some CS3 known issues

BUGS

...............................................................<=1.2.0(4).1.3.0(0).1.3.1(0).1.3.2(1).1.3.3(0).1.3.3(x)
Loco Config stuff
01- mFX ICON updates do not work correctly........................FAIL.....PASS.....<===.....<===.....<===
.....: Dynamic write to decoder corupted - ETA Bug fix Apr2017
02- mLD/3,mSD/3 sequence programming..............................FAIL....TO TEST ....................PASS
.....: does not show Timers distinctly
07- New Loco Address search function clunky.......................FAIL.....<===.....<===.....<===
.....: only possible after a loco has been defined initially (protocol known) and reedited
.....: appears to be WAD (I still consider it fault)
XX- New Loco "Read Address" repeats, even if Edit window closed.......................................FAIL
.....(MM) Loco can be seen to blink its like during normal scan
.....- but when the edit window closes, the loco continues to blink ever minute of so
13- Cancel = meaningless for CV/mFX devices.......................FAIL.....<===.....<===.....<===.....<===
.....: No undo as changes to decoder are applied immediately. Even before OK pressed
33- EditLoco = Read (DCC or MM] = No validation that protocol is enabled............FAIL
.....: SHOULD have an instructional reminder to place loco on the PROGRAMMING track..........PASS
.....: Message (as above) now seen, but the read attempt still continues in the background...FAIL.....FAIL
33- EditLoco => Configuration = AUTORead no real indication ........................FAIL
.....: Should NOT Autoread...................................................................PASS.....<===
.....: Should Provide a Cancel function (as above)



UI/UX Stuff
03- Poor messaging
.....:mFX (CV general) if loco not on track or track power off....FAIL...IMPROVED...<===.....<===.....<===
06- popup fahrpult ... opens but does not close...................FAIL.....<===.....<===.....<===.....<WAD
.....: Annoying - nonergonomic
15- small menu tabs overlap and operationally hide others ........FAIL.....<===.....<===.....<===.....<===
.....: (e.g. Loco list and keyboard opened so they boarder each other)
28- When in Loco Edit mode - BIG Stop causes onscreen "Stops" to show .....FAIL.....<===.....PASS.....<===
.....: i.e. it hides the "OK" button to "Save/Exit" the configuration
31- If "Klein Lok" is in view in either sidebar and a speed is
.....inadvertently selected, the speed bubble persists.....................FAIL.....<===.....<===.....<===
.....: i.e. keeps reappearing over the top of other window panes.
.....: because it is for "klein lok" the processing of the speed should not be done anyway!


16 - Full Turntable support (Jonas post#7)........................FAIL.....<===
.....: Issue relates to poor UX pullstop
.....: Need legacy means of operation
.....: Need intuitive Graphical UX


Help Stuff
19 - Decoder "I" files not scrollable .....................................FAIL.....<===.....<===.....<===
.....: CAN scroll with using VNC and PC MB3

System stuff
05- fahrstrasse.cs2 file faulty...................................FAIL.....<===.....<===.....<===.....<===
.....: fahrstrasse "shell" present but only ever one item (Mag address = 1)
20 - 60657 is reported as a 60653.................................FAIL.... <===.....<===.....<===.....<===
.....:

Wierd Stuff
10- mFX detection, observed loco rescan picked up mfx signal .....FAIL....TO_TEST..NoRepro...<===.....<===
.....: Concise Repro not defined .....
11- Scary display on startup......................................FAIL.....<===.....<===.....<===.....<===
.....: unnecessary blank screens - over unacceptable duration
12- Scary poweroff/on behaviour...................................FAIL.....<===.....<===.....<WAD.....<===
.....: if insufficient delay after powering off and re-powering (wait minutes rather than seconds before powering on again)
.....: suspect a hardware issue - unsolvable by code
18 - Rouge "large" Sensor icon (with "=" above it) ........................FAIL...NoRepro....<===.....<===
.....: first seen after using popup fahrpult

