the cut. The advertised 2.3 second (or whatever) tool change time is utterly irrelevant, because realistically it's 7-10 seconds chip to chip. Doesn't seem too bad, but when you're changing lots of tools in a part run, it adds many minutes to cycle time. If you're grabbing a tool from the 200-tool magazine, forget it... 20-30 seconds easily. Yeah, I know you can pre-stage tools, except you can't... because anything over 40IPM requires the high accuracy mode to be on, which flushes the look ahead buffer to avoid parameter writes during look-ahead... which means pre-staging the tool causes the spindle to freeze motion for many seconds while it puts the tool back or retrieves the tool from the library.

It is a joy to run the Speedio. It's so, so fast. For everything. Even probing routines run MUCH faster. It's a great user experience. The only thing I prefer about the Kitamura is the touch screen display.

Good grief, what year is that?


-----------------

Think Snow Eh!
Ox

Originally Posted by SRT Mike

I am running a Kitamura HX-500iG during most days, setting up a new production line/part. It is advertised as the fastest rapids in the world (IIRC for a box way machine?) and lightning fast tool changes. I want to lay down and take a nap when running that machine too. I don't give the slightest of shits about tool change speed. I care about chip to chip. If it takes 30 seconds to get back into the cut, who cares that the tool arm rotated in 0.2 seconds?

In the case of the Kitamura, it takes a while to decelerate the spindle. It has fast rapids, but it never achieves them because it is really sloooow to accelerate. Then it waits for a moment and thinks upon reaching the tool change position, then does the change, then it thinks for a moment and returns to the cut. The advertised 2.3 second (or whatever) tool change time is utterly irrelevant, because realistically it's 7-10 seconds chip to chip. Doesn't seem too bad, but when you're changing lots of tools in a part run, it adds many minutes to cycle time. If you're grabbing a tool from the 200-tool magazine, forget it... 20-30 seconds easily. Yeah, I know you can pre-stage tools, except you can't... because anything over 40IPM requires the high accuracy mode to be on, which flushes the look ahead buffer to avoid parameter writes during look-ahead... which means pre-staging the tool causes the spindle to freeze motion for many seconds while it puts the tool back or retrieves the tool from the library.

It is a joy to run the Speedio. It's so, so fast. For everything. Even probing routines run MUCH faster. It's a great user experience. The only thing I prefer about the Kitamura is the touch screen display.

That is as bad as the old Fadal I ran.

Originally Posted by wheelieking71

EGGFUGGINZACTLY!!!!!!!!!!!!!!!!!!!!! I had no damn clue. All I was told was basically: "it will go as fast as you want"

Ahem, as delivered? Hell no.

I've been bugging Yamazen for some time now that two things need to happen:

1- Part of the instillation procedure needs to be setting up the new high accuracy stuff, thermal comp, and turning off dog-leg rapids. Andy's crew on the West Coast is pretty proactive on this, but other markets? Not so much.

2- Document a more practical guide to everyday use than what the manuals provide. It would be nice of the Speedio folks had a simple, clear resource for doing common stuff.

I work pretty closely with Andy, and one of the barriers to all this is that the mothership in Kyria keeps upgrading/changing stuff. I had a whole thing set up documenting the differences between Mode A and Mode B/BII, with use guides and parameters, even finagled Autodesk to integrate some logic into their HSM/Fusion post... right as Brother up and installs a much improved, simpler, more standardized High Accuracy setting system. Brother does a great job pushing these machines forward and listening to customers quite intently (see, many of the nice details they've added to the X2 machines), but they do a very poor job documenting things or keeping Yamazen in the loop. It is frustrating when they have 50 applications engineers who all have tremendous knowledge in their heads, but no central database of guidelines/parameters/information that we can all draw from.

So we wind up with situations like this where Brother's old High Accuracy system was a powerful, finicky, obscure thing that many found unusable out of the box without a lot of work. So they fixed it! But didn't bother to give anyone guidance, so people are still under the impression that the system is finicky and complicated and wind up like Wheelie, where they just live with frustration. Literally 5 minutes in front of his machine and I could have his issue solved.

The new machines have things like tool load monitoring, G68.2, Rotary Fixture Offsets, new Inverse Time Feed, and it appears the new M200/M300 even have full 5 axis as an option (the catalog offers "Involute Cutting", which sorta sounds like 5 axis sim to me). If Brother doesn't step up and document this stuff to Yamazen the way they should, all of us (customers, and Yamazen AEs) are going to be stuck reading through clear as mud Japanese manuals trying to figure this all out. That would suck.

Originally Posted by Ox

Good grief, what year is that?


-----------------

Think Snow Eh!
Ox

2019... installed in September, the latest and greatest with Arumatik (Mitsubish M850) control.

?!?! - so, the Kitamura has controller debug mode on and is spewing internal status down a serial line somewhere? Been misconfigured so it thinks it only has 16 bytes of memory?