Blu Shark Part II:
My previous post on the Blu Shark project described operation with a bare-bones AC to DC converter: a simple full-wave bridge rectifier with smoothing/filter capacitor ahead of the Blunami card, with the Weaver/Samhongsa Baldwin Sharknose diesel's twin can motors wired in series. I continued onward to a more sophisticated conversion scheme, using a DC>DC buck converter board following the rectifier & filter to produce a regulated DC input to the Blunami card.
Here's a picture of the Nude Shark, stripped of its original QSI DCRU and electronic horn cards; the chassis and center rail roller wires were retained:
In the space above the Shark's speaker, I mounted a vertically-oriented piece of 3/16" clear acrylic to use as the (insulated) mounting surface for the convertor/regulator and Blunami cards. Here's a photo of the 3" x 1½" piece and its ½" aluminum mounting angle; loco front is too the right:
You can see the Diodes, Inc. GBU-1001 FWB rectifier mounted just ahead of the acrylic piece, using the hole from the QSU mounting cylinder.
The cheap ($3.50>$8) Chinese DC>DC buck convertor/regulator is mounted on one side, using double-sided foam tape squares. The Panasonic 1000µF 50V filter/smoothing capacitor is connected at buck converter input (left side), but is hard to see because it is the same color as the Shark carbody shell behind it.
I should note that the converter has two mounting holes, so using screws and stand-off insulators is a reasonable alternative mounting arrangement.
Don't ask about RF interference from this converter/regulator. It certainly increased the noise floor in my workshop's Bose Wave Radio on the FM band. I didn't work up the nerve to power-up my spectrum analyzer and take a detailed look at its emissions. Any correlation between its performance and FCC incidental emissions regulations is, undoubtedly, entirely coincidental! But it didn't clobber TMCC on my layout. And when you live out in the woods, the rest isn't of much concern.
I retained the 1000µF filter/smoothing capacitor used in the bare converter scheme previously noted, now mounted at the buck regulator input. The Blunami decoder/controller card is mounted on the other side of the acrylic piece, also using the foam tape squares. You can see the more-visible Panasonic filter/smooting capacitor between the acrylic piece and the forward motor, above the FWB rectifier:
You really don't have a good mounting alternative to foam tape for the Blunami, as the card has no mounting holes. I mounted it upside down, to ensure easy access to its "current keeper" port, in case that's needed. Operation to date suggests that it's not necessary. DC power input wires are red and black; motor output wires are yellow and green; speaker wires are both blue.
All wiring was #22 AWG. I contemplated using #18, but the original Weaver/Samhongsa wiring was #22, the lengths short, and the current normally under 1A. So the loss difference would be minimal and the flexibility of the #22 wire was much better. Before applying power for the first time, one MUST check the resistance to ground from each speaker terminal and the motors' two primary connections - resolving any measurement below a megohm. My first setup of this scheme resulted in a smoked Blunami when the voice coil lead of the speaker contacted its frame. On the reworked version, I put heat-shrink tubing over the speaker-end solder connections and carefully inspected the voice coil wires, in addition to checking the load-to-chassis resistances as noted above. The speaker was new, as well.
Sitting at idle, with track powered by a Lionel Powerhouse 180, the DC>DC buck converter/regulator was set for an output of 18V. The locomotive was run with a test train consisting of a dummy B unit, 16 "average" freight cars, and a lighted caboose. The BluRail iPad app includes a Blunami real-time input voltage measurement display. This never dropped below 17.8V, and then for only a short part of a 3% incline, when the entire train was headed uphill. Speed loss upgrade didn't seem significant. Shark trainspotters reported them as "luggers" that seldom exceeded 45 mph in real-life. The Blu Shark seems capable of at least that, uphill, with a heavy train, but won't be a ballast burner otherwise.
There are many theoretical reasons to favor the use of the rectifier + converter/regulator over the bare-bones FWB rectifier and smoothing/filter capacitor. The practical implementation proved those benefits - the locomotive performed consistently, regardless of train size or track gradient. Track voltage anomalies were essentially invisible. I'll be posting a more technical discussion of the measured data and its implications.
FWIW, while I had the Shark disassembled to implement this conversion, I also converted its pilot to a fixed mount, attached to the loco's base plate instead of to the front bogie (a/k/a truck). This resolved a problem of the uncoupler pad shorting the center rail in certain Ross switches, but at the cost of giving the Shark somewhat of a "high water pants" look from track elevation. Ignoring the latter, it seems to match prototype Shark pix better. Yet ...
Headlights, markers, and electrocouplers remain on the horizon in the Blu Shark. I set up the incandescent headlight for 3V, but it burned out after a half hour or so of operation, so that was probably too high. It will be replaced by an LED. A backup light LED will be installed in the B unit, connected to the Blu Shark by tether. Marker light LEDs have their mounting challenges, but I'll look into adding them.
The bottom line is that, for medium-sized, twin-motored diesels, the Blunami controller card appears to be a worthy alternative to Lionel's TMCC and the MTH DCS, as long as one is willing to accept a smartphone or tablet as the controller surface. I'm very pleased with the performance of my Blu Shark and look forward to implementing this scheme on an Atlas EMD switcher and (ancient) Weaver FA/FB pair. Passenger diesels are another matter for future investigation - likely next fall/winter.