New Marklin Decoder (LGB 24741) and Digitrax Super Chief (EU Wireless)

E

ernie

Registered
Hi All

It seems I may have good news for some?

I waited ages to see what Markling would produce to replace their LGB non DCC compliant system. But, having now seen specs, etc, I just don't like it for so many reasons. I believe that the Massoth DCC system would be technically fine but their system cost and upgrades costs are really high.

So, like I have done for my OO-gauge DCC system, I decided to go US. As wireless for outdoor G Scale is a must for me, I went for Digitrax Super Chief 8 (~120VA) DCC System and their PSU (~270VA) that can supply a controller and a booster. North Coast Engineering (NCE) do not produce EU wireless and Gaugemaster (MRC) does not go to G-Scale volts.

But, how well would the new Marklin/LGB decoders respond to fully NMRA compliant DCC systems?

So, some facts. I have have just purchased an LGB 24741 era III steam loco that comes factory fitted with the new Marklin decoder with sound and steam. And at a very good price point of ~£450. And it seems fine with my Digitrax DCC system. All 16 functions, most being sound, work as expected and so does the general sound for a 99 series steam loco.

I have, so far, programmed changes to a few CV's and they work as expected. I just couldn't read back CV's on the main track, could be that my Digitrax skills are at fault here? I will be setting up a program track over next few days and do some further testing.

Bye to all. One happy G-Scaler.
 
It is good to hear you are happy with your DCC arrangements.

ernie said:
I just couldn't read back CV's on the main track, ....
Click to expand...

It is not possible to read back CVs while programming on the main. (Well not without special arrangement such as Rail Com).

As you say you should be able to read back CVs on the programming track.

I believe there was a thread on here from Nick wanting to know the manufacturer Id in these new Marklin mfx decoders, perhaps you could help there.
 
ernie said:
... North Coast Engineering (NCE) do not produce EU wireless
Click to expand...
Actually it seems the fairly recent rule changes to frequencies has made the NCE radio gear legal here in the UK at last. I've noticed UK sellers now offering the NCE radio gear and stating it's legal. I must admit I'm really tempted to get a radio base station and wireless cab for my NCE system, I just need to justify it given that my line's not seen a lot of use over the last couple of years.

Cliff George said:
I believe there was a thread on here from Nick wanting to know the manufacturer Id in these new Marklin mfx decoders, perhaps you could help there.
Click to expand...
Yes would be really interesting to know - from a purely technical standpoint at the moment as I don't foresee any LGB purchases but I am a JMRI user and I've been following the threads with peoples' tales of woe about programming these Maerklin decoders.
 
Well, I tried to set a 4 digit address (the 4 digit running number) and it read back ok but it wouldn't respond to that loco number. Set it back to 03 and it took of like a banshee. Good job it was on the rolling road or it would have gone crashing off the table! In the end, I re-programmed nearly all the main CV's (as per LGB manual defaults and it works properly once again. It seems that OPS mode programming works best with my digitrax dcc set.

I am living in temp accommodation at the moment , so am struggling with most of my belongings being in storage but I am gradually getting organised again. I intend setting up the configuration again on Friday and will read CV7 and CV8 and report back. Here's hoping LGB follow NMRA RP 9.2.2!
 
There is a 'calculator' on the Marklin site for high-number addresses..
However, the Ger-English makes no sense what so ever!! :o ::)

Looks to be a variation on the 'normal' way of calculating long addresses.. You also need to set a bit in CV29 as well, of course.
 
Managed to spend some time early this morning, also getting the drift of how the Digitrax system works and used to the rather spongy buttons on the handset.......

Table below uses decimal values
CV 7 set to 17
CV 8 set to 131

After some playing around, I reset following sound values to suit my interim surroundings
reduced CV 63 (total volume) from 255 to 191 so not so deafening generally
reduced CV 140 (volume for whistle blast) from 255 to 127

Also experimented with loco inertia settings but cannot judge when only on a rolling road.

Over weekend, I will be trying out long addresses to see whether Digitrax system sets them up to suit this marklin decoder or not and then will do so manually by manual direct CV changes myself.

Does anyone know the use of 2 topmost bits (bits 6 and 7) of CV 18 (correction - CV 17) and why they must be set to 1, ie decimal values of 128 and 64 (Totalling 192).
 
ernie said:
Does anyone know the use of 2 topmost bits (bits 6 and 7) of CV 18 (correction - CV 17) and why they must be set to 1, ie decimal values of 128 and 64 (Totalling 192).
Click to expand...

Thanks for all of the info.

The NMRA spec says this about the extended address:

Configuration Variables 17, 18: Extended Address
The Extended Address is the locomotives address when the decoder is set up for extended addressing
(indicated by a value of "1" in bit location 5 of CV#29). CV#17 contains the most significant bits of the two
byte address and must have a value between 11000000 and 11100111, inclusive, in order for this two byte
address to be valid. CV 18 contains the least significant bits of the address and may contain any value.


