The Multi-protocol Digital system
for Motorola, Selectrix and DCC
about the "IB"
As with many new electronic devices especially those that do so many different things such as the Intellibox there are some situations that may require that you change some of the factory default settings in order to get optimum performance. Please read the following sections carefully and, if you decide to make changes to the settings on your Intellibox, be sure to read the instruction manual on how to implement these changes.
Differences in default settings for Marklin, DCC, and mixed mode operation
IB's from Modeltreno (Italy) have a slightly different configuration than IB's coming from Uhlenbrock (Germany). In particular, these two special options (SO) have different values:
SO #25 is 1 (instead of 0 for Germany). This forces the IB to always send a DCC signal, no matter if the user did not recall any DCC Lok. (This is needed for compatibility with older DCC decoders and/or for DCC decoders which haven't been programmed so as to turn "automatic analog mode recognition" off).
SO #907 = 5 (instead of 1 for Germany). This tells the IB to send - at startup - both the Marklin and the DCC idle signals to the tracks. That is: this(these) signal(s) are sent until the user takes control of the first Lok. After this, the idle signal is in theory - no more needed, therefore, it is not sent.
Furthermore, also this SO should be considered:
SO #22 = 7 (same value as in Germany). This is the amount of Marklin and DCC idle packets that the IB sends at each Power On. (More precisely: if the value is odd, then only that many Motorola and DCC packets are sent. If the value is even, then only that many Motorola packets are sent).
In a word: IB's from Modeltreno are configured so as to be "ready" for a mixed Motorola + DCC environment, while IB's from Uhlenbrock are configured for (almost: check SO #22) "pure Motorola".
Therefore, if you only change SO #907 (to the value 1), you have removed the DCC idle signal, but, since SO #25 "takes precedence", a DCC signal is still there. Therefore, in order to guarantee that no DCC signal is sent, also SO #25 must be changed (to 0). Furthermore, I'd recommend also changing SO #22 to the value 0. If you do this, then the IB shall not send any DCC packets anymore - provided, of course, that you do not explicitly configure a Lok or a Turnout for DCC!
Marklin Boosters (sometimes) need SO #901 to be changed from its default value (1) to a slightly higher value (e.g. 3). This forces the IB to stretch the negative part of the DCC signal, so as to make it "look like" (in a sense) the Marklin signal - thus making the Marklin Booster "happy". Therefore:
|a)||If you only use the Marklin-Motorola format:
SO #25 = 0
SO #907 = 1
SO #22 = 0 (or a low even value, e.g. 6)
(No need to change the default value (1) of SO #901)
|b)||If you only use the DCC format:
SO #25 = 1
SO #907 = 4
SO #22 = 0
(If using Marklin Boosters: SO #901 = 3)
|c)||If you use a mix of Motorola and DCC:
SO #25 = 1
SO #907 = 5
SO #22 = 7
(If using Marklin Boosters: SO #901 = 3)
It has been also noticed that 6017 Boosters are more critical than 6015 Boosters. In the couple of cases that we have had, the problem has been solved by carefully controlling the cabling which was used in order to supply these Boosters. For example, it can be important that a good common ground connection exists btw all Boosters: that is, that all brown wires of all Boosters be connected together - and to the IB too. Another issue is the cable connecting the Boosters to the IB (and the Boosters themselves). This should not be too long! Finally, we also had a case where the placement of the 6017 (at the beginning or end of the Booster cable) did make a difference!
The reason for experiencing "strange" behavior of the IB with a 6015 or 6017 and the recommendation that SO #901 = 3 is that the Marklin Motorola voltage has a negative average value, while the NMRA features an average voltage nominally =0. The 6015 and 6017 calculate "analogically" this average voltage. If it is <0 it's OK, if it is >0 or =0 they issue a "short" signal. By changing the SO#901 we oblige the IB to "stretch" the NMRA signal (but still remaining "NMRA compliant") in order to impose a negative average value also to the NMRA signal.
Problems with sending bursts of commands thru the IB interface via software (e.g. TPL)
If your IB has software version 1.000, we know about this problem. If more than approximately 10 turnout commands are sent to the IB one right after the other, the IB software goes in a "deadlock" condition (caused by the slow I2C bus); this, in turn, causes a "watchdog" reset shortly afterwards. This bug (along with a few minor others) has already been fixed. We are working on the update program which shall be downloadable via Internet.
For the moment being, you can use this work-around: please rise the minimum turnout "on" time from 100 ms to 300 ms. This is done from the main configuration menu of the IB (press [Menu] then [Mode]).Re-calling a locomotive on the IB after another controller address it causes it to stop
Note: This appears to only happen when another device
(e.g., software, infrared controller, etc.) takes control of an engine
previously controlled and released (another engine called) by the Intellibox.
Try changing LSO (Lok Special Option) #4 of any "Motorola new" loco (equipped with a Maerklin decoder) which shows the "wrong direction at power up" bug (only happens if PC operated) to the value 0 (the default for that protocol being 1). This should cure the problem.
The same LSO change may also cure another direction change peculiarity affecting some Maerklin ICE 3 trains, as well as the old TEE 3471. In the latter two cases what happens is: upon a direction change, the internal (to the model) relay which changes the pick-up shoe does a quick second "click": the decoder changes direction twice.
Each loco has a set of 10 LSO's. They are protocol dependent. These LSO's are edited from the loco menu: press the [lok#] key corresponding to the IB side through which you already have control on that loco. Now, while the cursor is blinking on the loco address, press the [menu] key and scroll to the "loco special settings" (or similar text). Choose the LSO you whish to review or edit on the left side. Jump to the right side (LSO value) with the usual [right arrow] key, press the key corresponding to the desired value (either  or ) end finally press [Enter].
Resetting the Intellibox to its factory default settings
To perform a configuration reset of the Intellibox:
- turn the Intellibox on while keeping the [Enter] key pressed;
- keep it pressed until the message:
- Release the [Enter] key.
The software version and Intellibox serial number will appear. Please write this down for future reference
- Now press (and release) the [Enter] key again. The
Intellibox shall ask whether you wish to perform a configuration reset. The default is "no".
|| Return to the Intellibox Page | Return to Intellibox FAQ's Page | Return to the Home Page ||