I've been asking a lot of questions on here lately about setting up my Atlas signal system. I've also been talking to a company that provides assistance in setting up signal systems. I sent NCE a question about this, but I haven't received a response yet and wanted some expert opinions on here as well (I contacted them just a few hours ago, so it may be a day or more before I hear back). I am using Digitrax DCC 5 AMP booster with Atlas N scale signal system and NCE BD20 block detector. I've been doing a lot of testing on this system to determine exactly what I can do with it and to discover any limitations or hidden things I didn't know about. The question I have is over the sensitivity of the BD20. According to the manual, when I place the wire through the BD20 just once, the sensitivity of the detector should be 8ma. In other words, I have to use something that draws 8ma or more for the detector to detect it. I'm sizing resistors to wire into the track to make it more sensitive so I can operate resistor equipped rolling stock. The problem is that the BD20 does not behave the way the manual says. When I place 5.45 milliamp resistors across the track, the detector's light goes red. When I place 4 milliamp resistors across the track, it does not glow red. Now, with 5.45 milliamps on the track, the detector shouldn't detect anything with the track wire looped through only once, but it doesn't work like this. Here's another example, when I looped the track wire through 4 times, it should have set the sensitivity at 1.9ma. I tried a resistor that was 1.2ma and it detected it. I dropped it down to 3 loops (2.5ma sensitivity), and it did not detect the 1.2ma resistor. This doesn't make any sense and makes it harder for me to determine how many resistors and of what value to wire into the track to make it more sensitive for the resistor equipped rolling stock. Here is what the manual says exactly: "Table 2 lists the number of milliamps through the track feeder wire required to bring the logic output of the detector down to 0.7 volts (a 5 volt supply with a 10k pullup resistor is assumed). This is low enough to give a reliable indication on the TTL logic chips used in most signal systems." Now, I've asked this before, but I'm beginning to wonder if DCC current doesn't somehow draw more from a resistor than DC or AC. Either that, or there's an issue with this pullup resistor setting that I'm missing (I don't know what a pullup resistor is or what it's for). Also, I'm using Kato N unitrack without any added ballast, glue, oil, etc... I don't have the track wires twisted around each other (I know this can cause crosstalk), but they are joined together like a household extension cord. Am I dealing with a little crosstalk from this or what?
I think you are making this way harder on yourself than it needs to be. The number of loops is determining the minimum current that the BD20 will detect. It's probably not an exact number, it's going to depend on a lot of things interacting... the length of the feeders, the wire size, the tightness of the wrap, etc. I rather suspect that the value they give for a number of wraps is the highest minimum they expect, in other words it WILL detect currents above that value, and it might detect currents slightly below that value. That seems consistent with your results. If we assume your DCC system is putting 12V to the rails, more or less, a 1k resistor would give you a nice 12ish mA, and you'd detect it reliably with one wrap. You could make the resistor 3k, and have about 4 mA current, which would probably work nicely with 3 loops.
As far as the pullup resistor, it's the resistor that goes on the output of the device to give you a high output when the device is inactive. Since the signalling system allows the detectors to be wire ORed on the input, I expect there is already a pullup there, though we don't know the value. It is possible that there could be a relationship between the value of the pullup, and even the supply voltage on the pullup, and the actual sensitivity of the device, so your experiments are verify that the setup you choose works are not a bad plan.
DCC follows Ohm's law, just as AC and DC do. However, because of the shape of the waveform, you probably are not getting valid measurements for voltage and current with your meter. But it really doesn't matter for these purposes. Assume 12V to the track, calculate a resistor that gives you a bit of margin in the detection current, test it, and be happy!
Jeff But it's a dry heat!
Contacted Logic Rail Techs and this is what they called this phenomena: "In all things electronic/electrical there is always "guard band". That is,
slop! Nothing is ever specified EXACTLY - there is always a tolerance orallowance for variance. For example, common resistors have 5 or 10%tolerance. So, a 10K resistor could actually be 9.0K or 11K!In this situation you have the combination of the tolerance of the voltageon the rails, the resistor, and the detector's threshold! So, I'm notsurprised that the detector tripped during your test!" Maybe I'm going overboard on this, but I want to have a full understanding of why this system works the way it does. This pretty well answers my question.
If you are concerned about sensitivity, order some of the suggested trimmer potentiometers and solder them on like the instructiosn show, this gives you a simple way to adjust the detection sensitivity so you can make sure all of your detectors trip the same way.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
rrinker If you are concerned about sensitivity, order some of the suggested trimmer potentiometers and solder them on like the instructiosn show, this gives you a simple way to adjust the detection sensitivity so you can make sure all of your detectors trip the same way. --Randy
We use Team Digital's DBD2's which is similar to (NCE's detector) to detect occupancy on our NTRAK club layout. I installed 0 to 5k pots on each one because conditions varied so much each time we set a layout up..
We have a 16 volt DC supply connected to LM7805 and LM7812 regulators. A SPDT switch between those and the DBD2 allows for changing input voltage.
Martin Myers
Thanks for the replies. I ordered some resistor wheels from Logic Rail Tech at 2800ohm, which should be approximately 4.29ma per resistor. What I'm going to do is wire in a resistor that's approximately 4ma to the track that will lower the amount of power needed to be drawn on the rails. Basically, I will increase sensitivity on the track. If 4ma causes the signals to drop to red all the time, I'll just replace it with a resistor that draws a little less power such as 3.5ma or 3ma.