ML Riechers Systems Engineering


[Photo of the Front Side of our rse-2c micro-controller card]

Taking the Time to do the Firmware Right.


Welcome to our pages on the changes to the Registration Controller Program. Here you will be able to check out the latest Versions of the Program, compare them to yours, and maybe decide to Upgrade your controller.

The Registration Controller Program has a name; it is called nc_regis.

Unless we note otherwise, the later Program Versions retain all of the changes we made to the earlier versions.

  • Version V1.9f, dated 30 December 2009: Added a modification to the SPEEDO to time-out after 30 seconds of Operator Wheel non-activity, and give a visual indication that SPEEDO is running. Ensure that all Operator Keyboard Key Presses end up at a defined state, namely selecting the Linear Register. Fixed a SPEEDO bug, where SPEEDO could start from inappropriate Key Presses. Also, implemented Direct Page Addressing for most Static Program Variables.

  • Version V1.9e, dated 9 November 2009: Added an Operator settable Error Zone Minimum Speed Value, to not ever sound the gong (Error Alarms) if the Press Speed (in Feet per Minute) is less than this value.

  • Version V1.9d , dated 27 July 2009: Added STRESS50, a factory maintenance test routine, to exercise and test the RS-485 communication drivers.

  • Version V1.9c , dated somewhere around April 2009, and probably incorrectly labeled V1.9b (but probably only one or two demos escaped): Modified the program to clear the Inspection Zone when changing Press Model, setting the number of teeth on the Impression Cylinder, or Switching Print Roll Drive Side.

  • Version V1.9b , dated 8 April 2009: Added the Force Jump (SPEEDO Overdrive) feature, changed the Error Output Relay Signal, as an Operator Option, to Gong using an inverted signal, added a Press Model for Die Station Auto-Register run off of an encoder on the Lower Die Shaft, and removed last vestiges of ARC 1000 Auto-sensing.

  • Version V1.9a , circa Feb 2009: Major release going forward. Uncirculated.

  • Version V1.8e , dated 10 February 2009: Took out the legacy -47 Offset for Axial (side/side) registration. This is the current MarkAndy/COMCO release, and the End of Life of the nc_regis Version 1.8 series.

  • Version V1.8d , dated 2 September 2008: Corrected the Flash memory updating for rse-2d's at power down, so the Inspection Zone Windowing value is actually saved.

  • Version V1.8c , dated 11 June 2008: Simplified the motor startup method for rse-2d's, possibly correcting disturbed transition timing.

  • Version V1.8b again , dated 4 June 2008: Corrected the uPfd100 processor timing value. The wrong value messes up Constant Error Correction and VariPrint.

  • Version V1.8b , dated 4 April 2008: Major, major upgrade to port nc_regis from MLRSE controller Version rse-2c to MLRSE controller Version rse-2d. Also, corrected the Gong routine to not ring the Gong in Auto, corrected the sense of the 5 fpm Operator Option, changed the pulse output in VariPrint slightly, changed the way variables are saved on power-down, and corrected a corner case in Receive from Host.

  • Version V1.7j , dated 5 August 2002: Added an Operator Option to use 5 fpm instead of 25 fpm as the minimum press speed, and corrected a corner case of incorrectly computing the AXIAL motor movement. This release marks the End of Life for both the MLRSE Version 2c controller (as of 30 April 2008), and the nc_regis Version 1.7 series.


    Here are the expanded versions:

  • Version V1.9f , dated 30 December 2009:

    1. Do a 30 second time-out on the SPEEDO, to work as follows: Give the Operator a 30 second time-out when he starts the SPEEDO. If he moves the wheel (adjusting the correction speed), then give him another 30 seconds before timing out. That way he gets plenty of time for actual adjustment, time for a half revolution of a 24 inch Repeat Roll before needing to move the wheel to get another 30 seconds.

    Why:

    Fear on our part that the Operator might leave the SPEEDO enabled during the actual Production Print Run. If that were to happen, that Station's Registration would be at best turned off, and at worst crept through the Print Image, ruining your whole day.

    Explanation:

    30 seconds after the Operator forgets about having the SPEEDO set up, it's canceled. That should be a short enough time so that the SPEEDO doesn't interfere with Print Production.

    How to use it:

    To keep the SPEEDO alive, be sure to adjust the Operator Wheel at least once every 30 seconds.

    2. Give a visual indication that SPEEDO is running: Make a flashing dash on the Operator Display in the place where the "Register in Auto" symbol appears, regardless of whether the Registration System's in AUTO or MANUAL.

    Why:

    1. Fear on our part that the Operator might leave the SPEEDO enabled during the actual Production Print Run. If that were to happen, that Station's Registration would be at best turned off, and at worst crept through the Print Image, ruining your whole day.

    2. It's the right thing to do. Every Operating Mode ought to have a visual that it is active.

    Explanation:

    This feature gives the Operator a Positive Indication that SPEEDO is active. If the Top Bar

    How to use it:

    Use Mark I Human Eyeball on Operator Display. If the Top Bar where the "Auto/Manual" symbol should be is flashing, then SPEEDO is active.

    3. Ensure that all Operator Keyboard Key Presses end up at a defined state, namely selecting the Linear Register. Fix false alternate key sequences that might let SPEEDO start.

