Welcome to the forum   
Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

Share
Options
View
Go to last post in this topic Go to first unread post in this topic
Offline AshleyH  
#1 Posted : 15 February 2016 11:33:54(UTC)
AshleyH

United Kingdom   
Joined: 15/02/2008(UTC)
Posts: 693
Location: Bournemouth, Dorset
With my gradually improving sight I have, quite literally, been dusting off my Marklin HO collection which has been stored in boxes for many years.

The other week I unboxed my 37990 Big Boy love and attempted to run it on a small oval of track. The loco did not run very well, so a friend helped me to strip it down, clean the brass contacts and add a little light lubrication, as per the instructions.

The loco now runs superbly once again, but there is a problem with the main chuffing steam loco sound effect. It starts out OK, but then at random the chuffing goes out of sync to the motion of the loco and makes almost a machine gun sound. This first version of the Big Boy has a pulse generator fitted to the front driving axle, later versions 37991 onwards do not have this. So my friend stripped and contact cleaned the magnets and sensor, but no change.

So after a bit of 'Googling' we found some information on the Stummi forum. It seems the sound decoder in the 37990 is not compatible with the 60213/60214 and 60215 Marklin Central Stations. I think the last time I ran the loco, maybe 9 or 10 years ago (yes really!) I would have used the Marklin 60212. Marklin service were kindly offering to upgrade a customer's 37990 for just over 300 Euros. However, mine was a gift and I would like to keep it in original condition. I also do not like the Marklin Service price and that was a few years ago!

So having no other digital controllers capable of generating the Motorola format, it is my intention to buy an additional controller as cheaply as possible that will operate my Big Boy correctly. There is mention of it working correctly with a Marklin Mobile Station, but which version? 60651/60652 or the later 60653? Does any member know the answer please? Also, I guess the 6021 or an Uhlenbrock Intellibox would be fine.

I already have a 60653, but not the 60113 track connection box, so that would probably be the cheapest option, but I would like to know if it works first, or whether I need to seek out one of the earlier Mobile Stations.

I hope this information will prove useful to other 37990 owners who may think their loco has developed a fault.

Best Regards
Ashley
thanks 5 users liked this useful post by AshleyH
Offline H0  
#2 Posted : 15 February 2016 11:52:44(UTC)
H0


Joined: 16/02/2004(UTC)
Posts: 15,254
Location: DE-NW
Hi, Ashley,

Just a guess, but maybe worth a try: you can set the CS2 to MM only (DCC and mfx will be disabled). This could help.

Do you use a switching-mode power supply with your CS2? Trying a transformer instead (16 V AC) could also make a difference.

I don't have the 37990 and cannot make any tests. Some CS2 seem to have problems with the 37990 while others don't.
Regards
Tom
---
"In all of the gauges, we particularly emphasize a high level of quality, the best possible fidelity to the prototype, and absolute precision. You will see that in all of our products." (from Märklin New Items Brochure 2015, page 1) ROFLBTCUTS
UserPostedImage
thanks 5 users liked this useful post by H0
Offline AshleyH  
#3 Posted : 15 February 2016 12:56:50(UTC)
AshleyH

United Kingdom   
Joined: 15/02/2008(UTC)
Posts: 693
Location: Bournemouth, Dorset
Hi Tom,

Thanks for your suggestions. I do use a switching mode power supply., so I will try my Roco 16v transformers. It may make a difference! I have already tried just setting the data format to MM only, it was a good idea, but did not solve the problem. I am more hopeful about the 16v transformer though, I will let you know when I have done the necessary wiring...

Best Regards
Ashley
thanks 2 users liked this useful post by AshleyH
Offline AshleyH  
#4 Posted : 15 February 2016 23:49:56(UTC)
AshleyH

United Kingdom   
Joined: 15/02/2008(UTC)
Posts: 693
Location: Bournemouth, Dorset
Hi Tom,

Well, this afternoon I went and found the box for my 60215 hoping that the cable with two wires at one end and the 4 pin plug at the other was still in the box, it was. So I connected the cable to the 16v AC outputs of my Marklin 6647, but sadly no difference, the same jumbled up chuffing sounds.

This evening I borrowed a friends 60652 Mobile Station, complete with a 60052 AC Transformer and the C track connection track. The Big Boys sound was perfect with this setup.
So I am lead to the conclusion that the MM signal generated by my CS2 is in some way incompatible with the 37990, and pissibly a few other Marklin locos. And it doesn't seem to matter what power supply you use with the CS2, though I have yet to try the 60052 with my CS2.

Many Thanks
Ashley
thanks 3 users liked this useful post by AshleyH
Offline dickinsonj  
#5 Posted : 16 February 2016 01:11:22(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Ashely,

I have had that same problem with my 37990 and I first had it while I was still running it under the first generation Marklin Central Unit (6020). When I removed the plastic cover over the axles where the signal generator is located, it had collected excess oil from oiling the drive wheels. I cleaned up the oil, cleaned the disk on the front axle that generates the signal and the sensor located above it. It did the out of beat sounds less often after that, but it still did it. It did not do that at all when it was new and running with the same digital setup. That was over 10 years ago and shortly after that my control units that went with that Central Unit stopped working. Since then all of my Marklins have also been packed away until just last month.

