Re: PC based storage oscilloscope

Vollrath, Don <dvollrath@...>


An interesting problem indeed. You described an ideal use for a miniature DC coupled clamp-on hall effect type current probe. I have borrowed one on occasion to help diagnose other current related issues.



From: WiringForDCC@... [mailto:WiringForDCC@...]
Sent: Friday, November 13, 2015 9:05 AM
To: WiringForDCC@...
Subject: RE: [WiringForDCC] PC based storage oscilloscope



I am not intending to dig into the detector internals but rather observe the detector inputs to, hopefully, be able to track down a noise source.  Only a couple of detectors are problematic so I should be able to differentiate good from failing and thus determine the problem source.

Having said all the above, what I discovered is that noise is not the real problem, although there is some present and appears to contribute due to the real problem source.  In other words, I found that the detectors are working correctly and the problem was real current flow in unoccupied track due to other issues.

So the scope did help point me in the right direction.

In one failing section the problem turned out to be a high resistance short between contacts on a switch machine in the block.  The contacts are used to power route the switch frog and was causing a small current flow which was reducing the sensitivity of the detector and allowing noise to cause detection.

I suppose I would have eventually found it but the scope made it a lot easier and quicker.

Dale Gloer

---In WiringForDCC@..., <dvollrath@...> wrote :

So you are digging into the inner workings of the detector itself?

One interesting thought is that if there is a constant rail-rail current (think resistor wheelset) detectable current should flow during the entire DCC waveform cycle (except during the 1-2 uSec or so during voltage reversals).


Most detectors are sensitive to peak currents during the switching interval as well as during the rest of the steady value portion of the  ½ DCC waveform. These can be fooled by peak currents that flow due to capacitive effects of wiring and/or track as this often causes polarity charging and sometimes oscillating currents to flow. NCE type boosters with fast voltage transitions aggravate this issue.


For CT type detectors with a doughnut hole there is an interesting way to connect a capacitor backwards to counteract the effects of wiring capacitance. Look for it within the bowels of wiringforDCC notes and photos.


An interesting twist would be for a smarter uP inside the multi-channel detector to look at the condition of current signals only after the voltage transients of DCC track voltage switching are likely to have passed. If the detector circuitry is powered by DCC that possibility is relatively easy. Sync the current detection interrogations of individual channels to be synchronous with but slightly delayed from DCC voltage transitions.


I’m not sure how Dick B of RR CirKits does it, but it is worth mentioning.




Join to automatically receive all group messages.