Trains.com

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!

QSI decoder losing long address

11958 views
47 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Saturday, April 13, 2013 4:42 PM

tstage

Jack,

Since the QSI decoder is losing its address, BLI wanted me to manually change those CVs to the said (calculated) values to see if the address would "stick" and the locomotive operate on that particular address.  If the address "stuck" and (after removing it from the track for several minutes then putting it back on the track) the locomotive ran fine then that would rule out the decoder as being the problem.

Tom

Tom,

Have you tried it yet?

Rich

Alton Junction

Moderator
  • Member since
    June 2003
  • From: Northeast OH
  • 17,249 posts
Posted by tstage on Saturday, April 13, 2013 1:42 PM

Jack,

Since the QSI decoder is losing its address, BLI wanted me to manually change those CVs to the said (calculated) values to see if the address would "stick" and the locomotive operate on that particular address.  If the address "stuck" and (after removing it from the track for several minutes then putting it back on the track) the locomotive ran fine then that would rule out the decoder as being the problem.

Tom

https://tstage9.wixsite.com/nyc-modeling

Time...It marches on...without ever turning around to see if anyone is even keeping in step.

  • Member since
    December 2001
  • From: Trois-Rivieres Quebec Canada
  • 1,063 posts
Posted by jalajoie on Saturday, April 13, 2013 8:53 AM

Thanks Tom

My question was out of curiosity, however I don't get the purpose of this testing. Do you?

I made my own CV17 CV18 calculator for my iPod and programmed an Excel sheet to reverse engineer the decoder address from values found in CV17 CV18. 

Jack W.

Moderator
  • Member since
    June 2003
  • From: Northeast OH
  • 17,249 posts
Posted by tstage on Friday, April 12, 2013 11:09 PM

Jack,

No, I can post them.  BLI gave me the following CVs to test the locomotive out on address #1234:

  • CV17=196
  • CV18=210
  • CV29=38

By entering those specific values for the above CVs, the locomotive should respond to address #1234.  These can easily be calculated for any 4-digit address using the following link:

http://www.2mm.org.uk/articles/cv29%20calculator.htm

Tom

https://tstage9.wixsite.com/nyc-modeling

Time...It marches on...without ever turning around to see if anyone is even keeping in step.

  • Member since
    December 2001
  • From: Trois-Rivieres Quebec Canada
  • 1,063 posts
Posted by jalajoie on Friday, April 12, 2013 10:01 PM

Tom, are these numbers a secret or can you post them.

Jack W.

Moderator
  • Member since
    June 2003
  • From: Northeast OH
  • 17,249 posts
Posted by tstage on Friday, April 12, 2013 9:49 PM

Derek,

I received an e-mail from the BLI Service Dept. today.  They said it could be a bad decoder but first wanted me to 1) reset the Hudson using the Power Cab then 2) change a CV17, 18, and 29 to specific values that were included in the body of the e-mail.  After I confirm that the locomotive works on the specified address, I'm supposed to remove it from the track for a few minutes then test it again.

I won't have time to check it tonight but will this weekend sometime.  They won't be reading the e-mail until Monday anyhow.  I'll keep you posted after I have a chance to run the test...

Tom

https://tstage9.wixsite.com/nyc-modeling

Time...It marches on...without ever turning around to see if anyone is even keeping in step.

  • Member since
    November 2012
  • 613 posts
Posted by UPinCT on Friday, April 12, 2013 7:08 PM

Bump,

Hi Tom,

Just wondering what BLI said.  How have you resolved the issue?  Or is it still outstanding

Derek

Moderator
  • Member since
    June 2003
  • From: Northeast OH
  • 17,249 posts
Posted by tstage on Tuesday, April 9, 2013 12:07 AM

Just sent an e-mail to BLI, explaining to them what I've described here.  I'll let you know what they have to say.

Tom

https://tstage9.wixsite.com/nyc-modeling

Time...It marches on...without ever turning around to see if anyone is even keeping in step.

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Monday, April 8, 2013 8:42 PM

 I'd just cut the reed switch out at this point. I also wouldn;t be too worried, my Atlas Trainmaster has been a good solid loco, I haven;t had to touch is after setting up the address and momentum, so far, despite months between runs (it's part of my club fleet, and I only get to play at 2-3 shows a year) it hasn;t had to be reprogrammed, it just works. And last show I had it at, it ran for hours as the middle loco in a 3-unit consist with a pair of GP7's, Protos. Despite the wierd 'things' that happen at the club - like people grabbing the wrong loco and messing up your cosists, even with basic throttles. And of course constant shorts when people run open switches, even with multple power districts you can get caught up in someone else's mess when a lot of trains are running. Never lost it's mind once.