I finally bought a CS2 last December and when I ran the 37990 the sound was almost perfect with just the very occasional erratic sound. I cleaned up the oil in the area again (still too much oil after 10 years storage!) and now it works fine with the CS2. Mine runs as beautifully as ever. Could you be more specific about what contacts your friend cleaned? That may be something that I need to keep an eye on as well.

I understand that my experience is not consistent with what you have seen with your 37990 but I definitely had exactly that issue many years before I ever had a CS2. I assume it is related to the signal generating disk being affected by the wheel oil. At least that appears to be the situation with my 37990. That is my number one favorite Lok although I doubt that I would pay that cost to have Marklin repair it, even if that is feasible from the US these days.

Good luck getting that sorted out and let us know what happens.

Jim
Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
thanks 3 users liked this useful post by dickinsonj
Offline AshleyH  
#6 Posted : 17 February 2016 08:30:34(UTC)
AshleyH

United Kingdom   
Joined: 15/02/2008(UTC)
Posts: 693
Location: Bournemouth, Dorset
Hi Jim,

The contacts that were cleaned were the brass strips that return the current through the axles. I think they may have oxidised slightly in the years of storage and non use. The loco was tested with a multimeter and a few of the axles were not returning the current to the rails. But fine running after cleaning.

My chuffing sound issue is now definately only diwn to some sort of incompatibility with the 60215 CS2, I guess the way it generates the MM2 data format



I am thinking of buying a 60113 connection box for my existing 60653 Mobile Ststion, but it would be helpful to know if any forum members have run a 37990 Big Boy with the current generation Mobile Ststion and whether or not the sound works correctly?

I thought about buying one of the earlier Mobile Stations like my friend lent me, but I don't really want to buy older, less reliable technology.

Many Thanks
Ashley

thanks 2 users liked this useful post by AshleyH
Offline dickinsonj  
#7 Posted : 17 May 2016 20:53:53(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Ashley,

I recently made a discovery regarding the 37990 Big Boy and the problem with the operating sound synchronization.

If I operate the Big Boy on a circuit controlled directly by my CS2 (60215) the sound quickly loses sync and I just turn if off. But if I operate the 37990 from a 6015/6017 booster connected to the port on the bottom of the CS2, the sound is fine. That suggests that the data is indeed corrupted in the 60215 data stream but that something happens when it is ported to the old style boosters that makes it work correctly.

I write software of this nature professionally and one of these days I will get the source code and figure out how this goes wrong. I am guessing that it is related to both decoders in this model presenting under just one address on the CS2. The first decoder controls the lights and motor and the second decoder controls the sounds. My other two decoder models let the the decoders present separately and I have to control the second decoder under a second address. Even though the decoders in the 37990 are on consecutive addresses they are both accessible under the address of the first decoder. However Marklin made that happen seems to corrupt the sound signal and it should be fixable in software. However that would require recompiling new software with my change every time the CS2 is updated.

In the short term I will probably just confine the Big Boy to a track controlled by the old style boosters, under which it works perfectly. I have tested my hypothesis multiple times with the same result, so I think I have a kludge that will get me by. If I come up with a better solution I will post it on the forum. Maybe Marklin would like to buy my solution - I can be easily paid in high quality Marklin equipment. Cool

I hope that you have found a workable solution to your problem.

Regards,

Jim

Edited by user 18 May 2016 01:01:32(UTC)  | Reason: Not specified

Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
thanks 2 users liked this useful post by dickinsonj
Offline H0  
#8 Posted : 18 May 2016 08:07:56(UTC)
H0


Joined: 16/02/2004(UTC)
Posts: 15,254
Location: DE-NW
Originally Posted by: dickinsonj Go to Quoted Post
That suggests that the data is indeed corrupted in the 60215 data stream but that something happens when it is ported to the old style boosters that makes it work correctly.
Nope, the boosters emit the same data stream - just at a higher voltage. 16 V AC give more than 20 V track voltage while 19 V DC give about 18 V track voltage.
The MS2 will give about 16 V track voltage from 18 V DC.

mfx may increase the problems, but mfx is also present in the 6017 data stream - as it is the same data stream.

Originally Posted by: dickinsonj Go to Quoted Post
Even though the decoders in the 37990 are on consecutive addresses they are both accessible under the address of the first decoder.
Nope. F5 through F8 are not sent to the first address, they are sent as F1 through F4 to the second address. Simple address translation in the CS2, no "decoder magic".
Regards
Tom
---
"In all of the gauges, we particularly emphasize a high level of quality, the best possible fidelity to the prototype, and absolute precision. You will see that in all of our products." (from Märklin New Items Brochure 2015, page 1) ROFLBTCUTS
UserPostedImage
thanks 1 user liked this useful post by H0
Offline dickinsonj  
#9 Posted : 18 May 2016 14:00:28(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Originally Posted by: H0 Go to Quoted Post
Nope, the boosters emit the same data stream - just at a higher voltage. 16 V AC give more than 20 V track voltage while 19 V DC give about 18 V track voltage.
The MS2 will give about 16 V track voltage from 18 V DC.

mfx may increase the problems, but mfx is also present in the 6017 data stream - as it is the same data stream.

Well, I assumed that it was the same data stream - as it should be. But the sound definitely works properly with the 6017 booster and not with the 60215. I don't see why the voltage difference should effect this function's operation but I suppose that is possible. Perhaps someone can explain how this voltage difference could cause this error in the sound.
Originally Posted by: H0 Go to Quoted Post
Nope. F5 through F8 are not sent to the first address, they are sent as F1 through F4 to the second address. Simple address translation in the CS2, no "decoder magic".

