Rigid Tapping on a PCNC1100

[TurboStep]

As should be apparent from my previous posts on the Tormach forum I'm not a fan of T/C Tapping and use thread milling exclusively. Now PathPilot is opening up another option - Rigid Tapping!
We know that LinuxCNC is capable of this but Tormach hasn't yet provided the necessary hardware and software extensions (the 1100 Mk4 can't be far away but who knows what it will bring ). Anyway, I just couldn't resist, and it turned out to be fairly easy. There are however a couple of issues that I still need to clear up. Here's a short video of my initial proof of concept. Now I've proven to myself that it works I can invest more effort an $ in a more professional approach....



Step

Similar Threads:

• Need Help!- Rigid tapping errors without rigid tapping installed
• What exactly is Rigid tapping? Why people always ask does it do rigid tapping?
• tapping on a mini mill without rigid tapping???
• Tapping head or rigid tapping
• Rigid tapping or tapping head

[MFchief]

Great video Step, I think you're going to spark a lot of interest with this.

Terry

[wtopace]

Very well done! I'm guessing you were able to read directly off the parallel port from within LCNC?

I haven't looked at the Tormach control board at all, but any thoughts on passing your new data line through the existing Tormach controller board through to the Mesa card?

I wasn't sure whether or not the Tormach would have the necessary deceleration / acceleration to do rigid tapping, but it's good to see that you had success! I also know nothing about rigid tapping, so this might not even be an issue.

[tmarks11]

awesome!

[mountaindew]

I have said this before! Your working knowledge of these machines is awesome.
I always read and take notes from your posts. Well done!

[jcposada]

Very nice! I can't wait to see more. Thanks for sharing.

[samco]

the 5i25/6i25 has 2 ports. I assume tormach is only using the rear one. If you are lucky the firmware in the mesa card has at least 1 encoder counter on the second port. If not - then you would need different firmware. Don't know how easy that would be in the pathpilot setup.

Did you add a base thread or are you only counting in the servo thread?

sam

[TurboStep]

@wtopace: like with the standard parallel port the Mesa card only provides 17 pins per port for input/output and those on the external connector are already fully used. As sam mentioned, the card does have a second port and I initially intended to use this but for the first prototype I decided to go for the parallel port to reduce the number of unknowns - this is all new-land for me as well!
@ sam: I added a base thread. From the information I've found the card doesn't appear to allocate the resources to either port (correct me if I'm wrong) so if an encoder has been provided at all by the firmware I'm presuming that it would also be available for the second port.
My next step is to replace my makeshift encoder and maybe Tormach will have released something comparable by then.
Step

[ltborg]

Awesome work! Any updates on the new encoder or the modifications to PathPilot required?

[TurboStep]

Awesome work! Any updates on the new encoder or the modifications to PathPilot required?

Thanks, but there wasn't that much to it. The biggest issue was getting the post to generate suitable code and finding a fix for the strange behaviour (see Youtube comments) which is still there on the latest PP versions, although I haven't tried 1.9.4b yet. If I remember correctly the only changes neede to PP were in the HAL to configure the parallel port and set up the encoder. I basically followed the standard configurations described in the cnclinux pages.

I've made no further progress for several reasons, including:
It aint broke... so why fix it!
I only wanted to try rigid tapping because I didn't have a suitable thread mill for 1/4-20 (I use metric multi form thread mills up to 8mm) but I had an unused 1/4-20 tap lying around "just in case", so I actually don't have much use for rigid tapping.
And then I don't actually like using taps! I much prefer thread milling because it fits better into my workflow. Thread milling is normally the very last operation and doesn't throw up any strings of metal that might scratch a finished surface.

Finally, I'm busy with other projects like a physical feed rate override knob (inspired by this thread http://www.cnczone.com/forums/tormac...ml#post1773868). I thought it would be fun, even though I didn't think I needed it. Now that I've tried it I find it really quite cool - a video will appear sometime - when I finally get around to it. I'm now working on a simple controller for a multi-nozzle Fogbuster type system inspired by the Datron spray mist coo!lant system. PathPilot is SOOO much more fun than Mach3 ever was!
Step

[samco]

huh. the rising z was an issue with the new TP. I know it was fixed in linuxcnc - maybe it never got back to pathpilot.

sam

[TurboStep]

huh. the rising z was an issue with the new TP. I know it was fixed in linuxcnc - maybe it never got back to pathpilot.

sam

Thanks for the feedback. I'll check it out when I've merged into 1.9.4b.
Step

[ltborg]

Thanks, but there wasn't that much to it. The biggest issue was getting the post to generate suitable code and finding a fix for the strange behaviour (see Youtube comments) which is still there on the latest PP versions, although I haven't tried 1.9.4b yet. If I remember correctly the only changes neede to PP were in the HAL to configure the parallel port and set up the encoder. I basically followed the standard configurations described in the cnclinux pages.