                --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
    November 2012
  • 613 posts
Posted by UPinCT on Monday, April 8, 2013 8:17 PM

Hi Tom,

I would agree with Chuck as well.  Contact BLI.  While you are waiting for a reply have you found the reset procedure using the reed switch in the manual?  I also thought I had disabled it in JMRI but because the reed switch is bad that CV setting had no effect.  The more I read about what steps you are taking the more it sounds like my problem.  

So did you try the magnet trick on the reed switch?

Needless to say, I'm not a big fan of QSI at the moment and me with 2 more Atlas QSI Locos on order at the LHS.

Derek

Moderator
  • Member since
    June 2003
  • From: Northeast OH
  • 17,249 posts
Posted by tstage on Monday, April 8, 2013 7:25 PM

Thanks for the reply, Chuck.  I usually send BLI an e-mail rather than call.  That way they have a written record of the problem.

Tom

https://tstage9.wixsite.com/nyc-modeling

Time...It marches on...without ever turning around to see if anyone is even keeping in step.

  • Member since
    July 2003
  • From: Sierra Vista, Arizona
  • 13,757 posts
Posted by cacole on Monday, April 8, 2013 6:44 PM

It seems that you have a bad decoder, and it should probably go back to BLI for replacement; but I would phone or e-mail them first, describe all the symptoms, and see what they say.

Moderator
  • Member since
    June 2003
  • From: Northeast OH
  • 17,249 posts
Posted by tstage on Sunday, April 7, 2013 9:30 PM

Unless the reed switch is causing all this, I think something is majorly wrong with the decoder.  I have tried innumerable combination of things and I get an innumerable amount of arbitrary responses each time:

  • Resets - albeit either DecoderPro or from the Power Cab - sometimes work; sometimes don't - The locomotive does respond to each of the three CV values after I press the ENTER button
  • I was able to get the locomotive to respond to the 4-digit address...once - Most of the time though it won't respond to either the short or the long address
  • Enabling or disabling the reed switch makes little to no difference
  • When I press the "Read full sheet" button, I many times get a number of "Program error - No locomotive detected (301)" messages - If and when I don't, most of the CV values get set to "255"
  • Either the whistle, bell, or the light will work when the locomotive does respond to commands but not all three

Update #1: The bell does respond to F1 and the horn, F2.  However, they will NOT respond to either the HORN/WHISTLE button or the BELL button.  The headlight does not respond to either the HEADLIGHT button or F0.

Update #2: Just out of curiosity I checked a couple of other sound locomotives and it seems that the issue was with the Power Cab itself.  (For whatever reason the BELL button turned on the headlights.)  I did another Power Cab throttle reset and everything is copacetic again.  Now I'll need to check it with the other locomotive again.

Update #3: Now I can get bell and lights to work on the troublesome locomotive but the horn will not play whether I press the HORN/WHISTLE button or F2.  Then I use DecoderPro to reduce the overall volume by 25%...now the locomotive doesn't respond at all.  That seems to be par for the course.

  • I performed a cab reset of my Power Cab to see if that made any difference but, sadly - no

Still a reed switch issue?  Or should this go back to BLI for repair?

Tom

P.S.  The locomotive is now responding to the 4-digit address again but I didn't do or change anything to make it do that. Tongue Tied

https://tstage9.wixsite.com/nyc-modeling

Time...It marches on...without ever turning around to see if anyone is even keeping in step.

  • Member since
    November 2012
  • 613 posts
Posted by UPinCT on Sunday, April 7, 2013 8:25 PM

Well before going to the drastic step of cutting the reed switch.  See if it is just stuck.  You should have a procedure for resetting the loco using the reed switch in the manual from BLI.  Try that first.  If not the reed switch could be stuck.  I had some success unsticking it by using the magnet that Atlas provided.  I did this by putting the magnet in the middle of the reed switch and dragging it to one side.  This seems to allow my loco to operate properly until the next time I get a short on the layout, then its right back to address 3.

Good Luck, Derek

Moderator
  • Member since
    June 2003
  • From: Northeast OH
  • 17,249 posts
Posted by tstage on Sunday, April 7, 2013 8:18 PM

According to DecoderPro under the "Sound Control" tab, this particular locomotive comes with a reed switch that can be disabled/enabled.  I glanced over the other thread that UP provided.

