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 BenP  
#1 Posted : 13 January 2023 23:25:33(UTC)
BenP

United States   
Joined: 04/02/2021(UTC)
Posts: 256
Location: Michigan, Ann Arbor
Is there a way to test the power output of old k83 decoder ports (4/decoder). I use about a dozen k83s and find that all are able to drive my M-track switches, but that some vintage signals do not get enough boost to switch. This is an annoyance, but also dangerous when the current keeps going through the coils as the metal core has not moved all the way. The coil quickly gets very hot. Many signals work one way (working red or green or arm moves, but the reverse does not go all the way).
Decoders get current from CS3+ using 18 gauge wires. Wires from decoder to nearby accessories are original (24/26 gauge?). The layout uses 3-4amp when running and never trips. Issue happens whether trains run or not. Scenery is on a separate transformer.
I wonder if a multimeter can be used to check functionality, but the switching pulse is 500milliseconds or less. Any suggestions?
Ben
Digital M track layout with mostly vintage rolling stock and accessories, and small Z scale layout.
https://youtube.com/play...0kgVYbh0CeDTF-bYXoD_2-V9
thanks 1 user liked this useful post by BenP
Offline marklinist5999  
#2 Posted : 14 January 2023 01:04:49(UTC)
marklinist5999

United States   
Joined: 10/02/2021(UTC)
Posts: 2,292
Location: Michigan, Troy
Ben, hot wires or solenoids isn't good. Are the K83's set for momentary contact, or steady? Can they be changed from one to the other? 18 gague wire is fine, even 24. 4 amp. is fine. I use my M83 for analog hobby semiphore signal, no problem.
Offline JohnjeanB  
#3 Posted : 14 January 2023 01:21:38(UTC)
JohnjeanB

France   
Joined: 04/02/2011(UTC)
Posts: 2,583
Location: Paris, France
Hi Ben
It may be the issue is not with the 6083 but with the signals themselves especially with the 7036, 7037, 7038, 7039, 7040, 7041, 7042 and 7043.
Another possibility is the digital signal feeding them is weak (long thin wires, etc).

When subjected to heat (train staying over a contact section, etc) then the nylon making the coil core becomes soft and the internal rectangular conduit changes shape and slows down the iron core sliding in a nickel-plated gutter. In some cases, the solenoids have to be replaced. Same issues with M Track solenoids (same design approach).

Another source of issues: bended rods on semaphores and foresignals.
Mine are operated by Märklin decoders and even some old 6083 are in use and working very fine (even my 7043 humping signal from 1953)
Here is a first version of the 6083 where the transistors are easy to find and to change.
6083.png

Now I prefer using 60832 (better time protection for solenoids.
Cheers
Jean
thanks 1 user liked this useful post by JohnjeanB
Offline BenP  
#4 Posted : 14 January 2023 02:43:20(UTC)
BenP

United States   
Joined: 04/02/2021(UTC)
Posts: 256
Location: Michigan, Ann Arbor
Originally Posted by: marklinist5999 Go to Quoted Post
Ben, hot wires or solenoids isn't good. Are the K83's set for momentary contact, or steady? Can they be changed from one to the other? 18 gague wire is fine, even 24. 4 amp. is fine. I use my M83 for analog hobby semiphore signal, no problem.


K83s for momentary pulse. I believe that overheating happens when solenoid only partially moves coil and leaves mechanical contacts live. Problem happens on a few units only. I ordered an m83 to see if that makes a difference with vintage signals. Replacing signals did not make a difference.
Digital M track layout with mostly vintage rolling stock and accessories, and small Z scale layout.
https://youtube.com/play...0kgVYbh0CeDTF-bYXoD_2-V9
thanks 1 user liked this useful post by BenP
Offline BenP  
#5 Posted : 14 January 2023 02:49:50(UTC)
BenP

United States   
Joined: 04/02/2021(UTC)
Posts: 256
Location: Michigan, Ann Arbor
Originally Posted by: JohnjeanB Go to Quoted Post
Hi Ben
It may be the issue is not with the 6083 but with the signals themselves especially with the 7036, 7037, 7038, 7039, 7040, 7041, 7042 and 7043.
Another possibility is the digital signal feeding them is weak (long thin wires, etc).

