Intermittent Reads from Sensors

John Russo

John Russo

Registered
Country flag
I am having issues with intermittent track sensor reads.

I am using LGB MTS III with a 50070 feedback module.

I am using Stellworks as the controlling software.

At Juergen's suggestions, I use these reed switches and have been trying all kinds of magnets:

MS-328-3-1-0500 PIC GmbH | Sensors, Transducers | DigiKey

The sensors are making reads 80% of the time.

Are there hard rules with the proximity of the magnet to the sensor?

Does the the sensor have to be precisely oriented?
 
How are you mounting the sensors and what orientation?
What magnets are you using? Again how and where are you mounting them?
 
The sensors are most sensitive on the 'front' face. - The face pointing vertically upwards in your photo.

The LGB magnets are not particularly strong.. you may be better with Neodinium (sp) 'super-magnets'?

PhilP.
 
The reeds look very low in the track, they need to be almost level with the rail top so the magnetic just clears them.
 
PhilP said:
The sensors are most sensitive on the 'front' face. - The face pointing vertically upwards in your photo.

The LGB magnets are not particularly strong.. you may be better with Neodinium (sp) 'super-magnets'?

PhilP.
Click to expand...

Phil,

When you say "The face pointing vertically upwards" do you mean A or B ?

If it is B, i.e. the thin edge, do you thing if I mounted it so that was facing upward with the thin edge laying horizontal?
 

Attachments

  • Picture1.png
    Picture1.png
    2.2 MB · Views: 0
Face B pointing up towards the magnets..

Where you have used two magnets, they are the same pole upwards? - You can test by putting a second magnet on top so it sticks, then move it to the other magnet. It should not flip over..

You can add extra magnets (on top of each other) to increase the strength, and get closer to the Reed..

PhilP.
 
I would place track sensors in line with the rails, not in line with the ties. This will give a moving engine more 'trip' time for the reed.
 
Dan said:
I would place track sensors in line with the rails, not in line with the ties. This will give a moving engine more 'trip' time for the reed.
Click to expand...

I like the idea, I tried it and did not have any change in operation. will fiddle with it more.

I also ordered more sensitive reed switches and will give those a try when they arrive.
 
A single magnet is best, 2 separate magnets may need "combining".

Experiment before gluing to the loco would be my suggestion.... you may find the centering the magnet over the reed switch body is NOT the most sensitive alignment. Again you can determine this by experimentation.

Strange you are having this issue,

Also, I thought the 50070 was a throttle, does it have some trigger input? I cannot find a manual on the 50070.

Greg
 
Greg Elmassian said:
A single magnet is best, 2 separate magnets may need "combining".

Experiment before gluing to the loco would be my suggestion.... you may find the centering the magnet over the reed switch body is NOT the most sensitive alignment. Again you can determine this by experimentation.

Strange you are having this issue,

Also, I thought the 50070 was a throttle, does it have some trigger input? I cannot find a manual on the 50070.

Greg
Click to expand...


Greg,

Good call on the centering. When I wrote to PIC, they sent me this link. Interesting.

Interactive Reed Switch

Regarding the 50070 manual: https://www.onlytrains.com/manuals/55070.pdf

I am not sure what yo mean by trigger input.

For me, I have the 50070 with with 8 reed contacts. The 50070 gets connected to the MTS III bus.

I then have an LGB 55060 connected to the bus and also a serial port on a PC.

I use Stellwerks software from Juergen Schwartz.

Once I get consistent reads from the reeds, I will be a happy camper.

John
 
Notice you typed 50070 again, but your link is to 55070..... now I understand what you are doing....

not an expert on this interface, but you may have to "debounce" the reed switches is my guess, especially they show these operated from more steady signals.

Reed switches are famous for bouncing, and debounce circuits are common.

I would ask Jurgen about this, unless you are familiar with "bouncing" and "debounce" circuits, and the concept of giving double sets of pulses.

Greg
 
Greg Elmassian said:
Notice you typed 50070 again, but your link is to 55070..... now I understand what you are doing....

not an expert on this interface, but you may have to "debounce" the reed switches is my guess, especially they show these operated from more steady signals.

Reed switches are famous for bouncing, and debounce circuits are common.

I would ask Jurgen about this, unless you are familiar with "bouncing" and "debounce" circuits, and the concept of giving double sets of pulses.

Greg
Click to expand...

I am familiar with the concept but not sure how to apply it in this situation.

I will reach out to Jürgen and also take another look at his documentation on the software.
 
Mystery solved.

I had a bad Feedback Module - LGB 55070.

I replaced it with a Massoth 817001 - 280R and voila.

One quirk is that I seem to have to unplug and replug the bus cable when I start up. Aside from that I get 100% consistent reads. Any of the magnets work.

Now it it is off to the races with programming routines for the trains.

Only problem is we are selling our house and moving. Everything is going into storage for 8 months and so I won't be unpacking until December. Maybe a break from trains will do me good. :smiley:
 
You must log in or register to reply here.
Back
Top Bottom