I’m a long time electronics guy but I have a question for the Forum. I have struggled for years with my layouts non track related power supplies. Over the years I have purchased and built all kinds of accessories with varying operating voltages. At one time I had more wall-warts than I had power strips to plug them into.
I decided to go to one 12 volt power supply with several regulators, early on that worked OK but again as time flew by the loading increased rapidly on the regulators creating heat. Then it was fans time to get rid of the heat.
Regulators are just too inefficient at high current loading.
I have the following power requirements:
1.4 volts at 5 amps (1½ volt micro bulb vehicle headlights)
4.5 volts at 1 amp (animated signs)
5 volts at 5 amps (Arduinos/many animated servo or motor driven g
It strikes me that the power supply from a desktop computer may be able to provide what you need. Depending on the wattage of the power supply used, you’ll be able to provide enough current for all your accessories. They also have feeds at multiple voltages. You’d be able to use the 3.3v, 5v and 12v outputs, then regulate those down to some of the other voltages as needed.
I took this approach when I needed a power supply for my layout lighting. Here’s how I did it, if you’re interested: http://bvlcorr.tumblr.com/post/135005793372/a-power-full-lesson-part-2. I only used the 12v feed for the lighting, as I already had my accessory buses set up and working. I also needed all the desktop power supply’s amperage for the lights.
My “other” hobby is playing with CCTV cameras. I have found some of the camera power supplies to be handy to use for auxiliary layout power. The voltage is stable and each output is protected. Some have fuses, the ones I have are auto-reset devices.
My layout has “evolved” over the years so I have several about a dozen supplies for lighting and signals. An old MRC Controlmaster 20 provides the bulk of the incandescent lighting. I use the variable DC out so I can dial down the voltage. When visitors are present I bump it up a little.
I have some of these volt/amp meters on my outputs, too:
30-35 amps? I prefer to have multiple low-amperage circuits rather than a big central source. My last incandescent lighting project was my “Union Station” platforms and the current draw for those alone was 7 ½ amps!
I used to go through my seemingly limitless supply of chargers from old phones. cordless drills and the like. I got tired of it.
Now, I have 12 VDC supplies set up for each section of my layout. They drive a bus for that section. Each supply has a toggle switch on the control panel, and a fuse to prevent burning out the supply on a short or overload.
This gives me 12 volts. For lower voltages, like Miller animated signs, I measure the current with a battery and a meter and compute the resistance needed to step down my 12 volts to what the device wants.
These supplies, at 4 or 5 amps, cost a few dollars including shipping from China. I’ve had no problems with them at all.
I’m sort of in the same boat. I don;t want to run a whole bunch of additional power busses around my layout for all these different voltages. I want one power feed with taps and circuits to get the actual voltage needed on site. Sort of like neighborhood electrical power. If you don’t mind all the extra wires, a PC power supply would be a good option because you would get a 12V bus, a 5V bus, and a 3.3V bus, which would allow fairly efficient drops to other similar levels via linear regulators.
I bought a 5 pack of some cheap buck converters, they seem to work ok. With these I can run maybe a 15V line and tap off as needed with buck converters set for the voltages I need. The biggest load will probbaly be the servos for my turnouts. I don’t want them to be on the same power drop as the Arduinos that will control them, servos are rather electrically noisy.
The buck boost convertors are looking pretty good. I’ve had a Dale 50 watt .25 Ω resistor as a load on one DC to DC convertors at adjusted to 1.45 volts (5.8 amps) for about 8 hours and the regulation holds very good switching between 100ma and 5 amps (.04 volts). The 12 volt switching power supply has remained cool to the touch and the DC to DC convertor is only slightly warm to the touch after hours running at 5.8 amps.
I’m in the process of loading a second convertor to about 5 amps at 8.5 volts testing it for stability and heat dissipation.
I was going to use one 5 volt supply but I’m going to be going the servo route also so maybe I need dual 5 volt convertors too.
I’m a little leery of the buck boost convertors and I was hoping to get some input from someone using them. For adjustable voltage, current, size, heat dissipation and price they can’t be beat.
After many hours of testing the cheapo eBay Buck Boost Convertors (under $5 for 12 amps) they passed. I have decided to go with them. They are very stable and hold voltage regulation to .05 volts from no load to full load. With a fixed load the output voltage held to .02 volts over an 8 hour period. I have three convertors and all three measured identical in my tests.
I also did a lot of research on MOSFETs with the following results (each convertor uses two MOSFET transistors).
From a MOSFET Manufacturers web site:
MOSFETs usually fail short circuit, causing the opposing device (the second MOSFET) to fail as well, (also short-circuit.) This short-circuits the supply, so current limiting or suitable fuse protection should be employed to prevent damage to the supply. When MOSFETs fail normally all 3 leads short together, that would permit the convertor to pass the supply voltage to the output of the convertor.
The ones I got are much lower power - 3A max. But I also got 5 of them for under $7. All I’ve done so far is verify they put out a stable voltage with a varying input - I do want to put my scope on them under load and check how much garbage makes it through tot he powered circuit, and how much it feeds back into the input side. For my CTC nodes, 3A is going to be plenty, and there will be a seperate one for servos. For 2 servos I should be good, for larger plants I will probably need a higher current model.
I did some more testing on my three modules, all three are very close. The switching is just under 300k and I don’t see any switching ripple but there is from 50 to 60 microvolt’s of grass/noise on the output. A 100 ųf 25 volt cap reduced the noise about 10 microvolt’s.
The noise varies slightly with load. I ran three UNOs on one convertor as a actual operating test without the cap and I didn’t see any problems. One UNO is programmed as a servo driver with two SG90s the other two are 14 port random light drivers.
The random light drivers do not get their lamp voltage from the 5 volt convertor, I’m using 7 channel 500ma driver chips from a separate power supply for lamp voltage.
My scope is like me, ancient (it’s almost 50 years old) Tektronix 321A. It still works as good as it did when I got it in the late 60s.
I remember using those at NASA in the 90’s along with stacked wire (Nixie tubes) voltage/current displays. Everything was reused there. Rarely was anything thrown out. Those were the good days.
The 321 is a great scope unless you need bandwidth. They weren’t worth much on battery operation unless you carried a crate of D cells with you. NiCds don’t work very long at all starting off at 1.28 volts per cell. The scope gets flaky at 1.25 volts per cell. Works great on 120 AC.
Mel
Modeling the early to mid 1950s SP in HO scale since 1951
