Advanced Search

Hi all,

I’m trying to troubleshoot a Heidenhain LS403 linear encoder (1Vpp analog output) that’s connected via a Heidenhain 1Vpp-to-TTL converter to a Mesa card in my LinuxCNC setup.

We weren’t seeing any motion or position data in HAL, so I decided to check whether the encoder is still producing valid analog signals.

I’m fairly new to working with encoders — and also still learning my way around the oscilloscope — so I might be missing something obvious. That said, here’s what I’ve done so far:

• The encoder is powered via the Mesa card: 5 V and 0 V are confirmed and stable.
• For testing, the signal wires are disconnected from the Mesa. I’m only using it to supply power.
• Based on voltage levels and scope behavior while moving the encoder, I’ve identified A+, A–, B+, B–, R+, R–.
• When I probe A+ or B+ relative to 0 V, I see a flat ~2.0 V line. A– and B– show ~1.5 V. Moving the encoder slightly only causes minor fuzz — no visible sine wave.
• When I move the encoder head, the line does become “noisier” — in the sense that it gets visibly fuzzy — but I never see a clean sine wave.
• I also tried using two probes and a CH1 - CH2 math channel on a Siglent SDS1104X-E to simulate differential probing, but still no sine wave appears.
• I’ve played with timebase, voltage scale, trigger settings, etc., without much change.
  
  At this point I’m unsure whether:
  • The encoder might be dead,
  • The signal needs a specific load to show up properly,
  • Or I’m simply not probing it correctly.
    
    Any tips or guidance would be very welcome — especially from folks who’ve scoped Heidenhain 1Vpp encoders before. Thanks in advance!
    

Just another off topic Milling machine vs router anecdote.

I bought a 2nd hand "what I thought" was a milling machine. I later was informed "Thats not a milling machine" (Indiana Jones rings a bell?)
So I learned everything below 5 tons is "not a milling machine".
Criteria for milling machine min. BT40 or HSK63, min. 15kW spindle power (continous load) tool changer

Point beeing, as long as it does what we want it does not matter much (all good).
For communication I now call my supposed mill a "tapping center", because then people know what I mean.

Joking aside I think the difference for hobby users is the ratio between z-height and x-y travel. A router is more flat and large a milling machine more a cube with a z-range in the magnitude of the x-y travel.

I sure did look at your thread and files.

What I am basically looking for is a minimal hal file that will allow PP to start. I don't know which parts of the various hal files posted are mandatory and needed to get PP to launch, leaving out anything not mandatory.

In the end I will use it on a bridgeport that has servo's using a 5i25 and 7i77 board. The mill has been operational for years using linuxcnc but I would like to switch to PP mostly for the excellent conversational. I have not seen any other conversational that even comes close. (I have used nativecam which is the next best thing)

Meanwhile, I have made progress. I got it to launch in sim mode. I also got a INI file that works with the mesa cards I have but crashes due the hal file various pins etc being wrong.
A absolute minimal hal file would help lots, I can add to that once I have PP launching withuot crashes.

I used Gemini 2.5 Pro to create a python script that adds retracts and C axis moves to a program.It could be turned into a filter program. Try it on the attached test file and see if it works properly.

Before:


After:

Just to substantiate that I am a "knowitall" (if that is the right word) NEMA means National Electrics Manufacturers Association. It is not a company, so no product piracy possible there.
The NEMA 23, 32 etc. solely refer to flange dimensions and bolt mounting patterns. Nothing to do with manufacturer, torque, voltage etc.

The only take is that usually bigger flange motors have bigger shafts, draw more current and have more torque.
There are exeptions, so a long high torque NEMA23 motor can have more torque than a "pancake" NEMA32.

As for a DIY router built. More bigger motor more better..... 

p.s. after looking at the pictures. They look like NEMA17. "Possibly pathetic", just get bigger ones that can move some machinery. For routing or milling you will not get anywhere with this. If you want to engrave some nameplates or want to drill some tiny holes, you might get away with it.



pps.: does not matter so much, you can figure out where the coils are at (my guess is 1st pin & 3rd pin for coil 1.... 4th pin & 6th pin for coil 2),
You will find lots of instructions how to figure out the pins on youtube. (tip someone gave me: just connect pin 1 & 3  and try to spin the motor by hand, if that is difficult you  found a coil)
Connect them to your motor driver. Set driver to one amp amp current, if they overheat reduce current. If they are cool enough to touch after running a while you can increase current. If they burn down, buy new ones.
That is what I always did with 2nd hand steppers and so far no fire, no damaged motor.

 

How self-righteous must one be to ask the component creator for help, only to turn around and explain why he's wrong? You’re always quick to highlight your own contributions; maybe it's worth pausing to genuinely consider the input provided.

In addition to what Phill said, when I looked at the config, I noticed that JOINT_3 STEPGEN_MAXVEL is not set to the recommended value of 25% larger than MAX_VELOCITY. Instead it's set only 4.16% greater than MAX_VELOCITY.

For Joints 0, 1, 2, STEPGEN_MAXACCEL is 50% greater than MAX_ACCELERATION (recommended is 25%).
For Joint 3 STEPGEN_MAXACCEL is 87.5% greater than MAX_ACCELERATION (recommended is 25%).

To highlight Joint 3 (the joint associated with the axis you're having trouble with) inconsistencies:

• STEPGEN_MAXVEL is only ~4% greater than MAX_VELOCITY
• STEPGEN_MAXACCEL is 87.5% greater than MAX_ACCELERATION

These inconsistencies stand out, and regardless of how long you’ve been using these settings, the values for both the axes and joints are still not aligned with best practices/recommendations. It's difficult to predict the exact repercussions because, generally speaking, most configurations follow these minimum recommendations. Yours is definitely the outlier.

Hi All. 

I apologise for posting this.  I'm sure the solution is out there somewhere but all the stuff I can find seems not to work.   

I want to run  a pci card as a second parallel port for my DIY cnc.   I want more inputs and some more outputs also. I am using a Generic breakout board that is identical to primary one I am using on Port 0. My board has a Parallel port on 0x378 This Port 0 in linux and works fine if I do in my hal file.

loadrt hal_parport cfg="0 out” or loadrt hal_parport cfg="0x378 out” 

The PCI Card is some sort of generic MOSCHIP based card and it is detected as Port 1 in linux.  It has two addressed associated with it 0xe480 & 0xe800I have tried the following in my Hal file

loadrt hal_parport cfg="0 out 1 out”  (port 0 works port 1 nothing)
loadrt hal_parport cfg="0 1 out”  (cant remember what this one did probably same as the rest)
loadrt hal_parport cfg="0x378 0xe480 0xe800 out “ (port 0 works port 1 nothing)
loadrt hal_parport cfg=”0x378 0xe480 out” (port 0 works port 1 nothing)
loadrt hal_parport cfg=”0x378 0xe800 out” (port 0 works port 1 nothing)
loadrt hal_parport cfg=”0x378 out  0xe480 out  0xe800 out “ (port 0 works port 1 nothing)
loadrt hal_parport cfg=”0x378 out 0xe480 out “ (port 0 works port 1 nothing)
loadrt hal_parport cfg=”0x378 out 0xe800 out “ (port 0 works port 1 nothing)

I have tried every variation I can think of but still get no output. I am testing using a multi-meter on two outputs on the port 1 breakout board had had set them to be amplifier enable and one output is inverted from the other just for the sake of testing.

I have also tried probing the cable it shows no changes when I press enable in linuxcnc. On port0 it works and hitting enable locks up the motors.

Can anyone point me the correct direction!  I feel like I’m going in circles.