S88 Stuff
08- Extraneous "Bus" field for S88s of CS2,CS3+,etc ..............FAIL.....<===.....<===.....<===.....<===
.....: only needed for L88 - unnecessary, superfluous, confusing - UNTIDY
09- L88 Sensor conflicts..........................................FAIL.....<===.....<===.....<===.....<===
.....: Sensors (2 separate definitions) able to be defined as both Direct and Matrix of the same L88
21 - This Red "warning" line under main menu icons - no sense or meaning...FAIL.....<===.....GONE.....<===
.....: If this is a warning there needs to be some place where the actual issue is stated - in English for the user!
.....: Determined from M-service and observation to relate to S88 (non)presence
25 - This Red "warning" line appears when the (TFP3) S88 bus length = 0....FAIL.....<===.....GONE.....<===
.....: Can temporarily fool the CS3+ but the line reappears on reboot
30 - This Red "warning" line does not "reduce" when offending S88 removed..FAIL.....<===.....GONE.....<===
.....: Line percentage only recalculates on an "Add"
32 - Conditioning for an Event START may only contain ONE condition........FAIL.....<===.....<===.....<===
.....: The workaround of using inline Delay/continue is a poor substitute

Acccessory management / configuration
14- processes have messages that come and go (mfx discovery)......FAIL.....<===.....<===...Improved...<===
.....: - no continuity within the same dialogue. If function fails (for what ever reason) - no message!
22 - mfx accessory registration (m83.2) gets confused about x4 boundaries..FAIL.....<===.....<===.....<===
.....: e.g. if last address in use was 2, then the m83 is given 3 not 5
.....: this is supported by the m8x, and operational as long as you do not edit the item in which case the address (3) fails validation
26 - mfx accessory registration wizard fails to detect existing config.....FAIL.....<===.....<===.....<===
.....: Specifically 4 or 8 mode addressing
.....: Needs validation
27 - mfx accessory registration wizard has no port setup assistance........FAIL.....<===.....<===.....<===
.....: Icons are "assumed" to be "left turnout" (m83)
.....: Port mode assumed without validation
29 - Editing a multidecoder accessory - "switching time" is not discrete...FAIL.....<===.....<===.....<===
.....: Cannot work through other ports of decoder to set time without saving and specifically selecting the item(port)


mFX+ World of Operation
04- mFX+ images stretched (120%) rather than new images - rough edges, font size incorrect
...........................................................................<===.....<===.....<===.....<===
17 - mFX WOO enablement (With Consumption) causes blank screens............FAIL.....<===.....<===...NoRepro
.....: Screen goes blank about 5 times, for about 5 secs and then stays blank
23 - Electric speed able to show "-1"......................................FAIL.....<===.....<===.....<===
24 - Diesel "fahrpultdiesel.cs2" has duplicate buttons for icon id=53......FAIL.....<===.....<===.....<===
.....: Actually legacy from the very first implementation of mFX+ on the CS2


0- UI poor rendering (English?) overlap of text and images or truncation
...: System->CS3_>Track ... Start Locomotives Automatically.......FAIL.....PASS.....<===.....<===.....<===
...: System->CS3->Track ... WoO Rotary Knob assignment............FAIL.....PASS.....<===.....<===.....<===
...: Loco - Edit->NewLocomotive->Info ... "Locomotive Database"............FAIL.....<===.....<===.....<===
...: Loco - Edit->Setup .. ".... Function .... 'Run Time Functi?"..........FAIL.....<===.....<===.....<===
.....:: OK with NO Function selected - appears as soon as you select a Function
...: System-> (various) "Itemnumber" should read "Item Number"....FAIL.....<===.....<===.....<===.....<===
.....:: better still "Product Number"
...: Web Interface - Should inherit Language setting of unit .....FAIL.... <===.....<===.....<===.....<===



