Mesa daughter cards: Parallel vs Serial

Hello Marc,

How far is your Mikron WF41C retrofit project?

I am planning to to similar job with Mikron WF32C. Did you use original Bosch drives and encoders? Where I can expect most problems?

Regards
Bojan

Hello Bojan,

we (me and two friends) are slowly but steadily moving along. We are planning to use the original Bosch Drives with their encoders. We developed a little board for the Heidenhain glas-scales (basically doing the same thing as an original Heidenhain EXE). I'm currently also working on trying to find a solution for the homing (not decided if I try to add a new homing method to linuxcnc, which looks pretty complicated, or create a hal-component that only creates an artificial limit-switch signal once two signals for the index track were received and the absolute position-offset has been calculated).
Further we developed a box for the control containing most the mesa cards with similar dimensions AS the original Heidenhain 355 control and basically has the same connectors as the Heidenhain. Only the 7i73 is in a second HMI interface box containing the computer, a touch-screen and some buttons, potentiometers, a jog-wheel...). The idea is that we can still switch between the Heidenhain and the LinuxCNC control. As we have more then enough inputs with our MesaCards we built some d-sub37 T-connecters, that allows us to tap into the input and output signals of the Heidenhain with linuxcnc to basically run the machine with the Heidenhain and record input and output signal sequences for certain operations like gear-changes etc.
So currently we just have completed the LinuxCNC control box. Last step was to mounted the exe replacement boards we had to wait for. Within the next few days will try to get them running as well as do some systematic recording of the Heidenhain signal sequences.
Once this is refined and done we will probably start try to use the 7i77 to control the Bosch servos which for me is the scariest step (relevant electric power involved and risk of some serious damage .

Best regards,

Marc

Marc
Would you be willing to share more information about your - " original Heidenhain EXE). " ? As in much - much more ----
Many others have run into the Heidenhain scale - ( micro-amp to TTL ) - problem , would be very helpful in not always having a new wheel invented .
Do you have a build thread for your Mikron WF41C retrofit project?
Thanks
Bob

Hello Bob,

the original Heidenhain glass scales of the series LS403 and similar produce two harmonically oscillating analog signals A and B when moved with constant speed. Depending on the direction of the movement B is trailing A or the opposite way around by 90 degrees allowing to get the direction of the movement. With the scales mentioned the period for A and B corresponds to 0.02mm. However the typically resolution the Heidenhain control gets out of these scales is 0.001mm and for consumption by nc controllers like Siemens, Fanuc, LinuxCNC we need a digital signal. To achieve this Heidenhain provided little boxes called EXEs. They basically have three tricks up their sleeve to achieve the increased resolution:
1. They do signal quadrature of A and B which increases the resolution by a factor 4.
2. They evaluate the analog signal not only once per period but several times. In our case specifically 5 times per period (this leading to an accuracy increase of 4 times 5 i.e. 0.001mm)
3. They output a 5v digital square-wave signal again of A and B but with the increased resolution (versus the original analog and also quite weak 1.3v signal)

If you are retrofitting a machine with Heidenhain scales but no Heidenhain control chances are that you find these little EXE boxes somewhere in your control cabinet. Heidenhain controllers 'unfortunately' embed this circuits into their control unit so they typically can not be re-used for a retrofit.

Further there is also an index track which produces an in my case an index signal approximately every 10mm along the scales. These index-signals are not equally spaced. Therefore for homing you just need to run a max of 2cm (even direction is irrelevant) and measure from the A and B signal the distance between two index-signals. Based on this information you can calculate the absolute position on the scale i.e. you can set your offset of the scale and are homed.

We now developed a little PCB to do exactly what these EXEs do. Last week we connected them to the machine for the first time by taping them into the signal line between the scale and the Heidenhain controller. When moving the axis with constant speed through the Heidenhain controller we could measure on an oscilloscope something that looked sometimes similar to the desired phase-shifted A and B TTL signals. However it was hard to get a stable image as the movement is more jerky on the micrometer scale than expected. The controller is constantly adjusting which gives the oscilloscope a hard time establishing a stable trigger. Further of taping into the signal line caused a signal error of the glass scale every now and then requiring a restart of the machine every time. Therefore the next step will be to temporarily remove the y-scale from the machine to allow for better testing.
If we get these PCBs to work we are happy to share more info about them. But please bear with us as this is not our full-time job and we can only work on it as times allows.
if we have some more stuff working we might also consider creating a build thread for our Mikron project.

Best regards,

Marc

Hi Mark
Thank you for up dating what you are finding out about the Ls 403 scales . Hope to read more --- yes I know how LIFE can get in the way of progress
Mikron project

---

Guess you have been holding out on us. If you get a chance show some photos . Think Micron is from a Swiss manufacture vers the Deckels an Mahos being of German manufacture . All I have read is that Micron is considered to be a very good machine . Wondering what your Micron has for a spindle drive an spindle transmission ?
Bob

Hello Bob,

if you are still interested, I started a thread regarding the retrofit of the Mikron WF41C here: forum.linuxcnc.org/12-milling/42195-retr...-mikron-wf41c#205460

Best regards,

Marc