We (the West Island Modular Railroad Club, www.wimrc.ca ) have concluded that trying to trouble-shoot problems with our DCC system with only a conventional Sperry Multi-meter is like trying to wallpaper with only one arm, or worse. We’re looking into what relatively inexpensive means are readily available that will allow us to measure voltage and signal characteristics at various points on the layout under different operating conditions eg. number of loco’s, sound/no sound, consists, number of throttles etc.
Any recommendations and suggestions would be appreciated.
You could use a bridge rectifier to measure the voltage. You can also use a decoder, and measure the function output voltage and add a couple of diode drops to calculate the voltage.
I don’t know of any handy way to look at ‘signal characteristics’ without a scope. You could verify that things on the rails are good by wiring up a decoder in a test box with LED indicators on the functions and the motor output (or put a meter or a motor on that) and using it to verify that commands get to the decoder (just as if it was a loco).
Signal characteristics is way out of my field, so I have no suggestions. But, Tony’s Rampmeter, as suggested, gets high marks. One advantage is that it can be placed against the rails anwhere, hand held, with the end-board contacts touching the railtops to get a local and instantaneous reading…pretty spiffy, don’t you think?
I seem to get a fairly close reading to my track voltage from a digital multimeter set to the 20 volt range and ac current. I get readings in the 15.7 volt range, which should be what my DB 150 is putting out to the tracks, plus or minus 0.5 volts.
Voltage and amperage are fairly easy, but wow, the only thing I know of to measure the actual signal characteristics is an ocilloscope. They aren’t inexpensive. There might be some computer programs that could tap in and display out the various settings the DCC system is sending out.
The other thing I’d say about signal characteristics is, what do you want to see? And what are you going to do about it? The problems you are most likely having to deal with are going to be shorts and opens. A vanilla meter will be all you need to track those down. The RRAmpmeter will tell you when you are drawing too much current and taxing a booster (which probably leads to bad signals). It isn’t like you are going to be adjusting the DCC signal. For a digital signal, DCC is pretty slow, and huge. So if your trains respond, it is probably good. If you have response time issues, you’ll want to make sure you don’t have a DC loco using address 0 running (zero stretching takes time), otherwise, it is probably dirty track, or possibly issues on the throttle bus side, though that ‘shouldn’t’ be.
You could get an I/O card in a PC, and esssentially build a special purpose scope, but you are talking about a bit of software, and think getting out of the range you were thinking about. Especially since I don’t know what it would really buy you.
to measure signal characteristics your only option is an O-scope. You can get one for a few hundred to a few thousand depending on how many channels and how many features (signal memory is great when packets are going through too fast for your eyes to see)but to get features like that you will spend more $$$. If you are having response problems your signal bus could be loaded too much causing a weak signal. My modular club runs into this problem when we have a large layout and no one brings transformers for the UP5 pannels. I am not familiar with Lenz, NCE or other DCC systems requirements aside from Digitrax. On a Digitrax system you dont need to power every UP5 pannel but there is a place to add power to the pannel. By adding power to a few of the UP5s you will boost the signal strength from that point on to the next. Another thing that will help with signal strength is soldering the rail joiners. If you dont have a feed for every section of track then you should solder the joiners. If you dont want to or cant solder the joiners then see about adding more power feeds.
Our four power buses, feeding four power districts from a PM42, are 14 AWG with feeders to tracks on each module. The longest bus is about 52 feet from the DCS100 8 Amp command station/booster. Each UP5 has its own 12 V DC wall wart. My understanding is that feeeding 12 V DC to the UP5 provides power to the plugged in throttle, avoiding pulling power from the DCS100 and dragging down the power/signal at that point (but not boosting the signal strength). We are currently using only plug-in throttles. Most track joiners are either soldered or have parallel jumpers. So far we haven’t found any significant difference in the track voltage on the modules fed by the 52 foot bus.
12V from a wall wart is not providing enough power to your DCC boosters. Most manufacturers recommend 16 to 18 Volts AC at 5 Amps or more output to a power booster in order to get sufficient power on the track. Internal losses within the booster are probably cutting your output voltage to around 10 Volts, which is too low.