I was trying to not be too technical, since it seems to put some people off. That is what I meant by "decoder magic" - the CS2 is clearly doing the address translation for the second decoder. Debugging from small amounts of info can often lead you astray, as I well know.

When Ashley suggested that the CS2 was at fault it didn't make sense to me either. The operating sound function is binary - the sound is either on or off, which is all the CS2 needs to send, since the actual sounds are stored in the decoder. That would be normal practice for a CAN bus system like this, so I assume this is the case for this system. But now I wonder if most iterations of the data message have the operating sounds "on" but occasionally it interjects a message with that bit "off", which might cause the stutter.

The decoders are on adjacent addresses but I don't know how the CS2 determines that this is a case where F5 - F8 should be send as F1 - F4 on the next consecutive address. Maybe it does that in any case where there are more than 4 functions but it is not an mfx decoder? When I first placed the 37990 on the programming track I was surprised to see all 8 functions show up under one address and I wondered how the CS2 determined that this was a special case.

I also wondered what would happen if I remapped the decoders, but I was not sure how to do that since the CS2 only presents the address of the first decoder. Both decoders are still on the factory setting, 40 and 41, but I assume it would work with any two consecutive addresses.

So I bow to your greater knowledge of this system, but I still have the problem with the operating sounds on my very nicest lok and I don't know what I can do to fix it! Confused
Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
Offline H0  
#10 Posted : 18 May 2016 15:31:53(UTC)
H0


Joined: 16/02/2004(UTC)
Posts: 15,254
Location: DE-NW
The MM protocol is limited to F0 through F4. So for MM protocol and F5 through F8 the next address will be used automatically. This is Encoder magic (the CS2 is the Encoder, not the Decoder).

If you have to MM locos with consecutive address and you define F5 through F8 for the loco with the lower address, then this entry in the CS2 will control the functions of the other loco.


Re effect of higher/lower track voltage:
When the first MäTrix mfx decoders came up in 2009 there were some complaints about distorted sounds. Märklin added buffering capacitors to the locos that were sent back and also to locos issued later.
While the CS2 does not emit power to the track, the power stored in a capacitor on the decoder must provide current for the sound. The voltage of the capacitor drops while it gets depleted.
With higher track voltage, the minimum voltage from this capacitor (in a semi-depleted state) will still be higher. This may make the difference between normal sound and disturbed sound.

So maybe a "power pack" (capacitor, resistor, diode) would be enough to cure the problems. In my opinion it is a good sign that it works when the 6017 is involved.

The Big Boy 37990 has two decoders - one with 10 DIP switches, one with 8 DIP switches. While it makes sense to use consecutive addresses, you can also set non-consecutive addresses.
Regards
Tom
---
"In all of the gauges, we particularly emphasize a high level of quality, the best possible fidelity to the prototype, and absolute precision. You will see that in all of our products." (from Märklin New Items Brochure 2015, page 1) ROFLBTCUTS
UserPostedImage
thanks 2 users liked this useful post by H0
Offline xxup  
#11 Posted : 18 May 2016 16:06:12(UTC)
xxup

Australia   
Joined: 15/03/2003(UTC)
Posts: 9,457
Location: Australia
Originally Posted by: H0 Go to Quoted Post
... but mfx is also present in the 6017 data stream - as it is the same data stream...


Okay there might be something subtle in here AND it has been many years since I tried it, but I seem to remember using my eCOS 1 with a 6017 booster and trying to drive a mFx loco into the boosted area - where it stopped.. My understanding was that the 6017 could not retransmit the mFx signal onto the track. This is why I bought the eCOSBoost when it came out some time later.
Adrian
UserPostedImage
Australia flag by abFlags.com
Offline H0  
#12 Posted : 18 May 2016 16:37:12(UTC)
H0


Joined: 16/02/2004(UTC)
Posts: 15,254
Location: DE-NW
The 6017 transmits mfx from the controller to the loco, but not the other way around. You cannot register or edit mfx locos on the 6017 section, but you can drive them.
Regards
Tom
---
"In all of the gauges, we particularly emphasize a high level of quality, the best possible fidelity to the prototype, and absolute precision. You will see that in all of our products." (from Märklin New Items Brochure 2015, page 1) ROFLBTCUTS
UserPostedImage
Offline Armando  
#13 Posted : 19 May 2016 06:05:19(UTC)
Armando

United States   
Joined: 21/07/2003(UTC)
Posts: 1,350
Location: Houston, Texas
Hi all,

I've too experienced exactly the same erratic, machine-gun-like, sound from my 37990 the first time that I ran it using my CS2. I thought that the sound decoder had become damaged after having had the locomotive stored, and not run for over 10 years. I tried operating it with my old faithful 6021 Control Unit, and the sound operation was perfect, just like before.

I still haven't tried operating it with a MS2 mobile station, though.
Best regards,
Armando García

Offline dickinsonj  
#14 Posted : 19 May 2016 13:44:09(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Originally Posted by: Armando Go to Quoted Post
Hi all,

I've too experienced exactly the same erratic, machine-gun-like, sound from my 37990 the first time that I ran it using my CS2. I thought that the sound decoder had become damaged after having had the locomotive stored, and not run for over 10 years. I tried operating it with my old faithful 6021 Control Unit, and the sound operation was perfect, just like before.