0- UI lack of English Translation (poor spelling)
...: Loco Sidebars, when no Loco selected =...................."Keine Lok" FAIL.....<===.....<===.....<===
...: Loco - Edit->Setup "Loc Symbol" dialogue field........................FAIL.....<===.....PASS.....<===
...: System->TFP3->Setting S88 config fields "Lange","Zykluszeit","Bitzeit"FAIL.....PASS.....<===.....<===
...: System->LS88->Settings->Setting1-16 combo "Einzeln","Tastaturematrix".FAIL.....<===.....<===.....PASS
...: Acc. config. item has not been assigned to a Trackboard = "Gleisbild".FAIL.....<===
...: Accessory->Edit->Discover mfx itmes = "Discover mfx itmes"............FAIL.....<===.....<===.....PASS
.....................: (as above) = "Ein neuer mfx-Decoder wurde entdeckt" FAIL.....PASS.....<===.....<===
...: Edit Track Diagram = "Selction" (missing "e").........................FAIL.....<===.....<===.....PASS
...: Settings for Area = "Name Areal"/"Farbe Areal":All colours in German..FAIL.....<===.....PASS.....<===
...: mfx article discovery.........."Ein neuer mfx-Decoder wurde entdeckt" FAIL.....PASS.....<===.....<===
...: mfx article discovery.."Das Mapping des Artikels m83 wird ausgelesen" FAIL.....PASS.....<===
...: Article configuration tab (mfx device)................................FAIL.....PASS.....<===.....<===
.............................."Der Artikel hat (noch) nichet geantwortet."
.................................."Eine CV-Bearbeitung ist nicht moeglich"
...: "Direct adress drive" title <-- note spelling of "Address"............FAIL.....<===.....<===.....<===
...: Writing to a Loco-Card message contains the phrase "rightg slot"......FAIL.....<===.....<===.....<===
...: "Die Mfx-Suche ist im Stop-Zurstand nicht moglich""...................FAIL.....PASS.....<===.....PASS
.......when attempting automatic mfx accessory scan with CS3 in STOP


DOCUMENT ERRATA
Page 7 of CS3 manual states "... Link S88(60833)..." should read "... Link S88(60883)..."
.....: Noted that with the 4 language version (available Apr 2017) that the nederlands ONLY section was corrected

EXPECTED FUNCTIONS - ABSENT
00- InApp help files missing......................................MISS.....PRO.....<===
.....: anticipate slow progress
01- English translation...........................................MISS.....<===....PROG
.....: anticipate slow progress
02- Decoder firmware updates......................................MISS.....<===
.....: CS3 v1.3.0(0) = up to Mar2017 = 3.0.1.6 (not 3.0.1.8 already available)
03- Decoder Soundfile upload

04- Output programming of mFX signals not operational

05- no mFX+ sensor definition for 'advanced' operation (as per CS2) or equivalent

06- System Bus devices not succinctly identifiable ...............MISS.....PASS
.....: i.e. ... by serial# (or Name = MS1)
.....: CS3 v1.3.0(0) = actual Serial# field (for connected device) now present
07- mFX function option of a timer missing (refer CS2)............MISS.....PASS

08- Shuttle function (Workaround with Event possible )............MISS.....<===
.....: = labourious - preference for the CS1 way of implementation)
09- USB devices not identified by physical position
.....:
10- Layout lock (from inadvertent panning)........................MISS.....<===
.....:
11- Single step edit (Spanner/Wrench) - instead of multiple steps with all levels lumped together for more clicks to wear out the finger
[color=white].....:

Subjective loss compared to the CS2
A- Loss of adjustable feet
B- Loss of tactile "click/detent" for lokocard (camofaged because of tighter (rough) tolerance on initial entry)
C- Loss of Tactile Function buttons
D- Loss of Tactile Menu button
E- Loss of circular speed display
F- Loss of singular/specific 'known' access method for layout, keyboard, memory(event), Control - multiple UI actions needed
G- Loss of supplied stylus and holder

INTEROPERABILITY CLARIFICATION

1- CS1 (60212) may only be used through a 60213/4/5 CS2 that is slaved to the CS3

