Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

false detection problem

807 views
16 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    July 2009
  • From: somerset, nj
  • 2,823 posts
false detection problem
Posted by gregc on Wednesday, January 8, 2020 5:51 AM

having a problem with detectors constantly active (indicating occupancy) when no train is in the block.

we're using the NMRA type of current detection circuit which is in series with one of the rails.  A sufficient current creates a voltage (2 diode drops) across a bridge rectifier which is sufficent to drive a current thru an opto, that drains the charge on a capacitor.   The 555 is used as a Schmitt trigger that is active when the capacitor voltage is < 1/3 Vcc and deactivates when the capacitor is allowed to recharge, > 2/3 Vcc, when the block is no longer occupied and no longer drawing current.

we're wiring up staging loops that will eventually be hidden, covered by a 2nd set of loops.   the loops are divided into 2 sets, one w/ 2 tracks and the other w/ 3 tracks.   each loop is intended to store 2 trains and the storage portion has 4 blocks, a long and a short block that can be switched off to automatically stop the train.

the problem blocks are the 2 blocks with 2 turnouts each (4,5 and 2,3) connecting the violet, orange and blue tracks.

  1. ignoring the detector problem, we completed wiring the tracks and were able to run locos thru all from main to main
  2. we tried a different detection circuit verified to work.
  3. we jumpered the bus to a short piece of flex track and the detector worked as expected when we put a loco on it.
  4. with the long (7') bus wire connecting the detector to drops from the rails disconnected from the rails, the detector would become active with me touching the exposed wire at the end.
  5. when falsely active, i measured ~0V at pins 2 and 6 of the 555, the capacitor.
  6. unlike the other 2 track blocks that are similar, the turnouts in these block have insulated frogs.
  7. haven't fully tested detection on all the storage track detectors.  may have see a similar problem on one
  8. all the blocks in the diagram except the mainline blocks (lower right) are powered using a DCC Specialties Frog/AR, used to power separate reverse loops for each group of tracks
  9. the auto-reverser is connected to the end of a ~60' DCC bus feeding many other tracks along the way

has anyone had a similar problem?

any thoughts?

greg - Philadelphia & Reading / Reading

  • Member since
    February 2002
  • From: Reading, PA
  • 27,301 posts
Posted by rrinker on Wednesday, January 8, 2020 7:33 AM

 Key is the point where you said it falsely detects with the track feeders disconnected by touching one of the bus wires. You're getting either an inductive load (are the bus wires twisted?) or a capacitive load.

                                           --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    July 2009
  • From: somerset, nj
  • 2,823 posts
Posted by gregc on Wednesday, January 8, 2020 7:45 AM

no

why just some blocks?

greg - Philadelphia & Reading / Reading

  • Member since
    February 2002
  • From: Reading, PA
  • 27,301 posts
Posted by rrinker on Wednesday, January 8, 2020 8:18 AM

 How are they run relative to the others which work fine? Near any other wiring? 

"Tried a different detection circuit known to work" - so you swapped the detector on this line with another one, with the same false detection? 

Maybe it's being overly sensitive due to something with the bus/feeder wiring. You can reduce that a little with a .01uF cap across the diode bridge (between the DCC input for the detected rail and the track output for the detected rail). I've seen that in other schematics as an option to reduce noise sensitivity. You can go as high as .1uF if .01 doesn't work.

                                          --Randy


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    January 2009
  • 4,492 posts
Posted by RR_Mel on Wednesday, January 8, 2020 8:55 AM

Greg
 
Before I gave up on current detection I used a Rob Paisley circuit using a H11AA4QT optoisolator with a LM339 chip and they never falsed.
 
When I was testing the circuit I found out that the H11AA4QT circuit board was sensitive to stray inductive fields so I used ¼” ferrite beads on the leads to the board as a preventive thing.  At the time I had SCR lighting dimmers in the house and they got into everything.
 
The reason I went with optical IR detection was that I didn’t want to have to put resistors on the axles of all my rolling stock.  I had good luck with the Paisley detection circuit but I wanted to be able to detect anything in my hidden blocks.
 
 
Mel
 
 
 
My Model Railroad   
 
Bakersfield, California
 
I'm beginning to realize that aging is not for wimps.
 