I still haven't tried operating it with a MS2 mobile station, though.

Yes, this appears to be a problem shared among all of the 37990's, which is too bad since it is such an amazing locomotive!

Ashely mentioned that Marklin had modified some to correct this problem but I have not yet found any details about this modification.

For now I am just glad that mine seems to operate correctly running under one of my old boosters - that gives me a reason to keep some of them around as I change over to 60174 boosters and switched mode power packs. You might be fine if you use yours with your MS2.

Regards,

Jim
Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
Offline dickinsonj  
#15 Posted : 24 March 2017 02:32:12(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
I now have some free time to update a few old threads, and this time I have some very good news to share!

For a variety of reasons I finally bit the bullet this year and devised a functional power bus that I could install under my C track seasonal layout. My mainlines are approximately 18 meters long and I really had no choice if I wanted things to operate properly. I have another older thread where I asked for info on how I might do that and I will update it soon with some pictures of what I did.

I have found that everything works much better now with fewer amps consumed and no dead spots anywhere on my mainlines. People have posted in other threads about other issues that they thought might be caused by not having a smooth, consistent power supply. That made me wonder if the sound problems with my 37990 might be due to that as well. I just got that loco out and running this week and I was very happy to find that the operating sound now works perfectly in all of the areas where I have my power bus installed. Since that is everywhere except my small yard, I can enjoy the original, if basic sound of that loco for the first time in many years. BigGrin

Apparently that model's decoder is just very sensitive to any areas of weak signaling due to a poor power distribution which resets the operating sounds and screws up the chuff sequencing. So there was not really anything wrong with the loco itself other than that and a good electrical supply fixes the problem. Adding a capacitor might be a good idea as well, but for now it is working just fine.

I do find it curious that the problem was worse with my CS2 than with the old boosters (6015/6017) that I used last year, but I am no longer using anything but 60174 and 60175 boosters. So here is more proof of how essential a power bus is to a layout that runs well, even if it is just a temporary one.

And as a bonus I can use my 37990 as my signal quality tester as I build out my real layout in a few years. Cool
Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
thanks 5 users liked this useful post by dickinsonj
Offline H0  
#16 Posted : 24 March 2017 09:05:10(UTC)
H0


Joined: 16/02/2004(UTC)
Posts: 15,254
Location: DE-NW
Hi!
Originally Posted by: dickinsonj Go to Quoted Post
I do find it curious that the problem was worse with my CS2 than with the old boosters (6015/6017) that I used last year, but I am no longer using anything but 60174 and 60175 boosters.
When the load is low, the 6017 will provide a higher track voltage than the 60174.

Thanks for the feedback. Everybody building a permanent layout should be aware that more feeder tracks can solve some issues.
Regards
Tom
---
"In all of the gauges, we particularly emphasize a high level of quality, the best possible fidelity to the prototype, and absolute precision. You will see that in all of our products." (from Märklin New Items Brochure 2015, page 1) ROFLBTCUTS
UserPostedImage
thanks 4 users liked this useful post by H0
Offline AshleyH  
#17 Posted : 24 March 2017 10:12:04(UTC)
AshleyH

United Kingdom   
Joined: 15/02/2008(UTC)
Posts: 693
Location: Bournemouth, Dorset
An interesting post. After trying both switch mode and traditional transformers and getting the same issues with my 37990 and CS2 60215, I had come to the conclusion that the problem is the way the CS2 generates its data format.

One thing I will try later is running my CS2 with the 100VA PSU in 5 Amp mode. However, if voltage or current are the issue, then it puzzles me that the Big Boy works perfectly when driven with a MS1 or MS2 with a PSU of only 1.9 Amps or lower......

Another possibility is has this issue been resolved by a recent firmware update to the CS2.

I'll report back after more testing.....

Best Regards
Ashley
thanks 1 user liked this useful post by AshleyH
Offline dickinsonj  
#18 Posted : 24 March 2017 12:49:16(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Originally Posted by: H0 Go to Quoted Post
Hi!
Originally Posted by: dickinsonj Go to Quoted Post
I do find it curious that the problem was worse with my CS2 than with the old boosters (6015/6017) that I used last year, but I am no longer using anything but 60174 and 60175 boosters.
When the load is low, the 6017 will provide a higher track voltage than the 60174.

Thanks for the feedback. Everybody building a permanent layout should be aware that more feeder tracks can solve some issues.

Hi Tom,

You mentioned last year that the main difference between the two boosters was voltage but I could never figure out exactly how that might cause the problem.

It was a real wake up call to me about how important it is to have a power bus with lots of feeder wires if you want things to work properly. It did take a while to hammer that lesson into my thick head. It was also instructive on how difficult it can be to pin down the specific cause of a subtle digital issue like the sound problems with that loco. I was provided a lot of good input here on the forum though and I got there in the end. BigGrin
Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
Offline dickinsonj  
#19 Posted : 24 March 2017 12:56:30(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Originally Posted by: AshleyH Go to Quoted Post
An interesting post. After trying both switch mode and traditional transformers and getting the same issues with my 37990 and CS2 60215, I had come to the conclusion that the problem is the way the CS2 generates its data format.

One thing I will try later is running my CS2 with the 100VA PSU in 5 Amp mode. However, if voltage or current are the issue, then it puzzles me that the Big Boy works perfectly when driven with a MS1 or MS2 with a PSU of only 1.9 Amps or lower......