I believe the extended address can only be between 1 and 9999 in which case the top two bits are not needed since 9999 is represented as 10011100001111, i.e only 14 of the 16 available bits are used. Not sure why the NMRA decided these spare bits should be one, however why would you want to set them to anything else. Are you wanting to use an address bigger that 9999? Actually you can have any value up to 10239 since that is binary 10011111111111 (CV18 is any value, and CV17 up to 100111.
 
Wot he said! ;)
 
Update on experimenting wirh marklin decoder:

I have had inconsistent results with programming a change of long addresses (held in CV's 17 (msb) and 18(lsb)) and also bit 5 of CV 29. My Digitrax DCC system allows me to select 4 ways of programming (damn confusing but maybe good to have). Their manuals imply/say Direct Mode and then Paged Mode are the ones best for modern decoders and cover full range of CV's. Ops Mode for PoM (programming on the main) also works providing your decoder is up to it. This marklin decoder does offer this.

My limited testing with changing various CV's have all worked other than long addresses. Digitrax complains of NoAk (no acknowledgement) in some cases or just plain Fail. BUT when I program each CV, ie 17, 18 and bit 5 of CV 29 individually and only on the Programming Track, then I get totally consistent and working results. To calculate decimal values to use is not difficult but a simple calculator helps, unless you are used to modulo 256 arithmetic. :D

As I guess changing loco identity is not a frequent requirement for many, for me anyway, I can happily live with that. I can rest happily! But, I shall nevertheless try to see what the issue is.

Cliff, my query regarding 2 topmost bits of CV 17 is purely academic.
 
The thing is, having to calculate CV17 & 18 values and CV29 bits individually is archaic and not at all user friendly. DCC systems these days do it for you, and even better (in my opinion) are the computer-based software programmers such as JMRI - I use this in combination with SPROG hardware or connected to my NCE system.

My question a while back on another thread regarding the manufacturer ID and decoder model CVs for this Marklin decoder was because there didn't seem to be a JMRI decoder template yet for this decoder, which usually simply means no-one in the JMRI community has got round to creating one yet.
 
ntpntpntp said:
The thing is, having to calculate CV17 & 18 values and CV29 bits individually is archaic and not at all user friendly. DCC systems these days do it for you, and even better (in my opinion) are the computer-based software programmers such as JMRI - I use this in combination with SPROG hardware or connected to my NCE system.

My question a while back on another thread regarding the manufacturer ID and decoder model CVs for this Marklin decoder was because there didn't seem to be a JMRI decoder template yet for this decoder, which usually simply means no-one in the JMRI community has got round to creating one yet.
Click to expand...

There is a calculator (in German0 on the website.. Google translate fails in this instance. :(

If you use a Massoth CS / Nav, and the change address option, it fails due to the way the software alters the CV's ( ?perhaps the order it does it? ) - The Massoth alters the address CV's, CV29 (and possibly CV49, but not sure on that).

I would guess the cost of the decoder(s), and the relative rarity of 'G' items needing them yet means no-one in the JMRI/SPROG community has had to purchase and fiddle with one yet. - You tend to buy pre-programmed, and they just work.
 
Actually I think if you really want to understand something it is best to get down to the bits and bytes level.

I haven't looked in JMRI yet but I did check with some Trix decoders and these new LGB ones are very similar in terms of CVs. Maybe there is a Trix decoder definition in JMRI that could be used for these LGB ones.
 
PhilP said:
If you use a Massoth CS / Nav, and the change address option, it fails due to the way the software alters the CV's ( ?perhaps the order it does it? ) - The Massoth alters the address CV's, CV29 (and possibly CV49, but not sure on that).
Click to expand...
... which is extremely disappointing and very bad design of the part of Marklin if indeed this decoder doesn't behave well compared to the vast majority of NMRA DCC complaint decoders. You'd have thought they'd thoroughly test it at the very least with systems well known to be used with LGB products. Shooting themselves in the foot?

Cliff George said:
I haven't looked in JMRI yet but I did check with some Trix decoders and these new LGB ones are very similar in terms of CVs. Maybe there is a Trix decoder definition in JMRI that could be used for these LGB ones.
Click to expand...
Yes, that's what needs to be examined in detail.
 
An interesting thread this. A number of folk have had problems with new locos fitted with Märklin MFX decoders - as soon as they try to programme them they run erratically or even go magic smoke and the resultant burnt out decoder.

I have one loco with such a decoder - left on address "3" I have made no attempt to alter anything. I am not as competent as the likes of Cliff or Nick in terms of programming so any helpful info is read and appreciated with interest.
 
You must log in or register to reply here.
Back
Top Bottom