So, clipping and/or unsoldering the reed switch would eliminate the issue happening again?  Do I need to bridge the gap at all?  I'm going to try changing CV29 first to see what happens before clipping the reed switch.

Tom

https://tstage9.wixsite.com/nyc-modeling

Time...It marches on...without ever turning around to see if anyone is even keeping in step.

Moderator
  • Member since
    June 2003
  • From: Northeast OH
  • 17,249 posts
Posted by tstage on Sunday, April 7, 2013 8:05 PM

UP,

This is a 1st run BLI Dreyfuss Hudson.  Thanks for the link to the thread.  I'll take a look at that and check about that reed switch.

Chuck,

I always address/reset my decoders on the programming track using the Power Cab (in this instance using CVs49, 50, and 56, respectively) so I never physically set CV29.  I will take a look at it though.  I also turn off DC operation so that it will only operate on DCC.

Actually, I just went into DecoderPro and looked at CV29 under the CV tab.  After the earlier reset, the value for CV29 is set at "50" and this is listed as "From file" under the "State" column.  I'll write "34" in that cell and see what happens.  Thanks for the idea...

Tom

https://tstage9.wixsite.com/nyc-modeling

Time...It marches on...without ever turning around to see if anyone is even keeping in step.

  • Member since
    November 2012
  • 613 posts
Posted by UPinCT on Sunday, April 7, 2013 7:13 PM

Hi Tom,

Is this an Atlas Loco, was the QSI factory installed, and does this decoder have a magnetic reed switch for factory resets?

I had a similar problem and found I had a bad reed switch.  Both Atlas and Randy suggested clipping it off.

Anyways here is the thread

http://cs.trains.com/mrr/f/744/t/215401.aspx?sort=ASC&pi314=1

  • Member since
    July 2003
  • From: Sierra Vista, Arizona
  • 13,757 posts
Posted by cacole on Sunday, April 7, 2013 7:09 PM

Instead of the decoder losing its long address, it may be a case of CV29 not being reset.  If CV29 does not stay set to tell the decoder to run on the long address, the decoder defaults to the short address of 3.

I usually set CV29 to a value of 34 to tell the decoder to turn of dual-mode DC operation.  The next time you have this problem, first try setting CV29 to a value of 34 and see if it will then run on the long address.

Moderator
  • Member since
    June 2003
  • From: Northeast OH
  • 17,249 posts
QSI decoder losing long address
Posted by tstage on Sunday, April 7, 2013 6:11 PM

I have a QSI-equipped locomotive that's giving me some addressing problems.

After about a year of not running this particular locomotive, I went to run it 3 weeks ago on a friend's layout and it wouldn't budge on it's long address.  My friend suggested that I try dialing in the short address and it ran fine on address "3".

I brought the locomotive back home with me and tested it out on my layout.  Same thing occurred so I reset the decoder to factory settings and readdressed to the long address.  It then worked on my layout using the long address.

A week or so later I was back at my friend's house and the same thing happened with the same locomotive: I'd dial in the long address and get nothing; I'd dial in the short address ("3") and it would go.  (Just to be clear, I didn't operate the locomotive between the time that I readdressed it at home the week before to the time I ran it on my friend's layout the 2nd time.)

Yesterday I finally installed and got my home computer hooked up to DecoderPro and started learning how to use it.  (So far I really like. Yes)  One-by-one I read and recorded the CVs to each of my locomotives so that I could have a record of them.  When I put the aforementioned QSI locomotive on the programming track I had all sorts of problems reading and writing to the decoder.

I'd set the short and long addresses, set the speed table, and read and write CVs but the locomotive wouldn't run using either my Power Cab or the built-in DecoderPro throttle.  When I looked at the locomotive addresses under the "Basic" tab, I noticed that their values had changed to something completely different than what I had entered earlier.  I'd reset the decoder and the locomotive would run on address "3".  However, as soon as I would address it again, it would stop running and wouldn't responded to either the short or long address.  Tried this several time but with the same result.

Does this sound like a faulty decoder?  I haven't had any issues with any of my other decoders losing their brains.  This one just refuses to either take or hold a long address anymore.  Any suggestions you all might have would be appreciated.  Thanks for the help ahead of time.

Tom

https://tstage9.wixsite.com/nyc-modeling

Time...It marches on...without ever turning around to see if anyone is even keeping in step.

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!