Another possibility is has this issue been resolved by a recent firmware update to the CS2.

I'll report back after more testing.....

Best Regards
Ashley

Hi Ashley,

Some obscure difference in the data format from the CS2 was my first guess as well. I am running the latest software on my CS2 and the sound will still lose synch in areas without a power bus, so I don't think the difference is the CS2.

Please do report what you find after more testing. Debugging something like this can be very difficult and I wandered down multiple wrong paths before finding my solution. Every time I wandered off into the weeds with some half formed theory though, Tom would pull me back on track with some helpful insights. There is a wealth of good information available here on this forum and I don't think that I could handle this stuff on my own without that.

Good Luck!
Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
Offline Incognito  
#20 Posted : 16 April 2017 18:21:28(UTC)
Incognito

United States   
Joined: 19/04/2015(UTC)
Posts: 30
Location: Maryland, Chevy Chase
I just acquired a Big Boy - 37990 and also noticed the problem with the sound. My CS2 is at the latest firmware version. I use a 60174 boosters for the track, but when I replaced it with an older 6017, the sound was perfect. That really eliminates the CS2 and the locomotive and only leaves the 60174 as the culprit.
Alex
Offline dickinsonj  
#21 Posted : 16 April 2017 22:48:46(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
I have another sound update on my 37990.

As I had reported earlier, it has been working perfectly on all on my circuits as long as they have a power bus. This year they are supplied by the booster in my CS2 and a 60174 and a 60175 booster. So the sound issue seemed to have been solved, at least for a while. But eventually there was another sound problem and I got the chance to learn one more detail about this situation. That train was parked for about 10 days on one of my main lines (running from the 60215) and when I got around to running it again the sound issues had resurfaced. They were not as bad as what I had seen in the past but they were definitely there.

Eventually I noticed that the problem only occurred when the loco was in the section where it had been parked for 10 days. That loco pulls 25 wagons with lots of newly oiled axles so I decided to try cleaning just the track where it had been sitting. After cleaning that track segment it is now back to working perfectly again. I have run it a good bit over the last 2-3 weeks without a single sound problem. My latest theory is that the decoder in this model is extremely sensitive to ANY disturbance in the digital signal, even just the signal degradation that dirty track can cause. A new decoder would no doubt solve the problem and a capacitor might fix it also, but for now I am leaving it alone.

Originally Posted by: Incognito Go to Quoted Post
I use a 60174 boosters for the track, but when I replaced it with an older 6017, the sound was perfect. That really eliminates the CS2 and the locomotive and only leaves the 60174 as the culprit.

At this time last year I was convinced that there was some data format difference between the 60215 and the 6017 which caused this issue. But all of my 6015/6017 boosters are now retired and my 37990 sound is working perfectly on all of the newer circuits as long as there is a power bus and the track is clean. I think the problem is less prevalent with the 6017 because of the voltage differences between the older and newer boosters as Tom suggested to me last year. I have seen the problem when it is on circuits powered by a 6020, 60215, 60174 and 60175, so I don't think it is the digital output from any particular booster causing this problem but rather a combination of track voltage and having a good, clean power supply.
Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
Offline Incognito  
#22 Posted : 20 April 2017 21:10:16(UTC)
Incognito

United States   
Joined: 19/04/2015(UTC)
Posts: 30
Location: Maryland, Chevy Chase
I will clean the tracks again since I havn't done it in a while and see if that resolves the issue. Thanks.
Alex
thanks 1 user liked this useful post by Incognito
Offline AshleyH  
#23 Posted : 08 February 2024 17:20:38(UTC)
AshleyH

United Kingdom   
Joined: 15/02/2008(UTC)
Posts: 693
Location: Bournemouth, Dorset
Well, nearly 7 years on, it’s time for an update to this thread.

Time and technology has moved on, I am now running a Marklin CS3 Plus and CS3 setup, powered by the LGB 51095 PSU at 5 amps.

My 37990 was recently tested, and the sound once gain works perfectly. However, if I change to the programming track output on the CS3, then the machine gun noise returns.
So 5 amps to the track, no issues, lower amps from the programming track output, old problem re5urnd.

I hope this information is of some use to 37990 Big Boy owners.
Nonetheless, the time has probably come for me to upgrade the decoder to MsD3.

Best Regards, Ashley

Edited by user 19 February 2024 12:19:18(UTC)  | Reason: Not specified

Offline JohnjeanB  
#24 Posted : 08 February 2024 21:38:06(UTC)
JohnjeanB

France   
Joined: 04/02/2011(UTC)
Posts: 3,084
Location: Paris, France
Hi Ashley
My experience is a little different.
I discovered that my 37990 Big Boy was not working with my CS2 60215.
So I though it was failing. As a last resort I used my CS1 and its feed rail and surprise surprise it worked perfectly.
Checking again, I found that the chuff chuff sound was triggered by a Hall Transistor located on the front bogie first driving wheel.
To bring this signal to the sound module in the tender, there is quite a long wire.
Depending on the digital signal (multi protocol of single) so electrical noise goes into this detection circuit and started this avalanche of sounds.
Probably this may also be linked to aging capacitor (whose value reduces over the years).
So I did not went any fiurther as my loco is a collector and poorly suited for my European-style layout.
Back to its wooden box without any change.
For those willing to fix the issue my advice is to replace the normal wire with a thin coaxial one.
Cheers
Jean
Offline dickinsonj  
#25 Posted : 09 February 2024 01:00:39(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Originally Posted by: JohnjeanB Go to Quoted Post

For those willing to fix the issue my advice is to replace the normal wire with a thin coaxial one.