Measured with a digital AC Voltmeter, the DCC power on the track should be in the range of 14 to 14.5 Volts. Even though most AC voltmeters can’t respond fast enough to accurately measure the DCC high frequency square wave, they can usually give you a reading that is within one volt accuracy.
A DCC signal of too low a voltage is also draining your available amperage in an attempt to make up for the low voltage.
I think you misread one of the previous posts cacole. He (Isambard) is using the 12V Wall-Warts to power UP5 expansion panels (and consequently the throttles and possibly boosting the loconet signal as well…), not the boosters themselves.
He’s using the wall warts for his UP5’s, not his boosters. Big difference. Depending on how many throttles you have plugged in at once, you only need a few hundred ma’s. If they actually have one w-w per UP5, that’s probably overkill. Won’t hurt anything as long as the polarity on all of them is correct, but it’s probably way more than they need.
That aside, I’m still curious as to what kind of problem he’s trying to troubleshoot. As others have mentioned, a 'scope is really the only way to look at the waveform. But you have to know what you’re looking at, and what you’re looking for. Here’s some shots from a guy who knows about DCC… http://homepage.mac.com/jacobsen/PhotoAlbum8.html
In re-reading his initial post, it seems that he’s looking for a way to troubleshoot things for when they start going bad, as well as just general usage specs - kind of like Alan’s use of the Rampmeter on his BAH RR.
Our search for a better way of trouble shooting problems was prompted by intermittent problems with a new sound-equipped Spectrum 2-8-0, that appear to be more evident on some sections of track and certain modules e.g. is the problem with the loco decoder or with our system? (see my 29-04-2007 Re 2-8-0 as a Switcher below) Easy for me to say it’s the loco since the problem doesn’t show with other locos, but then we do have occassional unexplained operating glitches that may suggest our system isn’t as clean as it could be. A multimeter may say that the voltage is adequate but it doesn’t tell a thing about the signal quality, spurious signals or other transient effects, hence our interest in additional tools. We’ll be looking into testers like the Rampmeter for sure.
I would suspect that the loco is more sensitive to others to dirty track, but that’s a hard one to prove. Or maybe dirty wheels.
Anyway, it is going to be really hard to do any real good eveluation of the signal without considerable expertise and expense. So, you really are stuck with mainly assessment of symptoms and analysis rather than hard facts.
So, you have a few tools to gather the best info you can get, then the ‘assuming’ comes in. The RRampmeter can help you tell if you are drawing too much current, which I would expect to be the most likely thing to actually cause a problem with the signal on the rails. It’s reading will be useful since you can use it while you are operating, in ‘real life’ situations. As I suggested before, a decoder wired up as a t
All of this had me thinking that is wouldn’t be hard to make a DCC signal test box, it woul dpretty much be a decoder that stored and processed commands rather then running a loco. So, I Googled a bit and found this:
I have no idea if it is any good, or even really exists, but the concept is sound. Whether this fits inexpensive is a matter of opinion, of course. And whether it is really necessary is a topic for another discussion. But there you go…
A Tony’s RRampmeter is probably a worthwhile investment, especially for a club. However, you can measure ‘close enough’ with a conventional meter. What sort of issues are you having? While the reading on a conventional meter won;t be 100% accurate, if the meter says 15 volts on one section and 12 volts on another, you have a problem and it’s NOT the meter.
The biggest factor is having heavy enough wire to handle the current load at the set voltage over the distance required. If the wire is too small, there will be too much of a voltage drop and locos will slow, but also it can prevent the short circuit protection from activating. Hence the ‘quarter test’ - set a quarter on the rails anywhere and it should trip the circuit breaker. If it doesn’t - there is probably insufficient wiring between that spot and the booster driving it.
The RRampmeter looks like the best solution for our needs, skills and what we can do with the information (a point made earlier). Last evening I checked out the voltage levels around the club layout using a Sperry Multimeter and a Radio Shack Multimeter. Testing around the layout one gave 10-11 volt readings the other 8-9 volt readings, with the DCS100 output set on N Gauge, our normal setting. The readings on both meters oscillated a bit, plus/minus 1 volt, either due to digital meter characteristics or to power demand changes with a number of loco’s operating. At the DCS100 output terminals the oscillation was less frequent.
Regarding my problem with a new sound-equipped Spectrum 2-8-0, following recommendations I reset to factory levels and disabled analo