[Bret W Tesson]

After two weeks of operation, my QSI Aristo board has ceased to function in a really odd way. My setup involves the QSI Aristo board and g-wire receiver installed in a Bachmann K-27 that is controlled with the Airwire T9000 transmitter. I now get zero response on track power. On battery power (with my Airwire throttle turn on) I get zero response. With battery power on and the Airwire throttle off, the sound card starts up and the engine begins to move forward (uncommanded). I have no control even if I then turn the throttle on. I put the locomotive on the programming track and have zero communication with the quantum programmer.

What's really odd about this is the way it happened. I had been running on battery power for a little over an hour when suddenly my K-27 stopped abruptly. Assuming the battery was dead, I switched to track power and used the locomotive for another 5 minutes to put it away. During that time, I did notice the dreaded burning electrical smell, but couldn't find specific source (I had other engines operating at the time). The next time I took the engine out, everything was dead as described above. This is my first QSI install and I hope this is not indicative of QSI's quality.


For now it's back to straight DC track power for my K.

Bret

Update: I spoke with Lewis at QSI and he was very helpful. He had a couple of suggestions which I'll try tonight. The first was to reset the decoder using the reed switch and a magnet. He also mentioned that a few of the new boards got released with the ever problematic poly-fuse installed on the board. This yellow colored thermal fuse IS on my board. He suggested I remove the fuse and solder the wires together. The lastest releases of the board have this fuse removed. He also said that this work on the board won't void the warrant if this doesn't fix the problem. I'll let you know what I find tonight.

 

[Greg Elmassian]

Bret, I have about 10 QSI's and only had a problem on one, a very early one where it ceased to power the lights.

I have run these so hard that I overheated it, and it just sat there, sound going, motor off, announcing the temperature until it was cool enough to keep going (about 40 cars behind an Aristo mallet).

The polyfuses themselves generate heat when near full load, and they help overheat the board! The board has temperature sensing built in, and the polyfuses have been proven to be problematic when run near their rated value (they actually wear out).

I shorted the leads of mine, and the Mallet no longer stops. You can just put a piece of wire in the "kinks" of the leads of the polyswitch, rather than mess with removing it from the board.

Did you have any trouble programming it before this incident?

Regards, Greg

 

[Bret W Tesson]

Update 2: I followed the suggestion from Lewis at QSI and did a reset of the sound card/decoder and that did the trick. The reset is very simple. Just hookup the included reed switch to the decoder, unplug the G-wire receiver, put a magnet next to the reed switch and apply power to the decoder. After about 2 seconds it will announce "Reset". Of course this causes all of you CV values to go back to defaults, but if you use the Quantum Programmer, you should be able to restore everything pretty easy.

What I don't understand is why this happened in the first place. Like I stated earlier, it was working fine when I put the engine away for the day and dead the next time I used it.

Greg, I was successful in initially programming the K with the Quantum Programmer (thanks for the help). I'm sure I'll have no trouble putting the CVs back to the desired states using this system.

Bret

 

[Greg Elmassian]

Strange, more like the decoder got the lights disabled. Good news though on the reset, will remember that.

I have an NCE system, so I can use DecoderPro, and save complete configuration files, and just blast all the settings in one shot.

Regards, Greg

 

[Bret W Tesson]

One last update. When you reset the decoder, CV1 goes back to 3 (default). Using the Quantum Programmer I can then change the CV17/18 (long address) to 453 for the engine (like I did before this thing died). Everything works fine on the programming track, but as soon as I disconnect the Quantum Programmer, reattach the G-wire receiver, and apply power, I again have ZERO control of the engine. I repeated this at least 4-5 times with the same results. Just for kicks, I left CV1 at 3 and didn't change CV 17/18 (left them at default 0). After doing this I again hooked up the G-wire and now everything works fine....with CV1 set at 3. Obviously, something is still not right in the decoder. For some reason, I'm stuck using CV1=3. It's not the end of the world, but it's not right and will give QSI another call or return the decoder for repair.

 

[blattan]

Bret, I suggest that you check the value in CV 29. It needs to be one of the four values in the G wire instructions. The specific value depends on two or four digit address and forward or reverse normal direction of travel. Bert

 

[Greg Elmassian]

Sure sounds like it Bert, since Bret mentions going from the programming track to the Gwire, so it's possible he did not test the long address being active (the programming track is independent of address).

Bret, you HAVE to change cv29 to have the decoder USE the long address...

Regards, Greg

 

[Bret W Tesson]

Bert & Greg,

Thanks guys. I'll try that tonight. This DCC stuff is obviously new to me, but I see a lot of potential. I'll let you know.

Bret

 

[aceinspp]

Just for info I did as Greg recommended and jumpered the poly fuse as one of my boards was over heating and shutting the engine down. works fine now. later RJD