  • Member since
    July 2009
  • From: somerset, nj
  • 2,823 posts
Posted by gregc on Wednesday, January 8, 2020 10:05 AM

i have toroids.   I could try putting one on the bus wire near the detector 

greg - Philadelphia & Reading / Reading

  • Member since
    July 2009
  • From: somerset, nj
  • 2,823 posts
Posted by gregc on Friday, January 10, 2020 3:45 PM

we found that using 2 spare detectors on a different board solved the immediate problem.

not sure what the issue is with that particular board

greg - Philadelphia & Reading / Reading

  • Member since
    February 2002
  • From: Reading, PA
  • 27,301 posts
Posted by rrinker on Friday, January 10, 2020 5:31 PM

 Trying to figure out what would make it more sensitive. The goofy one - check the 10uF cap, make sure it's not damaged and/or leaking. And is the pullup the 120K shown on the schematic, and not a 12K, or 1.2M?

                                   --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    July 2009
  • From: somerset, nj
  • 2,823 posts
Posted by gregc on Saturday, January 11, 2020 5:09 AM

i will check, but while capacitor and resistor size could possibly make the detector less sensitive, i don't see how their values can make a detector more sensitive.

the detector requires a voltage across the bridge and sufficient current to turn on the opto transistor repeatedly to keep the capacitor drained.   what is the source if there's nothing across the rails?

greg - Philadelphia & Reading / Reading

  • Member since
    February 2002
  • From: Reading, PA
  • 27,301 posts
Posted by rrinker on Saturday, January 11, 2020 12:54 PM

 Bad bridge rectifier, or conencted to the wrong side? Or maybe the 33 ohm resistor in the opto. Or just a failing opto, or out of spec and turning on the transistor with much less brightness. Or maybe it's within spec and just near one side instead of being more to the center.

                                            --Randy


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    July 2009
  • From: somerset, nj
  • 2,823 posts
Posted by gregc on Saturday, January 11, 2020 1:51 PM

it could be components just on that board.   three circuits seem to have the same problem while 6 others on different boards do not.  will know when other circuits are wired

greg - Philadelphia & Reading / Reading

  • Member since
    July 2009
  • From: somerset, nj
  • 2,823 posts
Posted by gregc on Friday, January 17, 2020 4:04 PM

we completed more of the detector wiring.   all of the longer blocks were active w/o occupancy.

the circuit we're using does not have the 0.1uF cap across the output of the bridge that is shown in the NMRA circuit (see full featured)

i have my guess.  wondering what other may think

greg - Philadelphia & Reading / Reading

  • Member since
    February 2002
  • From: Reading, PA
  • 27,301 posts
Posted by rrinker on Friday, January 17, 2020 5:22 PM

 Probably keeps noise and spikes from false triggering with that capacitor.

                             --Randy


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    July 2009
  • From: somerset, nj
  • 2,823 posts
Posted by gregc on Saturday, January 25, 2020 3:17 AM

adding a 0.1 uF cap across the bridge fixed 8  of 10 blocks.  tried 2 different type caps (ceramic disc, polyester) thinking first damaged by heat.

measured ~0.7VAC across bridge of one detector falsely active, as low as ~0.012VAC across others not having any problems.

still unclear what signal (noise) is the cause of the problem on these larger blocks.   must be higher in frequency than ~10kHz DCC for cap to supress.

we plan to look again at resistance across rails and wiring, as well as trying different detection circuits.

greg - Philadelphia & Reading / Reading

  • Member since
    February 2002
  • From: Reading, PA
  • 27,301 posts
Posted by rrinker on Saturday, January 25, 2020 11:56 AM

 It could be that the bridge rectifiers are not fast enough to deal with DCC frequency. Instead of a canned bridge, you might have to use a set of fast switching diodes and make your own, although there are fast enough bridges out there.

                              --Randy


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    July 2009
  • From: somerset, nj
  • 2,823 posts
Posted by gregc on Saturday, January 25, 2020 2:14 PM

why on these 2 blocks and not the other 26?   baffled

greg - Philadelphia & Reading / Reading

  • Member since
    February 2002
  • From: Reading, PA
  • 27,301 posts
Posted by rrinker on Saturday, January 25, 2020 2:58 PM

 Are they the same part number? Same manufacturer?

                           --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!