That is an interesting observation Jean.

My 37990 was fine with my old 6020 and the problem only started when I got a 60215. It was also present after I upgraded to a CS3+.

But after I installed a power bus the problem went away, although I was never sure exactly why. Maybe the root cause is the thin wire between the Hall sensor and the decoder board, s you suggest. That could also explain why Ashley sees the problem on the programming track and not on the main.

I have moving to a modern decoder for that loco on my very long list of upgrades. I have done many other upgrades but the originality of that loco and it's complexity have kept that project on long term hold. I no longer have the sound problem but the gigantic speed steps make it seem very old compared to my newer Big Boys and Challengers. It is the one loco that I will probably get professional help with when (if) I decide to move ahead with that.

It would be interesting to know if your wire replacement idea was the answer to this well known problem all along. I have read some posts that indicated that Marklin service had fixed some 37990s, but I never learned exactly how they accomplished that.

Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
Offline einotuominen  
#26 Posted : 09 February 2024 07:28:26(UTC)
einotuominen

Finland   
Joined: 19/09/2022(UTC)
Posts: 378
Location: Kaarina
Originally Posted by: AshleyH Go to Quoted Post

Nonetheless, the time had probably come fi4 m3 to upgrade the deciders to MsD3.

Best Regards, Ashley


If you decide to upgrade, go with ESU Loksound 5. You’ll get better running characteristics and to be honest, better sound files.

-Eino
thanks 2 users liked this useful post by einotuominen
Offline bph  
#27 Posted : 09 February 2024 09:39:37(UTC)
bph

Norway   
Joined: 04/08/2018(UTC)
Posts: 985
Originally Posted by: AshleyH Go to Quoted Post
Well, nearly 7 years on, it’s time for an update to this thread.

Time and technology has moved on, I am now running a Marklin CS3 Plus and CS3 setup, powered by the LGB 51095 PSU at 5 amps.

My 37990 was recently tested, and the sound once gain works perfectly. However, if I change to the programming track output on the CS3, then the machine gun noise returns.
So 5 amps to the track, no issues, lower amps from the programming track output, old problem re5urnd.

I hope this information is if some use to 37990 Big Boy owners.
Nonetheless, the time had probably come fi4 m3 to upgrade the deciders to MsD3.

Best Regards, Ashley


Hi

what track voltage do you read out on the CS3 when using the LGB 51095?

Before I upgraded my 37990 with a mSD3, I did some quick tests.
1. it struggled on the cs3 with a 60041
2. went back to my old 6021 + 6002, and it worked perfectly fine
3. tested with CS3 and the booster 60175 powered by the 6002 transformation, and the same track voltage as 6021. and it worked perfectly fine

based on that, it is possible that the 19,3v track voltage is on the borderline of working the voltage for the old decoders in 37990. so it becomes very sensitive to working properly. Of course, other things might also affect it.

Since the wheel sensor on mine worked fine, I wired it up to the new mSD3, and it now works fine on the CS3 +60041. (note it's not necessary, but lie like to use it as long as it works.)
And the Maxon motor used in early big boys is a beast. ThumpUp


thanks 2 users liked this useful post by bph
Offline dickinsonj  
#28 Posted : 09 February 2024 16:57:57(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Originally Posted by: bph Go to Quoted Post

Since the wheel sensor on mine worked fine, I wired it up to the new mSD3, and it now works fine on the CS3 +60041. (note it's not necessary, but lie like to use it as long as it works.)
And the Maxon motor used in early big boys is a beast. ThumpUp


I also love that axle sensor and it has been a large part of why I have not converted this loco.

So were you were able to get the Hall sensor to trigger chuffs on an mSD3? I was also never sure what the full purpose of the daughter board is. I seem to remember that the sensor connected to it. I refer to the one in the front of the boiler and how that would impact things.

I fully agree about those Maxon motors. In my perfect MRR world all of my locos would have motors of that quality. ThumpUp

A typical engineer overthinking something that is probably quite simple. BigGrin
Originally Posted by: AshleyH Go to Quoted Post

My 37990 was recently tested, and the sound once gain works perfectly. However, if I change to the programming track output on the CS3, then the machine gun noise returns.
So 5 amps to the track, no issues, lower amps from the programming track output, old problem sound.

That is what I think as well Ashley. It fits in with my experience of fixing that problem by improving the uniformity of my track power with a well designed bus.
Originally Posted by: einotuominen Go to Quoted Post
If you decide to upgrade, go with ESU Loksound 5. You’ll get better running characteristics and to be honest, better sound files.

I agree that he ESU v5 loksound is an excellent decoder and has access to better sound files in my experience. I would probably use an mSD3 just for admin uniformity, where most of my locos are mfx. But I have DCC and MM and ESU decoders in the mix and it all works just fine. So I should reconsider using a Loksound whenever I go ahead with this long, long delayed project.
Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
Offline bph  
#29 Posted : 09 February 2024 18:54:14(UTC)
bph

Norway   
Joined: 04/08/2018(UTC)
Posts: 985
Originally Posted by: dickinsonj Go to Quoted Post
Originally Posted by: bph Go to Quoted Post

Since the wheel sensor on mine worked fine, I wired it up to the new mSD3, and it now works fine on the CS3 +60041. (note it's not necessary, but lie like to use it as long as it works.)
And the Maxon motor used in early big boys is a beast. ThumpUp


I also love that axle sensor and it has been a large part of why I have not converted this loco.

