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 intruder  
#1 Posted : 20 October 2008 21:38:11(UTC)
intruder

Norway   
Joined: 16/08/2006(UTC)
Posts: 5,382
Location: Akershus, Norway
Hi all

I have red a lot about Alzheimer, that certain decoders suffer from this.
But I have not got the general idea about what the decoder Alzheimer really is. How does a locomoyive work / not work when it has got the Alzheimer?
Best regards Svein, Norway
grumpy old sod
Offline RayF  
#2 Posted : 20 October 2008 22:00:00(UTC)
RayF

Gibraltar   
Joined: 14/03/2005(UTC)
Posts: 15,840
Location: Gibraltar, Europe
Svein,

The best way to think of it is that the loco "forgets" the speed it was going at when it hits a momentary break in track current. It goes straight to speed step zero.

When, a moment or two later, it sees the track current again, it reads what speed it has been requested and proceeds to accelerate slowly up to it.

This is my interpretation of the "alzheimer effect".

Ray
Ray
Mostly Marklin.Selection of different eras and European railways
Small C track layout, control by MS2, 100+ trains but run 4-5 at a time.
Offline WelshMatt  
#3 Posted : 20 October 2008 22:22:36(UTC)
WelshMatt


Joined: 06/10/2006(UTC)
Posts: 1,345
Location: ,
That sounds more like the problems most decoders have with very dirty track. I have a G scale loco with an ESU DCC decoder that did this until I added a capacitor (as advised in the instructions) which smooths the power delivery and keeps it moving over brief power interruptions.
Matt from Wales.

When you pay Range Rover prices, don't accept Lada quality
Offline perz  
#4 Posted : 20 October 2008 22:41:44(UTC)
perz

Sweden   
Joined: 12/01/2002(UTC)
Posts: 2,578
Location: Sweden


Ray gave a rather accurate description. However, normally the lok remembers the requested speed, it only forgets that it was already driving at that speed. So usually it does not need to get a new command from the control station to accelerate up again, it just needs to get power back. Ray was a bit unclear by saying "it reads what speed it has been requested". It should in most cases mean "it reads what speed it has been requested, from its internal memory".
Offline intruder  
#5 Posted : 20 October 2008 23:39:17(UTC)
intruder

Norway   
Joined: 16/08/2006(UTC)
Posts: 5,382
Location: Akershus, Norway
Thanks for the explanation, folks.

I think I have a couple of them:
One loco with the decoder from a 60760 set behaves exactly as explained.
First I thought it was poor electrical connection, as this is an analog locomotive that has rested for many years before I converted it. I removed the axels and cleaned very carefully - no change. Also the pick-us shoe and it's connections are checked.
After running as you describe for a while it gets warm and works quite well.

Anotherone, converted with an even simpler decoder than the one from 60760 (marked 102164 on a label), works the same way, but in addition it sometimes stops and is completely dead.
When switching the track current off and on (stop and go on the CS), the locomotive continue as before, but in reverse direction. When reducing the speed to zero and the locomotive comes to a complete stop, it automatically swithes to the correct driving direction and runs as before.

I think another decoder is the answer to both these locomotives.

Case 1:
UserPostedImage

Case 2:
UserPostedImage
Best regards Svein, Norway
grumpy old sod
Offline RayF  
#6 Posted : 21 October 2008 00:06:13(UTC)
RayF

Gibraltar   
Joined: 14/03/2005(UTC)
Posts: 15,840
Location: Gibraltar, Europe
Per, thanks for the clarification. The effect you see is the same in either case.

The main point is that it makes a sudden attempt to stop, before accelerating slowly up to the original speed again. This is not very desirable, as it makes the interruption more pronounced and noticable.

Ray
Ray
Mostly Marklin.Selection of different eras and European railways
Small C track layout, control by MS2, 100+ trains but run 4-5 at a time.
Offline RayF  
#7 Posted : 21 October 2008 00:18:04(UTC)
RayF

Gibraltar   
Joined: 14/03/2005(UTC)
Posts: 15,840
Location: Gibraltar, Europe
Svein,

I have a similar problem with my 3088 DHG700, which has a Lokpilot V2 decoder. Before fitting this I used it with a Delta decoder and had perfect running, apart from the poor low speed performance typical of Delta. As soon as I fitted the lokpilot It would stop on the slightest bit of dirt on the track. Even after careful cleaning of all the contact surfaces, it occasionally still stutters and hesitates, where other locos simply drive cleanly through.

