Easy Fix for QSI Reversing Problem

Buy Broadway Ltd Locos Here
More About Switchers
Reversing Issue
Programming Switchers with PowerPax

A few days after we posted the notice on th BLI Switchers from Broadway Limited, Tom M. from Maryland contacted us with this information:

Ken,

"Several members of the Atlas Forum have encountered the direction change/momentary short problem with their QSI sound-equipped Proto 2000 GP9s as well as with BLI switchers. One Atlas Forum member discovered a quick and easy fix. If you change the decoder's momentum settings to any value other than zero, the problem goes away. Specifically, you need to change both CV3 and CV4.

In my case, the factory decoder settings on the QSI sound-equipped Proto GP9s would trip the circuit overload protection on my Digitrax PM42s every time I changed the direction of travel. This would, of course, cause all trains within the affected power district to stop. Once I changed CV3 and CV4 to a value of 1, the problem disappeared. This minimal momentum setting is barely noticeable and has no adverse affects on normal train operations."

We proceeded to check this information out for ourselves and received this test result from our senior technical consultant:

Ken:

I used the LLP2K GP9 w/QSI since I have already repacked the BLI SW7. With CV3 and CV4 equal to 0, I was able to easily reproduce the current spike that I documented in my report to Tony on the BLI Switcher. With CV3 and CV4 both set to 1, I was still able to generate the high current spike on some reversals. I increased CV3 and CV 4 in unit increments until I could not reproduce the current spike. The minimum CV value for this condition is 5. A value of 3 or 4 reduces the occurence of the spike, but CV3 and CV4 set to 5 is the lowest value for which no current spike was observed. At this momentum setting, there is a discernable momentum effect, but it is probably about the right value for the engine all by itself. This is certainly an easier solution than ripping the unit open and replacing the software.

So although the two testimonies differ by degree, there is a simple solution. Even though the oscilloscope sees a spike at a value of 1, it may not be enough to cause the problem. If you don't want any momentum, gradually increase the values of CV3 & 4 until the problem goes away.

We welcome comments or suggestions from readers; please write or call.