So were you were able to get the Hall sensor to trigger chuffs on an mSD3? I was also never sure what the full purpose of the daughter board is. I seem to remember that the sensor connected to it. I refer to the one in the front of the boiler and how that would impact things.

I fully agree about those Maxon motors. In my perfect MRR world all of my locos would have motors of that quality. ThumpUp

A typical engineer overthinking something that is probably quite simple. BigGrin


I did not change any wiring on the front daughterboard when I upgraded, and it just passes on the wheel sensor wires.

the wheel sensor is wired up to the same PCB as the front light (marked as 67 on the Marklin exploded drawing), and it also uses decoder orange and purple.

The wheel sensor wire is the black wire and the decoder purple changes color to red on the front daughterboard.
https://www.hovland.net/bilder/mj/R5_A3309.jpg
note the Red and black wire in the lower left corner, that goes "out" of the daughter board and "down" into the chassis. And the yellow cable in the picture (upper left corner) is actually the orange feed to the wheel sensor and front light. (note I have replaced some aux wires on the picture)

On the msd3 the sensor will only work if you wire up the black wire to the IN3 pad/pin. Note, that the updated Märklin sound project for Big Boy is quite a bit better than the ones first released for the Marklin mSD decoders.

Edited by user 09 February 2024 22:59:11(UTC)  | Reason: Not specified

Offline dickinsonj  
#30 Posted : 09 February 2024 20:23:25(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Thanks bph.

That is good information, and makes that system a lot clearer.

The sounds on my 4014 are very nice and it would be a big improvement to have those on my 4013.
Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
Offline bph  
#31 Posted : 09 February 2024 22:49:04(UTC)
bph

Norway   
Joined: 04/08/2018(UTC)
Posts: 985
Is your 4014 the 37997? if it is, was it delivered with the old or new sound project?

The new one is the one currently available on Märklin website. I only noticed because I downloaded the original 37997 project, and used that as a base for my 37990. and later noticed that Marklin had updated the 37997 project. (I have I copy of both versions)

A side note on the 37990, is that according to Maxon, the max. continuous current on the motor used is 0,681 A, but the stall current can be 2.17 A. So be careful and try to avoid stalling it. Hopefully, the msd3 decoder overload protection should be able to handle it. but I have not tested that. But it's worth keeping that in mind regardless of what decoder brand you choose to use. (some Zimo decoders can handle that within specifications)
thanks 2 users liked this useful post by bph
Offline dickinsonj  
#32 Posted : 10 February 2024 01:37:10(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Originally Posted by: bph Go to Quoted Post
Is your 4014 the 37997? if it is, was it delivered with the old or new sound project?
A side note on the 37990, is that according to Maxon, the max. continuous current on the motor used is 0,681 A, but the stall current can be 2.17 A. )

My 4014 is a 37997, but I don't know which sound files it came with. It still has the factory decoder so there isn't really much I can do anyway, but it does sound nice.

I will keep that in mind about the Maxon current draw. The 37990 runs so well and so strongly that something catastrophic needs to occur to stall that beast. But it has happened and I believe the Earth's rotation was slightly changed. BigGrin
Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
Offline bph  
#33 Posted : 10 February 2024 13:31:36(UTC)
bph

Norway   
Joined: 04/08/2018(UTC)
Posts: 985
Originally Posted by: dickinsonj Go to Quoted Post
Originally Posted by: bph Go to Quoted Post
Is your 4014 the 37997? if it is, was it delivered with the old or new sound project?
A side note on the 37990, is that according to Maxon, the max. continuous current on the motor used is 0,681 A, but the stall current can be 2.17 A. )

My 4014 is a 37997, but I don't know which sound files it came with. It still has the factory decoder so there isn't really much I can do anyway, but it does sound nice.

I will keep that in mind about the Maxon current draw. The 37990 runs so well and so strongly that something catastrophic needs to occur to stall that beast. But it has happened and I believe the Earth's rotation was slightly changed. BigGrin


a 37997 with the new sound project:

and with the old sound sound project.

I'm not sure but it seems like Märklin delivered some early 37997 with the old project and that they changed it during production.
the original 37997 project is 7,3 MB and the updated one is 13,1 MB, the "downside" with the new project is that they removed some of the deeper tones from the sound files.

And yes, if you have a CS3 and it has a network connection to your PC you should be able to update the sound project on the 37997 original factory-fitted decoder. I don't have the 37997, but I have done it on many other newer locomotives. (using the 60971 don't work)
thanks 2 users liked this useful post by bph
Offline dickinsonj  
#34 Posted : 10 February 2024 15:01:15(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Originally Posted by: bph Go to Quoted Post

And yes, if you have a CS3 and it has a network connection to your PC you should be able to update the sound project on the 37997 original factory-fitted decoder. I don't have the 37997, but I have done it on many other newer locomotives. (using the 60971 don't work)

That is interesting, and thanks for posting the two sets of Big Boy sounds.

