How to assign and retrieve the horn value (i.e. diesel air-chime whistle) in a Titan2 equipped diesel locomotive


The way to assign a chime horn is as follows:  Let's say you want the beautiful Leslie S5T sound for your loco, as was used on the Frisco, Santa Fe, CNW, N&W, etc.  You first assign a "1" to CV 50, then assign a "1" to CV 49, and finally assign the horn value you want, which in this example is listed at page 127 in the August 5, 2015 QSI Manual (a pdf file that you can download from the web).  You then assign value "35" to CV 55, which is the value for the Leslie S5T, and voila, you engine now sports a Leslie 5-chime sound.

On my NCE Power Cab, to retrieve the horn value assigned to CV 55, remove all locos but the one in question from the program track, hit the prog key four times to use the program track, enter, hit "2" for for CV, and enter "55".

For a good primer on diesel air-chime whistles (we called them "whistles" when I fired on the Milwaukee in the '60's, as did the Leslie Company), visit this web site, which hopefully will be updated with Nathan air chimes at some point:

Greg Elmassian

you can also announce the cv settings by using cv64, which reads back the value... you can do this on the main (POM)....

do POM CV64 = 2 (edited to correct) and the decoder will announce the short address.

similarly, you can use index values, so you do (POM)

CV49 "X"
CV50 "Y"
CV64 "Z" where you want to read the CV of Z.x.y  using the nomenclature from the manual...

This way you don't need the program track and don't need to take locos off the track (although I submit if you have a dedicated program track, you should already be in the habit of never having more than one loco on this at ANY time!)




Greg has highlighted one of the really useful features in the QSI Titan, Revolution and all version 7 decoders.   The decoders can 'speak' using the CV 64 query.  The example Greg used needs a slight tweak, in that the short address query is Program On Main --> CV64 = 2

POM CV64 = 3 will get you the value for acceleration momentum.

Best regards,



KJ mentioned the "talk-back" feature in his response about Greg email. As someone that programs a LOT of decoders for our club and my friends, I must caution new users about this. The "feature" is actually a real problem unless you are only programming ONE CV at a time. If you are using any software like QSI CV manager with the programmer, or especially JMRI and any DCC system command station, the "talk-back" feature actually causes problems. Most "old guys" like me that have been doing this since the version 7 came out are aware of the problem and we disable that feature as the very first step in programming a decoder.

Information, without an explanation, is only opinion unless proof is supplied. When programming indexed CVs, with software, the CVs are sent in a multiple command message to the decoder. When talk-back is enabled, the decoder immediately attempts to provide the verbal information and then "tramples" the rest of the incoming input to the decoder. This results in a failed programming attempt that often causes inexperienced modelers to complain that the decoder will not accept commands.

It has been documented often in different groups and even in information packs sent to BETA testers like me. I felt it was necessary to make sure you have this information.

As a BETA tester I also learned a very important piece of information about the version 7 and Revolution decoders. After programming the decoder, if you are using the programmer and QSI CV Manager in OPS Modes (mainline programming), ALWAYS cycle power before testing the decoder. If you fail to do this, the decoder will not reflect the new information. Since I've been doing this since the Version 7 came out, I continue this process with the Titan as well but I do not think it is needed with the Titan.

Take care my friends and have a great (USA) Thanksgiving

Nick Kulp

Dave Robinson

Um, that would be "1", not "2", right KJ?
I hadn't noticed the error in Greg's but you're right in spotting it.

I too think this is a great feature of QSIs. I wish other brands would have this.

In snowy Edmonton AB,


Dave, thanks for pointing out my error pointing out Greg's error.   Hehehehe!   Yes, the correct CV number is 1 for short address.   CV2 is for slow speed minimum.  

Nick, thanks for pointing out the talk back issue.   I think there needs to be a distinction made between 'talk back' and 'Query'.    Talk back is the decoder confirming a single CV change.   As Nick points out, the talk back feature will 'rudely interrupt' a programming sequence to report on the first CV change in the sequence.   Further, Nick disables the talk back feature as the very first step in programming the decoder.   I follow the same practice.   For those unaware how to turn off the 'talk back' feature:    CV62 = 0       (side note: CV62=1 will turn the talk back feature on)

 Query uses CV 64 to 'speak' the value of whatever CV one wishes to read (while still on the main as opposed to on the programming track).   Simply program CV64 to whatever CV# you wish to 'read'.   Handy!


Greg Elmassian

Thanks for finding my error, CV2 is indeed what is needed for the short address, not CV3 ha ha!

I forgot to mention I turned off the talkback function from that issue, the problem is described clearly here, but I want to emphasize that with the talkback function off the CV64 query function works perfectly, and is very helpful for POM, or just checking settings instead of hauling the loco off the the programming track, I run G scale, the locos are big and heavy, and the programming track is sometimes far away in direct sunlight!