I think that some decoders are much more sensitive to poor electrical contact, and it's made worse by locos with few axles for ground connection.

From experience, the best in this respect are 6080, Delta and 6090x.

The worst are Lokpilot V2, ANDI, (and 60760 to a lesser extent).

Lokpilot V3 are not too bad.

Ray
Ray
Mostly Marklin.Selection of different eras and European railways
Small C track layout, control by MS2, 100+ trains but run 4-5 at a time.
Offline intruder  
#8 Posted : 21 October 2008 00:28:46(UTC)
intruder

Norway   
Joined: 16/08/2006(UTC)
Posts: 5,382
Location: Akershus, Norway
Thanks, Ray.

I have other locomotives with the 60760 decoder that runs perfectly.

The advantage (if running OK) with this kind of decoder is that the maximum speed adjustment has a very big range.
I don't care so much about model accuracy, but the max speed as close to prototypical as possible is important to me. That is why I try to use the remotely adjustable decoders in the slow running locomotives.

I will test another 60760 decoder in the orange locomotive.
Best regards Svein, Norway
grumpy old sod
Offline WelshMatt  
#9 Posted : 21 October 2008 01:39:09(UTC)
WelshMatt


Joined: 06/10/2006(UTC)
Posts: 1,345
Location: ,
Interesting that other people have had this happen with ESU decoders - the standard LGB decoders (which my lone ESU-equipped G loco shares the rails with) are made by Massoth and have circuitry built in to handle this problem.

I have had an identical problem with cheap DCC decoders on dirty track, ones with BEMF don't seem to be as prone.
Matt from Wales.

When you pay Range Rover prices, don't accept Lada quality
Offline perz  
#10 Posted : 21 October 2008 02:07:22(UTC)
perz

Sweden   
Joined: 12/01/2002(UTC)
Posts: 2,578
Location: Sweden
In the ESU decoders and in at least some Hobby loks, (and I think also in the 60760 but I'm not sure) there is CV 73 which selects whether the lok should remember its current actual speed or not. CV 73 set to 7 gives the "Alzheimer" problem, CV 73 set to 3 causes the lok to start abruptly after a power on, which can also be a problem. You can select which problem you dislike the least. Newer Märklin loks have either the CV 73 set to 3 from the factory or have the problem solved completely.

The "Alzheimer" is a secondary effect of a power drop. It is probably caused by a reset to the processor, followed by a non-optimal start-up behaviour. The Alzheimer may in some cases be cured by a capacitor that prevents the processor to go into reset.

There is of course also the primary effect of the power drop, where the motor stops because it gets no current. DC motors tend to stop more abruptly than the AC motors. That is one reason why you won't see this problem on the Delta and 6080 loks. The reason why DC motors are worse in this respect is that the magnet field is not shut off with the power as it is in the AC motors. There are two ways to compensate for this. Either you add a flywheel, or you add a capacitor. Unfortunately the capacitor needs to be rather big to make any real difference.
Offline jeehring  
#11 Posted : 21 October 2008 18:16:35(UTC)
jeehring


Joined: 25/09/2003(UTC)
Posts: 2,786
Location: ,
Quote:
[size=1" face="Verdana" id="quote]quote:Originally posted by WelshMatt
<br />That sounds more like the problems most decoders have with very dirty track. I have a G scale loco with an ESU DCC decoder that did this until I added a capacitor (as advised in the instructions) which smooths the power delivery and keeps it moving over brief power interruptions.


With DCC , Alzheimer behaviour seemed to be the rule for a long time.
Before 2004/2005 all DCC decoders I've seen , were showing Alzheimer behaviour .And my "DCC friends" were considering it as a normal behaviour due to digital system .They were surprised about Marklin decoders . Only Marklin motorola decoders (609xx) didn't show this uncomfortable behaviour .
Today I'm not 100% sure , it seems ESU has brought a solution in DCC too , using a software solution .About recent decoders from other brand , I don't know .....
Offline rschaffr  
#12 Posted : 21 October 2008 18:39:43(UTC)
rschaffr

United States   
Joined: 03/01/2003(UTC)
Posts: 5,180
Location: Southern New Jersey, USA
This problem is even worse with loksounds. I have a loksound in my ICE and when it gets the interruption, before it starts accelerating it has to go through all the start-up sounds again, which takes a second or two (seems like an eternity). Then it accelerates to speed.
-Ron
Digital, Epoch IV-V(K-track/CS3/6021Connect/60216051), Epoch III(C-track/6021/6036/6051)
http://www.sem-co.com/~rschaffr/trains/trains.html
Offline laalves  
#13 Posted : 21 October 2008 19:54:00(UTC)
laalves