And now this functionality has been removed in the latest CS2 update

NEW FUNCTIONALITY - COULD HAVE BETTER USEBILITY

1- Popup FahrPult
- - YET "another click" to get the functions up ?
- - YET another click on the minuscule tab selection to get F17+
- - NEED toggle click to close
2- Any twisty (Chevron rotate) to expand a sub menu ( e.g. System Info ) should NOT require yet another scroll function to see the content
3- Events have a countdown of steps BUT also need a countdown of time!
4- Get rid of the penguin
5- Virtual Sensors ("Kontrolling Contacts") could be multistate - non binary
6- Offer (pan) lock function for layout
7- better 'finger friendly" distinction for the small menu items

FOR THE NEXT CONTROL UNIT PLEASE CONSIDER ...

1- Return all the lost hardware functions (see above)
2- Bigger screen - does not have to be touchscreen - or an external display port for users to choose their own screen
3- Edge buttons
- - (top brow maybe) for Brightness (Up/Down) - Sound (Up/Down/Mute) - with onscreen notification when changed
- - Scroll Wheel
4- Separate GO button

PLEASE TRY HARDER

1 - the mFX loco "Quick ID" should be configurable by the user
2 - Visual realestate hides lists - any drop-down or horizontal list should not hide items because the window is too small. If there is room (on the screen) then the list window should be able to show all items without having to scroll (again) for one or two items
3 - Consistency in UI for scrolling - some "windows" move up (data UP) when scrolling [b]UP while other "pages" move up (data [b]DOWN) for the same action.
4 - Need "confirmation" (feel good) indication - even if action successful
5 - DO NOT present a "Cancel" option when the data has already been sent/programmed (e.g. mfx function icon changes)
6 - Throttle granularity (40 detent steps) means that first detent equates to about speed 3 on a 127 step decoder
7 - DCC programming (possibly more of a "poor messaging" issue)

Edited by moderator 23 September 2021 00:10:25(UTC)  | Reason: Not specified

The more and more I read about the CS3 and what others are telling me just reiterates that my not rushing to replace my CS2 is the best thing to do. I'd go so far as to recommend to someone looking to upgrade an old CS1 or from a mobile station to in the short term find a deal on a CS2 and hold off a while on the CS3 IF they can't wait a little longer with their existing equipment. Just my humble opinion. Personally, I find te CS2 to fit my needs quite admirably for right now.

The hardware technology is not what I would call bleeding edge but it is sufficient for the intended use and way, way better than what is in a CS2. The software unfortunately is not mature enough or competent enough for me at this point in time. Hopefully Märklin will get the software sorted out sooner rather than later. If they don't then we will still have the option of getting an ECoS if we do not purchase now.

I also find the CS2 quite adequate for my current needs and I don't plan to upgrade anytime soon.

Thanks Peter,

I think you can add turntable when support said that it doesn't work yet....

The update release seems very slow as always with Märklin

CS3 help please.

New to CS3. I have Lok 39500 E50 with previous address 50 showing up when I press Lok Edit as address #41. I want to change it back to 50

but the area on Lok edit window is not highlighted so changes can't be made. What am I doing wrong here? Thanks, Peter

It is a replaced decoder. I believe it's an ESU one. No address area on any Lok is highlighted for changes? I just checked other MFX Loks too.Peter.

Thanks for the info, Tom. No, it's not a big deal. Cheers, Peter.

Thanks a lot, Peter. It's good to know there's a cure! I will leave it as is for now. Cheers, Peter.

Hi all,

I regular switch between my old CS2 and the new CS3 as my daughter prefers the old speed dial where you can understand the speed settings.

I only have one central station connected to layout at any given time powered by 60061 60w switched mode power pack.

I have noticed that when my new CS3 is connected there is a significant higher humming noise from some decoders. One of my inbuilt turnout track decoder is so loud I can hear it several meters away. I believe the humming noise is associated with the frequency used for the PWM between decoder and motor, but I was wondering if any other users have noticed similiar with the new CS3 (maybe it uses another frequency)? Or maybe it's just a conincidence, but why would many of my locomotive decoder also be humming more now?

My track voltage as per CS is identical between CS2 and CS3 around 19 volts.