When subjected to heat (train staying over a contact section, etc) then the nylon making the coil core becomes soft and the internal rectangular conduit changes shape and slows down the iron core sliding in a nickel-plated gutter. In some cases, the solenoids have to be replaced. Same issues with M Track solenoids (same design approach).

Another source of issues: bended rods on semaphores and foresignals.
Mine are operated by Märklin decoders and even some old 6083 are in use and working very fine (even my 7043 humping signal from 1953)
Here is a first version of the 6083 where the transistors are easy to find and to change.
6083.png

Now I prefer using 60832 (better time protection for solenoids.
Cheers
Jean


Seems that decoders are powering vintage signals just enough, or just not. Perhaps old decoders (capacitors?) just wear out. Particularly red/green signal 7188 and 2 arm semaphores give problems; switches work fine, incl 1960s stock. I ordered M83 to continue my testing.

Digital M track layout with mostly vintage rolling stock and accessories, and small Z scale layout.
https://youtube.com/play...0kgVYbh0CeDTF-bYXoD_2-V9
Offline PeFu  
#6 Posted : 14 January 2023 09:26:29(UTC)
PeFu

Sweden   
Joined: 30/08/2002(UTC)
Posts: 1,075
Reading this post, I remember that I made a very simple capacitor solution for a jerking M track turnout in a k83 context, on my previous layout. I failed finding a link, but found another fellow forum member having the reference:

https://www.marklin-user...ength-duration#post62916

This made the trick for me back then. ThumpUp

I also know there are forum members that successfully have used two k83:s in parallel, both having the same digital address.
Inspired by Swiss railways SBB and BLS | C and K track | CS2 | TrainController Gold V10
Andreasburg-Mattiasberg Bahn on the Marklin-Users Forum and on Youtube
thanks 2 users liked this useful post by PeFu
Offline BenP  
#7 Posted : 30 January 2023 02:06:06(UTC)
BenP

United States   
Joined: 04/02/2021(UTC)
Posts: 256
Location: Michigan, Ann Arbor
Intermittent decoder progress report.

I added a power pack with 470 uF capacitor (+ resistor and diode to avoid cs3 startup surge) for signals (lights and semaphores), but that did not make a difference. Sometimes they work, often not. Gr/Red signal 7188 is most troubled. Track turnouts on the same decoders work all of the time, so no electronics fault.

I bought an M83, which did not appreciably improve performance. I did not include a switching power supply.

When I use decoders (old and new) on my short test track, they work fine. So, it seems a large layout (pulse?) power issue. My layout uses 3.5+ amps, based on cs3 reporting. Red/brown power supply wires are 18 gauge. Together, that should be enough juice from 5amp system, but perhaps not for start pulse of signals' solenoid.

Another test would be adding a 60175 mfx booster for switches+signals only, but that is an expensive test ($240+). I looked for a used 60174, but they are rare.

Ideas?
Ben

Edited by user 30 January 2023 05:48:18(UTC)  | Reason: Not specified

Digital M track layout with mostly vintage rolling stock and accessories, and small Z scale layout.
https://youtube.com/play...0kgVYbh0CeDTF-bYXoD_2-V9
thanks 1 user liked this useful post by BenP
Offline marklinist5999  
#8 Posted : 30 January 2023 14:43:19(UTC)
marklinist5999

United States   
Joined: 10/02/2021(UTC)
Posts: 2,292
Location: Michigan, Troy
Look online for a digital booster innards schematic, and make your own? It can't be super comples, because an old Insider issue I have shows how to use the Delta control as a booster with a second transformer. Of course you'll want it on a switched mode power supply, and the tricky part will be getting the correct cables, and or adapters to connect it up.
Amazon has electronics project boxes of various sizes, and all sorts of parts from I/C boards, transistors, resistors, capacitors, rectifiers, etc. If you have a high quality digital soldering set, clamps, etc. you can do it.
Offline BenP  
#9 Posted : 30 January 2023 15:32:51(UTC)
BenP