Joined: 10/02/2004(UTC)
Posts: 2,162
Location: Portugal
Regarding ESU decoders, that and other problems may be solved/mitigated (depending on the model of the decoder) with a firmware update available from ESU's webpage. Alas, one will need a Lokprogrammer to upload the updates.
Offline rschaffr  
#14 Posted : 22 October 2008 02:21:11(UTC)
rschaffr

United States   
Joined: 03/01/2003(UTC)
Posts: 5,180
Location: Southern New Jersey, USA
Well, I have one (lokprogrammer). I must look into that

Thanks, Luis
-Ron
Digital, Epoch IV-V(K-track/CS3/6021Connect/60216051), Epoch III(C-track/6021/6036/6051)
http://www.sem-co.com/~rschaffr/trains/trains.html
Offline intruder  
#15 Posted : 23 May 2009 19:40:49(UTC)
intruder

Norway   
Joined: 16/08/2006(UTC)
Posts: 5,382
Location: Akershus, Norway
A short report about my two Alzheimer locomotives 2847 and 3044:

When I bought and converted these locomotives, they had both been standing still for quite a long time, probably several years.

Some time late October last year I oiled the axel shafts very well, thinking that some deposites may be stuck or hardened in the axel bearings. Today I tested them again, same problem as before, as described by Ray in the second post of this topic.

When running stationary in my test bench they behaved the same way.
But, after a very simple fix today, they both run very nice on my oval test track.

1. Removed the axels, cleaned them for all the oil and assembeled them back in. Probobly has all the fresh oil from last year desolved the old stuff in the axel bushing.
2. Both locomotives have only one friction tyre on the rear axel, to improve the ground connection (the friction tyre on the non-cogged side is moved to the center axel, which has less ground contact anyhow)

So now they run lap after lap in they prototypical speed of 32 and 50 km/h. I am happy.
Best regards Svein, Norway
grumpy old sod
Offline Drongo  
#16 Posted : 25 May 2009 14:28:36(UTC)
Drongo

Australia   
Joined: 03/06/2008(UTC)
Posts: 1,226
Location: Sydney, NSW
I'm confused about this Alzheimer problem - more directly, how does one "cure" the problem? I have 2 locos with the problem and I want to know how to exactly cure them. Can someone please explain the procedure?
Take it easy . . . . or any other way you can get it !!!!
Offline RayF  
#17 Posted : 25 May 2009 14:41:53(UTC)
RayF

Gibraltar   
Joined: 14/03/2005(UTC)
Posts: 15,840
Location: Gibraltar, Europe
Hi Greg.

Alzheimers, as in the real illness (unfortunately), has no cure. It is a behaviour shown by some decoders and not by others. If your decoder shows the problem, you can mitigate it by having cleaner track and well maintained locos, so that the power supply to the loco is not interrupted. This is what Svein achieved by cleaning the axles on his locos from old oil.

You can also improve the behaviour by fitting a large capacitor across the decoder so that it "holds up" the voltage when you get a momentary interruption to the supply. Perz mentions this in his post above.

The latest Lokpilot V3 does not seem to show the effect, but the older V2 has it very bad! According to Luis, ESU have a firmware upgrade that can be uploaded which helps.

I hope this summarises your options in a way that makes sense.
Ray
Mostly Marklin.Selection of different eras and European railways
Small C track layout, control by MS2, 100+ trains but run 4-5 at a time.
Offline Bigdaddynz  
#18 Posted : 25 May 2009 14:45:59(UTC)
Bigdaddynz

New Zealand   
Joined: 17/09/2006(UTC)
Posts: 18,669
Location: New Zealand
Like the real life Alzheimer's, there is no real cure.