I had forgotten about the hack where you could update those decoders from the CS3, but not with the decoder programmer. I find that odd and irritating, because using the 60971 is my preferred way of updating Märklin decoders.

The next time I have access to my 37997 I will have to see which set of sounds I have, and go from there.

Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
thanks 1 user liked this useful post by dickinsonj
bph
Offline rhfil  
#35 Posted : 11 February 2024 22:10:20(UTC)
rhfil

United States   
Joined: 05/09/2014(UTC)
Posts: 422
Location: NEW HAMPSHIRE, Somersworth
Marklin seems to use files and pointers to them for things like loco images and cab images. I wonder if they did the same with engine sounds and somehow the cv is directing the controller to a sound location that now has the wrong file?
thanks 1 user liked this useful post by rhfil
Offline H0  
#36 Posted : 12 February 2024 08:38:14(UTC)
H0


Joined: 16/02/2004(UTC)
Posts: 15,254
Location: DE-NW
Originally Posted by: rhfil Go to Quoted Post
I wonder if they did the same with engine sounds and somehow the cv is directing the controller to a sound location that now has the wrong file?
We are talking about a loco that has two decoders because one decoder is limited to 5 functions. Those decoders have DIP switches for the address and everything else is hard-coded.
Electrical interference at the Hall sensor sounds more plausible.
Regards
Tom
---
"In all of the gauges, we particularly emphasize a high level of quality, the best possible fidelity to the prototype, and absolute precision. You will see that in all of our products." (from Märklin New Items Brochure 2015, page 1) ROFLBTCUTS
UserPostedImage
thanks 1 user liked this useful post by H0
Offline dickinsonj  
#37 Posted : 12 February 2024 12:36:44(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Tom is correct. The image files that are accessed by pointers are for things stored in the controller, not on the loco decoder.

I think that Jean has the right idea about the long wire from the Hall sensor to the decoder being the culprit here.
Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
thanks 1 user liked this useful post by dickinsonj
H0
Offline bph  
#38 Posted : 14 February 2024 23:20:27(UTC)
bph

Norway   
Joined: 04/08/2018(UTC)
Posts: 985
Could it also be a combined cause?

the 37990 was made around the time when 6021 was the standard with approx 21v track voltage.

so could it be the following?

the lover standard track voltage from the CS3 is just on the borderline of trouble-free operation. (CS3 powered by a standard 60041)

the wheel sensor gets lover voltage and thus sends back a weaker signal to the sound decoder.
(the tracking sensor is powered by the locomotive decoder located in the boiler)

The long wire with the signal goes from the front of the locomotive all the way back to the sound decoder. (black wire)
And in the signal wire goes over the speaker in the tender.

So could it be that the speaker magnet actually introduces interference into the signal wire and that because of the weaker signal from the sensor, the sound decoder struggles to separate the real signal and the "noise" signals created by the speaker? so in i way, when the speaker makes chuff sounds, the speaker magnet triggers more chuff signals and thus creates the chain gun sound?

I'm not an electronic expert so others with more knowledge have to comment on how plausible this is ?.
or someone who has the original decoder in their 37990, can test to temporarily move the sensor cable away from the speaker etc..
Offline dickinsonj  
#39 Posted : 15 February 2024 01:16:58(UTC)
dickinsonj

United States   
Joined: 05/12/2008(UTC)
Posts: 1,676
Location: Crozet, Virginia
Originally Posted by: bph Go to Quoted Post
Could it also be a combined cause?

the 37990 was made around the time when 6021 was the standard with approx 21v track voltage.

I think that is what is happening, whether interference from the speaker plays into it or not. With the reduced voltage this system is no longer fully reliable and is on the edge of functioning. I think that also explains why my 37990 was fine again after a serious power bus upgrade on my layout.

Originally Posted by: bph Go to Quoted Post
or someone who has the original decoder in their 37990, can test to temporarily move the sensor cable away from the speaker etc.
.
I might give that a try when I service that loco. It is at top of the maintenance queue when I start working on trains again, so I can take a look. Mine is my power distribution tester, because if it keeps its sound intact my wiring is doing great. If not I can figure out why.

One of my crocs from my dual set also has sound issues if anything is less than perfect electrically and it is more nimble, so that helps. I like to define locos that are fine when things are perfect to help me get to my layout quality goals. BigGrin
Regards,
Jim

I have almost all Märklin and mostly HO, although I do have a small number of Z gauge trains!
So many trains and so little time.
thanks 1 user liked this useful post by dickinsonj
bph
Offline Goofy  
#40 Posted : 16 February 2024 18:28:51(UTC)
Goofy


Joined: 12/08/2006(UTC)
Posts: 8,993
I did checked about 37990 and Märklins CS for the past years.
I must saying that there was a lot digital interference with those CS too.
37990 have old fx sound decoder and it has two adress.
I would saying that old Märklin trains does not always works perfect with older CS due to data transmission problems.
To change out sound decoder in the 37990 would help a lot but it will cost.
Märklins latest Big Boy was 37997 and have mfx+ sound decoder which are also totally difference from other models before.
To use all data in the mfx+ sound decoder you must update older CS 60213, 60214 and 60215 to software 4.2.
H0
DCC = Digital Command Control
Users browsing this topic
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.

| Powered by YAF.NET | YAF.NET © 2003-2024, Yet Another Forum.NET
This page was generated in 1.446 seconds.