I've made no further progress for several reasons, including:
It aint broke... so why fix it!
I only wanted to try rigid tapping because I didn't have a suitable thread mill for 1/4-20 (I use metric multi form thread mills up to 8mm) but I had an unused 1/4-20 tap lying around "just in case", so I actually don't have much use for rigid tapping.
And then I don't actually like using taps! I much prefer thread milling because it fits better into my workflow. Thread milling is normally the very last operation and doesn't throw up any strings of metal that might scratch a finished surface.

Finally, I'm busy with other projects like a physical feed rate override knob (inspired by this thread http://www.cnczone.com/forums/tormac...ml#post1773868). I thought it would be fun, even though I didn't think I needed it. Now that I've tried it I find it really quite cool - a video will appear sometime - when I finally get around to it. I'm now working on a simple controller for a multi-nozzle Fogbuster type system inspired by the Datron spray mist coo!lant system. PathPilot is SOOO much more fun than Mach3 ever was!
Step

Thanks for the feedback. I didn't realize it was just a matter of changing things like LinuxCNC. I'll do some searching then for the code. Thanks again!

[tmarks11]

TS-

You should try to sell your tapping mod to Tormach. I can't believe that the profit margin on T/C heads and the like is high enough that they wouldn't be interested in a commercialization of what you did.

Maybe no money in it for you, but at least you would get bragging rights.

I also can't believe that they haven't already got something like this in the works; after all, it is an application of the same principles that they use on the 15L lathe.

[TurboStep]

TS-

You should try to sell your tapping mod to Tormach. I can't believe that the profit margin on T/C heads and the like is high enough that they wouldn't be interested in a commercialization of what you did.

Maybe no money in it for you, but at least you would get bragging rights.

I also can't believe that they haven't already got something like this in the works; after all, it is an application of the same principles that they use on the 15L lathe.

I'd also be very surprised if Tormach hadn't done this long ago. There maybe reasons why they feel that the Tormach (or the market segment) isn't suitable or perhaps this might simply follow after work the PCNC440 accessories is concluded.
The implementation for the mill and the lathe shouldn't be any different except that additional connections between the PC and mill will be necessary - this may be an issue that is holding them back.
I merged my changes into v1.9.4b (I found some interesting new features ) but my workshop went under water the day after so it will take a while before I'll be able to test it!
Step

[mountaindew]

v1.9.4b (I found some interesting new features ) but my workshop went under water the day after !
Step

I Always make note when step finds interesting features .
Under water , what snow melt?

[tmarks11]

...except that additional connections between the PC and mill will be necessary - this may be an issue that is holding them back.

True, they would need to add encoder connections to their BOB (The 15L BOB has the encoder connection, but is short a few axis). That would be a bit of a bigger re-wiring chore if they sold a modification them most of the stuff they sell. A second BOB and a second parallel port cable would be easier, but they probably wouldn't go that route.

[jttoner]

I'd also be very surprised if Tormach hadn't done this long ago. There maybe reasons why they feel that the Tormach (or the market segment) isn't suitable or perhaps this might simply follow after work the PCNC440 accessories is concluded.
The implementation for the mill and the lathe shouldn't be any different except that additional connections between the PC and mill will be necessary - this may be an issue that is holding them back.
I merged my changes into v1.9.4b (I found some interesting new features ) but my workshop went under water the day after so it will take a while before I'll be able to test it!
Step

Maybe becaue it would kill sales of their compression thing for tapping.

[kstrauss]

.
I merged my changes into v1.9.4b (I found some interesting new features ) but my workshop went under water the day after so it will take a while before I'll be able to test it!
Step

Please don't keep us guessing about what is new!

[tmarks11]

Maybe becaue it would kill sales of their compression thing for tapping.

I doubt the profit margin on the T/C tapping head is enough to make that a consideration. I suspect they will get around to it eventually, partly because at least one competitor already offers it standard on their machine.

Anybody bought a new 1100 recently? It would be interesting to see if the BOB on the new machines has any un-used terminal blocks that the older machines don't have. Sort of like the Tormach controllers sold for a year had a "future use only" tag over the Mesa card parallel port connector.