TrainWeb.org Facebook Page
MRC-CVP Compatibility Solution

 

MRC-CVP Compatibility Solution‏

 

Wanted to share an experience integrating my MRC PA2 system with a CVP Zonemaster saga. Basically I had the PA power about 1/3 of the layout. As the layout expanded, I bought a CVP Zonemaster to power the other 2/3.

 

The Problem

 

All worked well, except I had an intermittent problem with boosters tripping out when a loco crosses between two boosters. One booster is my MRC command station; rated at 3.5 Amps. The other is a CVP Zonemaster Single, rated at 7 Amps. I measured the voltage and the rails were properly wired. In fact, if it is just a single loco, usually there is no problem. However if I had a 2-3 unit lashup of sound equipped locos, one or both boosters trip out on an overload. Both boosters handle all the loads just fine normally. I checked the wiring and all the tracks in both booster districts are truly separate, and the DCC signals are in phase. So all the usual suspects were accounted for.

 

The Issue

 

Three sources of problems were identified. Most important was that the CVP uses a different power switching scheme than the MRC. Some help from the MRR Electronics list identified incompatibility between different brands of boosters. MRC, CVP, NCE and Digitrax all use a single DC supply inside that reverses the power supply supply to the rails for each signal phase. However MRC and most European systems have a split 14 volt supply that alternate which rail gets power instead of switching each rail between the positive and negative supplies.  Thanks to Graham Dredge of Westimation for his help in sorting this out and drawing up a sketch of the differences….

 

 

 

 

This has caused lots of problems for NCE users that have tried to add a CVP ZoneMaster booster. (His suggestion is never to mix boosters – a wise thought.) The second issue might have been a slight phase shift in the DCC signal. Because the DCC input signal to the CVP has to go thought the opto-isolated input and it's logic, it is likely that a slight time delay was introduced. When locos crossed the gaps between boosters, this created signal noise to the decoders. Finally, the two DCC systems ran at slightly different voltages, about a 1.8 v difference.

 

The Solution

 

So the solution I came up with was to buy a second CVP booster and have it power what the MRC did. OK, not exactly an earth shattering lighting bolt idea, but what it does is creates an identical power feed scheme for both booster districts.  Symmetry.  It’s a good thing.  Even the Romans and Greeks thoughts so.

 

The Result

 

I ordered a second Zonemaster and connected it into the layout identically as the first. As such, the MRC PA is no longer directly connected to the tracks in any way. Lo and behold, it works great! Not the slightest hesitation when crossing boosters. I jumpered one rail together and measured the voltage across the gap on the other rail - 0.2 volts difference. Much, much better than the 1.8 volts I had before. So between the booster design matching, the phasing issue being removed and voltage matching, we have success.

 

During all this testing, apparently I did create a surge somewhere along the line and fried an output from a stationary decoder. (I also had to reset and reprogram a couple locos as they did not appreciate all the DCC noise that this problem caused.) As I was thinking out the new DCC system topology, it occurred to me that the 3.5 A out of the MRC PA would essentially create a 'quiet' bus. So I rewired all my stationary decoders to the MRC output. Now they are in a place on the system that will never see shorts or surges. Very happy about that. I have considered extending this lead out to my workbench for the same reason. I drew the new system topology up and posted it in the files section. Not everything is shown, but it gives the idea.

 

 

 

Thought it might be useful to those of you who have a PA and want to expand its capacity beyond what the command station can supply.

 

Regards,

 

Detlef

 

 

 

 

 

 

Back to Index