Advanced Search

I pulled the big SSD and installed QTpladmaC and disabled the normal bios stuff. Then ran a test with and without hyper threading on. You can tell which is which from the top left core count. Is one better than the other? Also no cores are isolated. These are at 500sec run times

Just install 2.94 from the downloads page for either the pi or a PC. Just use pnccconf to make a basic config and confirm movement, then cut and paste your current details into it. Take care with that as there are few different things now than 2.7. Mainly the spindle pin is now spindle.0.on

I might have found a clue as to what causes this problem with G33.

I just noticed on my test setup that synchronized motion (eg G33, G33.1) shows this erratic behavior if 'joint.n.index-enable' (where n is the joint for the caxis) is connected to the encoder index-enable and subsequently also to spindle.n.index-enable. This is used to home the caxis to index on startup.

In your rev 1.713 config you have
Seems to me like a bug in how joint index-enable is handled.

[edit]
I just confirmed this on my test bed (caxis is on joint 4):
1. start config and home all (joint 4 is set to home to index)
2. testing  g33 shows broken behavior (does a rapid move to target coordinate)
3. open 'hal configuration' tool and issue HAL command: 'unlinkp joint.4.index-enable'
4. testing  g33 shows expected behavior
5. issue HAL command: 'net c-index-enable joint.4.index-enable'
6. testing  g33 shows broken behavior (does a rapid move to target coordinate)

Big note here: Philips and Maho named the axes differently to the current standard, so your manual may have Z and Y swapped. On My machine with the Heidenhain controller the top slide is the Y axis. I redid my manual to show that, but I may have missed a few places.
Looks like we are on the right path. Here is the page out of my manual for it. It looks exactly as I have been saying.
The feather key is item 6. I'd pull it right out in case it catches just retracting it 6 mm in the instruction. That holds the quill from spinning and sets it's maximum travel. It looks like you will have to set the Y axis full forward to get to it. The drain screw is also visible under there.
I'm not sure about Item 2. Mine has a clamping bolt somewhere there. Take the round black plate (from a previous photo) off the side to look in.
I think you will find the quill assembly nearly identical to the vertical assembly.

David

There is a driver for the omron MX2 VFD.  Or use RS485

exactly.

e.g. analog threshold:

if( value >= anaValue ) X = true
else X = false

or touch button:

if( isButtonPressed(btn, touchX, touchY )
X = true
else
X = false

the index-enable pin is of type 'io' which _is_ bidirectional. All you need to do is move the line 'index-enable = 0; ' inside the 'if (enable)' block as in my version (or possibly just remove it as the user in the post I linked above.

i assume he means the overshooting on the lower right in the last image.

I just tried it and solved the issue for myself for the future.
I have difficulty updating to the new version of qtvcp and probebasic. I decided to postpone this process. I don't have time yet. But in the test configuration I tried the encoding cp1251, if I remember correctly. The test configuration reads my files correctly, but my machine does not work. I hope I am expressing my thoughts clearly.

In the first position of the encode_utils.py file I inserted the cp1251 encoding
But you can check it yourself. At the beginning of the topic I attached a file with Russian text.

CSS is using spindle.speed-in to do something, correct? What does the control loop look like for that? Does it influence the speed-out based on the speed-in feedback or something with the feed based on speed-in?

Geesh the first thing I remeber using a parallel port for was tone generation using a R2R digital to analogue converter (well a buffer connected to 2R2 network), almost a very frequency function generator.
Then using it to power some a steppers via a H-Bridge for a silly little robot platorm. Then some 433MHz remote control, a "sledge hammer" logic analyser knocked up a breadboard, a DCC model train control station, first running under DOS, then wrote a vxd in assembly for win95 to replace the DOS TSR. Probably some other stupid stuff as well.

Never got to play with thing like zip drives, ethernet or sound with the parallel port, used to get most of my stuff 2nd hand at computer markets, this was before ebay took off. Kind of miss wondering around the markets and seeing what exotic stuff you could get.
AT one time I actually used it for connecting to a printer, had to get rid of my HP Laser Jet just last year, to be honest I found it on the side of the road and got a good 10 years of service out of it, made quite a few PCBs using the toner transfer method. That doesn't make sense nowadays when you can get 5 double sided PCB shipped down under for about $7 AUD, just a bit of a wait that's all.

Ahhh that's right I used laplink as well for transferring files. MSODS 6.22 had Intersvr & Interlnk for file transfer aswell.

post your gcode file, i can test it out this week. I havent tested out CSS yet, and ive been trying to figure out how it works. because with threading it runs fine, but with CSS i think it is using the spindle speed-in pin.

If spindle0.speed-in has control on the rpm signal based on the feedback it receives, that would make more sense of your issue.
You removed the lowpass filter gain from the hal file. Im not sure what it defaults to. 0.001 was the gain that worked for me without jumping around, but it took longer to stabilize at speed
Monitor this pin on the halshow, spindle.0.speed-in when trying to css, and see what it looks like.

Your spindle PID in the ini file arent hooked up to anything, so they wont do anything.

RE: spindle gain, sorry about that murphy, I got confused between some configs and got some decimals mixed up. The correct way is 100/maxrpm.

Also when I set the gain to 0.0025 which should be the correct value for 4000rpm. It doesn't scale right. I need to input 0.02 for it to work.

if it works, whatever, but maybe look into it. with gain at .02 that would make 4000rpm 8volts.

Pretty sure 2.7.4 will not work with Mesa 7i96, but it will work with Mesa 5i25 and some others, not important, use 2.7.4 with parallel port and for Mesa 2.8.4 or 2.9.4 or Master.

Thanks for the heads up re: that thread. I'll take a look. And good to know that the RPi5 is so fast. Haven't had a chance to try one yet. Maybe someday. But yes, if my RPi 4 doesn't work out, definitely I'll switch back to a PC. (I don't have high hopes for the RPi.

My current PC still works (in fact, running Wheezy with 2.7.4, from ram off of a thumb drive), but it gives nasty warnings if I try and install something new, and it's old enough that I'd rather replace it on my schedule than wait for it to fail. And with the 7i96, I'll get to clean up some really messy wiring that's waiting to fail, as well. So I'm kind of looking forward to the change.

My "obsession" with parallel port is not a result of using it for CNC machines, oh no, it started in the PC/XT and PC/AT era when i used it for transferring files between PC's, and using floppies was considered "mild torture", had some DOS software named "fx.exe", boot from floppy, run it and choose the send/recieve PC and it moved.
Never failed, ever, and was much, much faster than floppy or serial or InfraRed.
Unfortunately, i can not find it anymore, i did have a need for it last year when i fixed all my old laptops, as one of those has a really old and thick HDD that can not be powered through USB-IDE adapters. Had to use IR...got some more gray hairs!
LinuxCNC (EMC2 back then) only reinforced that obsession.