United States   
Joined: 04/02/2021(UTC)
Posts: 256
Location: Michigan, Ann Arbor
Originally Posted by: marklinist5999 Go to Quoted Post
Look online for a digital booster innards schematic, and make your own? It can't be super comples, because an old Insider issue I have shows how to use the Delta control as a booster with a second transformer. Of course you'll want it on a switched mode power supply, and the tricky part will be getting the correct cables, and or adapters to connect it up.
Amazon has electronics project boxes of various sizes, and all sorts of parts from I/C boards, transistors, resistors, capacitors, rectifiers, etc. If you have a high quality digital soldering set, clamps, etc. you can do it.



I appreciate the confidence you express in my electronics skills :-) I will first open the innards of signals and try a deep clean, which better fits my basic skills. Will report back.
Digital M track layout with mostly vintage rolling stock and accessories, and small Z scale layout.
https://youtube.com/play...0kgVYbh0CeDTF-bYXoD_2-V9
thanks 1 user liked this useful post by BenP
Offline PMPeter  
#10 Posted : 30 January 2023 17:01:42(UTC)
PMPeter

Canada   
Joined: 04/04/2013(UTC)
Posts: 1,258
Location: Port Moody, BC
I would agree with others suggesting that the problem is with the signals, not the K83 output. I have a K83 set up for testing items before I sell them. From the same output I can usually see the problem quickly. I can successfully switch a 7188 in both directions and the next one will only switch one way or partially in either direction. Increasing the activation time usually makes little difference. So I know the signal has a problem that needs to be corrected, or it is simply worn out.
thanks 2 users liked this useful post by PMPeter
Offline BenP  
#11 Posted : 30 January 2023 18:14:01(UTC)
BenP

United States   
Joined: 04/02/2021(UTC)
Posts: 256
Location: Michigan, Ann Arbor
Originally Posted by: PMPeter Go to Quoted Post
I would agree with others suggesting that the problem is with the signals, not the K83 output. I have a K83 set up for testing items before I sell them. From the same output I can usually see the problem quickly. I can successfully switch a 7188 in both directions and the next one will only switch one way or partially in either direction. Increasing the activation time usually makes little difference. So I know the signal has a problem that needs to be corrected, or it is simply worn out.


Thanks PMPeter. I agree that signals themselves are key to the problem, but I find that the same light/semaphore signal on my large layout will switch partially (one way), but working flawlessly on my test setup (using exact same decoder used for both experiments). Using old k83 or modern m83 makes no difference; changing pulse time (at 300-500 millisecs) has no obvious effect. Somehow, there is power loss in the system that hovers around the tipping point for the coil mechanics.
Long wires? The k83s are 1-3m from cs3 power, but working and failing signals are not systematic with length of 18-guage wiring.
Power use from other devices? Trains are on the track, but are not running (coaches are lighted). There are also another 20 signals and 20 switches connected, drawing (diluting?) power. Total power use is 3.5-4 amps.
Mechanical resistance/wear within the signal? This is likely a key issue as some signals work fine and track turnouts have no problem. I'll do some deep cleaning today and test again.

AMBen
Digital M track layout with mostly vintage rolling stock and accessories, and small Z scale layout.
https://youtube.com/play...0kgVYbh0CeDTF-bYXoD_2-V9
Offline PMPeter  
#12 Posted : 30 January 2023 20:48:46(UTC)
PMPeter

Canada   
Joined: 04/04/2013(UTC)
Posts: 1,258
Location: Port Moody, BC
Originally Posted by: BenP Go to Quoted Post
I agree that signals themselves are key to the problem, but I find that the same light/semaphore signal on my large layout will switch partially (one way), but working flawlessly on my test setup (using exact same decoder used for both experiments).