A summary of the suggested fixes in this thread are:

- Replace the decoder with a newer one that is less susceptible (Lokpilot 3, 6091x)
- Set CV73 to 3 and see if that helps
- Apply firmware updates to older decoders - Lokpilot 2. Requires an ESU Lokprogrammer
- Investigate whether a capacitor can be fitted to the decoder to prevent momentary power loss.
- Make sure the loco's mechanicals and electrical connections are clean. Make sure the loco has good grounding
- Keep your track clean and fix any dead spots.
Offline river6109  
#19 Posted : 25 May 2009 15:51:26(UTC)
river6109

Australia   
Joined: 22/01/2009(UTC)
Posts: 14,740
Location: On 1965 Märklin Boulevard just around from Roco Square
Quote:
[size=1" face="Verdana" id="quote]quote:Originally posted by RayPayas
<br />Svein,

I have a similar problem with my 3088 DHG700, which has a Lokpilot V2 decoder. Before fitting this I used it with a Delta decoder and had perfect running, apart from the poor low speed performance typical of Delta. As soon as I fitted the lokpilot It would stop on the slightest bit of dirt on the track. Even after careful cleaning of all the contact surfaces, it occasionally still stutters and hesitates, where other locos simply drive cleanly through.

I think that some decoders are much more sensitive to poor electrical contact, and it's made worse by locos with few axles for ground connection.

From experience, the best in this respect are 6080, Delta and 6090x.

The worst are Lokpilot V2, ANDI, (and 60760 to a lesser extent).

Lokpilot V3 are not too bad.

Ray


most the time it relates to track contact.
Locos as you've mentioned with 3 axles sometimes loose the contact, although this should not happen on a track without including turnouts.
With original metal couplings (loco)& metal couplings on carriages (coupled to loco) does eliminate the problem of loosing power temporarly.
With my tracks, it is frustrating having locos which suffer from this illness, stopping and starting.
I also noticed Bo-Bo locos, e.g. BR 110, BR 10 (3039) etc. not having proper contact between front bogy and frame. later models include a springwasher.

John


https://www.youtube.com/river6109
https://www.youtube.com/6109river
5 years in Destruction mode
50 years in Repairing mode
Offline Drongo  
#20 Posted : 27 May 2009 06:26:06(UTC)
Drongo

Australia   
Joined: 03/06/2008(UTC)
Posts: 1,226
Location: Sydney, NSW
I'd like to add my results to this topic. A friend suggested to me that I setup a small section of track with a MS. I had 2 locos with the problem. I managed to correct one loco and not the other, so I guess there is some chance of rectifing the problem without changing the decoder, etc. I hope this helps someone else.
Take it easy . . . . or any other way you can get it !!!!
Offline RayF  
#21 Posted : 27 May 2009 10:39:42(UTC)
RayF

Gibraltar   
Joined: 14/03/2005(UTC)
Posts: 15,840
Location: Gibraltar, Europe
Quote:
[size=1" face="Verdana" id="quote]quote:Originally posted by youngagain
<br />I'd like to add my results to this topic. A friend suggested to me that I setup a small section of track with a MS. I had 2 locos with the problem. I managed to correct one loco and not the other, so I guess there is some chance of rectifing the problem without changing the decoder, etc. I hope this helps someone else.


What did you do?
Ray
Mostly Marklin.Selection of different eras and European railways
Small C track layout, control by MS2, 100+ trains but run 4-5 at a time.
Offline Drongo  
#22 Posted : 27 May 2009 14:35:17(UTC)
Drongo

Australia   
Joined: 03/06/2008(UTC)
Posts: 1,226
Location: Sydney, NSW
What did you do?


I placed the loco on the small loop I setup and manually typed in the loco's number - this is all done with the "stop" set on the MS. Then I pressed the "OK" and the loco took off as it has done in the past with Alzheimer's, however, after a few seconds the MS then can control the loco like normal. As I said it worked for one loco and not the other, so it's not a guaranteed cure. I hope this helps.
Take it easy . . . . or any other way you can get it !!!!
Offline RayF  
#23 Posted : 27 May 2009 15:05:09(UTC)
RayF

Gibraltar   
Joined: 14/03/2005(UTC)
Posts: 15,840
Location: Gibraltar, Europe
Thanks, Greg.

This is not so much a cure, but a way of regaining control when the loco remembers that last time it was on the track it was at a speed other than zero. This sometimes happens when you end a running session with the "stop button" instead of stopping each individual loco.

None of my "alzheimers" locos take off like that when I put them on the track, but they do exhibit the behaviour explained above, where a momentary power drop makes them stop suddenly and then accelerate slowly back up to the desired speed.

Ray
Ray
Mostly Marklin.Selection of different eras and European railways
Small C track layout, control by MS2, 100+ trains but run 4-5 at a time.
Users browsing this topic
Guest
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 1.039 seconds.