    Why:

    1. For Regularizing the Button Sequences, Customer Request.

    2. Our bug find, and our bug fix for the false SPEEDO start.

    3. It's the right thing to do.

    Explanation:

    Prior to this release, some Operator Keyboard Key Press sequences would end up in an indeterminate state. Setting the Dead Zone, Error Zone, linear/Axial/CEC averaging didn't return to "Select Linear Adjust." Also, selecting CEC for display select ought to "Select Linear Adjust", instead of nothing at all. Furthermore, we found that SPEEDO could be turned on after pressing the Pre-Reg key, and also during the "NEXT MARK" sequence &mdash very bad, and embarrassing for us.

    How to use it:

    Correctly, I hope.

    4. Implemented Direct Page Addressing for most Static Program Variables.

    Why:

    Our internal programming needs.

    Explanation:

    No customer action needed, except to note that this was a fairly big programming change. If you see any program wierdness, let us know.

    How to use it:

    n/a

  • Version V1.9e , dated 9 November 2009:

    1. Added an Operator settable Error Zone Minimum Speed Value, to not ever sound the gong (Error Alarms) if the Press Speed (in Feet per Minute) is less than this value.

    Why:

    Customer Request.

    Explanation:

    Using this feature the Operator can set the Error Zone Minimum Speed Value to say, 100fpm, and work in peace while he's setting the Press up at 50fmp, regardless of whether stations are in Auto or Manual. Only when the Press Speed is at or exceeds 100fpm, could the Alarms gong.

    The Error Zone Minimum Speed Value is set to zero out of the Factory, so, unless the Operator changed the value, the controllers with this enhancement would act just like previous versions, and hence this is not a special, and no need for a special part number. Just a version enhancement.

    How to use it:

    The Operator sets the Minimum Speed Value by:

    1. Pressing the Station Keyboard "A" key and "Dead Zone" keys together, then

    2. Adjusting the Minimum Speed Value from 0 to 2550 fpm in 10 fmp increments, and then

    3. Pressing the Station Keyboard "A" key and "Dead Zone" keys together once more, "setting" the value into the controller's memory.

  • Version V1.9d , dated 27 July 2009:

    Detail to be expanded, time willing.

  • Version V1.9c , dated somewhere around April 2009, and probably incorrectly labeled V1.9b (but probably only one or two demos escaped):

    Detail to be expanded, time willing.

  • Version V1.9b , dated 8 April 2009:

    Detail to be expanded, time willing.

  • Version V1.9a , circa Feb 2009:

    Detail to be expanded, time willing.

  • Version V1.8e , dated 10 February 2009:

    Detail to be expanded, time willing.

  • Version V1.8d , dated 2 September 2008:

    Detail to be expanded, time willing.

  • Version V1.8c , dated 11 June 2008:

    Detail to be expanded, time willing.

  • Version V1.8b again , dated 4 June 2008:

    Detail to be expanded, time willing.

  • Version V1.8b , dated 4 April 2008:

    Detail to be expanded, time willing.

  • Version V1.7j , dated 5 August 2002:

    1. Added an Operator Option to use 5 fpm instead of 25 fpm as the minimum press speed:

    Why:

    Customer Request.

    Explanation:

    nc_regis uses a Minimum Press Speed for many things. For instance, it will not ring the Error Gong, or calculate a Registration Error when it thinks the Press is not moving, and it will do a Pre-Registration Roll move only when it thinks the web is still. nc_regis believes that the web is still at 25 fpm and slower.

    However, some COMCO customers have purchased presses with idle speeds well under 25 fpm. This option allows the Auto Register to be active at these low idle speeds.

    How to use it:

    Please see "How to set up the `Set Operating Parameters' feature", "use 5 fpm min speed, not 25 fpm, for WebStill" on line 6 in Operational Modes Preferences.

    2. Corrected a corner case of incorrectly computing the AXIAL motor movement:

    Why:

    Programming Error.

    Explanation:

    If the GAIN times a negative axial error equals zero, then we were setting a zero motor movement, but _also_ setting the negative bit. This commanded the negative max motor movement, thus throwing the axial motor erroneously and wildly to the negative, until caught on the next mark.

    Corrective Action:

    Set the negative bit to the result of the final two's complement operation, rather than assuming the original negative sign.

    3. This release marks the End of Life for both the MLRSE Version 2c controller (as of 30 April 2008), and the nc_regis Version 1.7 series.

    Why:

    Integrated Circuit obsolescence.

    Explanation:

    In 2006, we introduced the rse-2d controller to replace the rse-2c, upgrading it in 2007. The rse-2d controller we use for nc_regis uses a modern Freescale microprocessor, MC9S12XD64 by name, which has vastly more memory and speed than the MC68HC11 processor used on the -2c. Additionally, it has a RISC co-processor, is easier to program, and is less expensive.

    'Nuff said?

    Benefits to you:

    The additional capabilities of the rse-2c will ensure continued enhancements to the the nc_regis program in Versions 1.8 and beyond. Additionally, we can re-program your controllers to upgrade them much faster and more easily.

    Contact Information

    Please send inquiries about MLRSE to
    M. L. Riechers Systems Engineering     Voice:  +1-513-844-2220
    530 Main Street                        Fax:    +1-513-889-0488
    Hamilton, Ohio 45013-3222, USA         mlr(at)rse(dot)com
    


    This page was created by Mike Riechers, e-mail at: mlr(at)rse(dot)com

    The Webmaster of this site is free to use the images below on an Apache/NetBSD-powered Web server. Thanks for using Apache on NetBSD!

    Site driven by NetBSD - NetBSD rocks!