I believe you possibly answered your own question. If it works on the same K83 output on the test bench, but not when it is mounted on the layout it most likely is one of two things. Possibly the physical mounting is twisting or binding the signal solenoids so that they don't function properly. I have certainly had that with the turnout mechanisms where when I screw them down to the layout top they sometimes didn't work whereas when I left them loose they worked just fine. The other thing could be that your electrical connections when it is mounted on the layout could be loose or not making contact with all the strands of the connecting wire. If you use the entire length of wires on your test setup and it works, then this is not the problem.

thanks 3 users liked this useful post by PMPeter
Offline Crazy Harry  
#13 Posted : 30 January 2023 23:31:54(UTC)
Crazy Harry

Canada   
Joined: 18/11/2008(UTC)
Posts: 460
Location: Oakville, Ontario
Originally Posted by: PMPeter Go to Quoted Post
Possibly the physical mounting is twisting or binding the signal solenoids so that they don't function properly.


I think this is the most probable cause of your issues with the signals. Have you tested the signal on the layout without physically screwing it down?

Cheers,

Harold.

thanks 1 user liked this useful post by Crazy Harry
Offline BenP  
#14 Posted : 31 January 2023 05:29:52(UTC)
BenP

United States   
Joined: 04/02/2021(UTC)
Posts: 256
Location: Michigan, Ann Arbor
I took signal apart for deep cleaning and discovered that plastic outer casing creates a little extra friction. It might be a wire pressing on the delicate copper connectors or on tumbler handle. Without the case pushed all the way down, the signal usually works.
This still does not explain the unit working fine on my test track, suggesting to me that power is still a limiting issue.

Digital M track layout with mostly vintage rolling stock and accessories, and small Z scale layout.
https://youtube.com/play...0kgVYbh0CeDTF-bYXoD_2-V9
thanks 2 users liked this useful post by BenP
Offline PeFu  
#15 Posted : 31 January 2023 07:37:27(UTC)
PeFu

Sweden   
Joined: 30/08/2002(UTC)
Posts: 1,075
I’m using the S-DEC-4 turnout decoders from LIttfinski (LDT). You can add a separate power feed to the S-DEC-4, which you can’t on the old Märklin K83 decoder. Here is the solution, when a C track double-slip was struggling:

https://www.marklin-user...blem---Solved#post540799

I don’t have a Märklin M83, but know they can have a separate power feed. However, Märklin recommends their own power source having a specific plug on the M83, which makes it more tricky to test different power sources? (like old laptop chargers etc.)
Inspired by Swiss railways SBB and BLS | C and K track | CS2 | TrainController Gold V10
Andreasburg-Mattiasberg Bahn on the Marklin-Users Forum and on Youtube
thanks 1 user liked this useful post by PeFu
Offline BenP  
#16 Posted : 16 April 2023 02:08:18(UTC)
BenP

United States   
Joined: 04/02/2021(UTC)
Posts: 256
Location: Michigan, Ann Arbor
I gave up and accepted 7188 defeat. Got 8 of 'm.
Tried one k83 for one signal, added capacitor, added booster, tweaked switching times, thicker wires, etc. Units work fine on analog and digital test benches, which has short track and short wires. My hunch is that digital AC power from k83 is just enough, or often not, to fully move solenoid of light signal (or not). Semaphore doesn't show the issue, using same k83 at same location.
I now installed basic LED red/green signals, using digital AC to drive controlling k84 decoder, and 12V DC to power LEDs (not using digital AC, which is 16+V and creates flicker). The nostalgic 7188 lights have been retired.

Screenshot_20230415_200631_eBay.jpg
Digital M track layout with mostly vintage rolling stock and accessories, and small Z scale layout.
https://youtube.com/play...0kgVYbh0CeDTF-bYXoD_2-V9
thanks 1 user liked this useful post by BenP
Users browsing this topic
OceanSpiders 2.0
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 © 2003-2023, Yet Another Forum.NET
This page was generated in 0.447 seconds.