Said another way, in a simplied example where there is only an AIU with a direct connection to tracks (that is, no intervening relay), the PBW would be connected to the inputs to the AIU and the signal is only received by the track that is connected when the AIU internal relay is energized. In that example, and assuming no AIU relays are closed, and with the transformer having its handle engaged, the PBW would actively be generating a signal since it’s positive and negative is connected even though no tracks are powered.
Correct. If using just the AIU's ACC port relays, if none of the ACC ports are "ON", the PBW would be generating the blasting out the watchdog signal once per second...but no one would be listening! Oh well. There are bigger problems in the world!
Note that the AIU ACC port internal relays are capable of 3 or 4 Amps (something like that). This should not be a problem for roundtable whiskers, sidings, etc. since you shouldn't be zooming engines with long loads like you might on a main-line section. In which case you'd have the ACC port power an external $1 relay that can handle the "full" 10 Amps for a TIU channel.
Just an update, I decided to go ahead with the 555 timer version as it has less parts and also requires no calibration. So I built half a dozen and put one in my remaining spare DCS-RC box in case one of the ones on the layout ever craps out.
Of the half dozen I built, I only had a 3% variation in the time interval that they trigger, well within the tolerance needed to generate a reliable watchdog signal. I didn't have one of the 1% resistors, they're trapped in COVID delays. That being the case, I simply hand selected from a batch of 5% resistors using a 1% resistor as a standard to match them using a Fluke bench meter. Most of the timing variation I see probably comes from the 5% capacitor, a .47uf 1% capacitor is price prohibitive, it cost more than all the other parts, including the PCB!
555 Timer Based Watchdog Generator Board
The three pin female socket mounts as illustrated in the three holes at the end of J2.
The WD board plugs into the female socket.
I do a lot of SMD, this was laid out as a kit so I did it with thru-hole stuff as many people have difficulty with SMD. For this application, there was no benefit to SMD, there's plenty of room for the board inside the DCS-RC where it installs.
I do have a size limit on hand assembled SMD, usually 0603 for resistors and caps. They're still pretty easy to handle. I once screwed up and specified some 0201 resistors on a board, I couldn't even see the little things! It took me forever to build the prototype, and I immediately sent that sucker out for a rework with larger parts!
Most of my products that are sold through Hennings are SMD. Also, most of the stuff I've build for local use here is SMD, only a handful of the projects end up being thru-hole. If it might be offered as a kit, I look at thru-hole. Also, if I'm going to hand assemble a bunch of them, I may consider thru-hole for at least some parts, it's usually quicker to do.
IF you make more of these I will take two of them.
John, interesting comment on the 0603 smd size. With my limited (but growing) smd experience, thats pretty much the smallest size that seems practical. I accidentally ordered a few 0402 resistors a while back, and promptly tossed them out when they arrived.
I was happy to see that all parts for your smart motherboard project are at least 0603, and it went together quite nicely. I finally figured out an smd soldering technique that seems to get the job done with minimal muss and fuss. The only pesky part on that board is the tiny little indicator led, and being polarity sensitive, you need to get it right!
Anyway finally after all your prodding I took the smd plunge, and here we are. It seems to be working! Nice when an old dog can still pick up a new trick or two.
Once you get used to handling them, they're not too bad. I don't think you'll ever be quite as fast as with thru-hole, but you can make much smaller boards with more functionality, which is why I took the plunge. Also, nowadays, it's harder and harder to find some functionality in thru-hole parts, everything is going SMD.