Brgds Lasse

If it is really a hum rather than a buzz, I would want to know details on your accessory (Yellow 16VAC) schematics and whether you have a common ground.

You do need to record which hardware version of CS2 you have and if you really mean a CS3 (60226) or CS3+ (60216)

That's about loco decoders.
A turnout decoder that is just listening (waiting for commands) does not generate any PWM - at least it should not.
Sometimes you hear components on the decoder humming with the track voltage.
The CS3 has a new TFP (track format processor, aka GFP or Gleisformatprozessor) and that may have a different pattern of track voltage.
Identical boosters can generator different "tunes" with different software (called TFP in this case).

The humming varies depending on which locos you have previously addressed.

I wonder if it's only my CS3 controller but my throttle can only do 3 speed steps at a time (from 0 to 3 and onwards)?

With each click your rotate you will gain 3 speed steps. There is a work around. If you rotate the knob halfway between clicks you can get to speed step 1 and 2.

Alternatively you have to use the mouse and the speed graph.

Some may not for it any issues but when you are programing CV and optimizing the decoder is quite handy to test the loco at the lowest speed level 1.

Interested if anyone has observed the same?

And how is on mobile stations etc.? And what about a ESU controller?

Thanks!

Brgds Lasse

The CS3 throttle appears to have been designed/calibrated such that one detent is equal to one bar of the 40 bar speed gauge.
Unlike the other throttles there is no variation if you turn the dial slow or fast

If you happen to have set your locos max speed to 120 then it will display in 3 km/h increments
If your decoder is defined as one with 128 steps the each bar would equate to approximately 3 steps.

Watching the gauge while driving with a CS2/MS2 attached you do seem to have 1 step increments (when turning slowly)
(by watching I mean noting the number displayed nudge up even if the next bar does not highlight until 3 nudges. If you set the loc max speed to 512(CS2)/160(MS2) then the number will pretty much match 1 for each nudge)

So. The issue is noted that the CS3 detenting throttle has a fixed course granularity (1/40th of range), that is an apparent backward step compared to the MS2/CS2.

With all that in mind the various emmitters (throttle) are an analogue device and the value is a fuzzy interpretation. If you want to be precise you would use a PC, keeping in mind the the CS2/CS3 API uses a speed range of 0-1000.

Tracing the system bus we see APPROXIMATELY.
CS3 = steps of 24/1000
MS2 = steps of 6/1000
CS2 = steps of 2/1000

Marklin document the formula for translation to a decoder value as ..
Systemfahrstufe = 1 + (Gleisfahrstufe - 1) * Schrittweite

with statement
The decoder driving stage 1 is always also system driving stage 1.

and examples
Decoder driving stage 11 is thus at:

for 14 Driving steps ==> System driving stage 771
for 27 Driving stages ==> System driving stage 381
for 28 Driving stages ==> System driving stage 371
for 31 Driving stages ==> System driving stage 331
for 126 Driving stages ==> System driving stage 81

I think this statement needs to be teased out a bit. Are we talking about the decoder or the motor.

Yes, the change in controller granularity compared to other controllers is an issue worth noting

For one the motor, the gearbox and transmission design, and the other mechanical aspects cannot be overlooked. When using a common motor its starting voltage ( the effective PWM voltage at which of the locomotive motion is detected after overcoming all resistance) is probably different for a plastic BR89 v a metal BR218

A motors characteristics are inherently analogue. Any imposition of "steps" is nothing more than a convenience of design and operation (e.g. how many relays and associated wiring can fit under the hood of in a E44). And the association of "a step" to a particular speed is performed by where the tap off the transformer is done or, in the case of a mfx decoder the value configured into the speed graph.

Unless you really need the granularity of 127 different speeds in normal operation, I would suggest treating the speed table of the decoder such that command steps 1,2 & 3 are set to whatever value makes the particular locomotive crawls reliably.
Note: also taking into account the starting voltage setting.

How many of Marklins locomotives actually leave the factory calibrated such that speed command 1/127 makes it move?