Bob:
Like Mike said, a voltage rectifier is required. Otherwise you will be feeding the capacitor AC which it is not designed for. They are not expensive. These guys are quite small but still easy to work with. The pins are marked but you might need magnification to read them:
http://www.digikey.com/product-search/en?keywords=MB12S-TPMSCT-ND
Dave
I'm just a dude with a bad back having a lot of fun with model trains, and finally building a layout!
Basically. Look at the previous diagrams. Note that the voltage regulator is also required to provide + and - power. Caps are typically polarized, so connect the cap accordingly, + to + and - to -. The cap will keep supplying power when the DCC wave reverses many times a second, doing away with the flicker.
Mike Lehman
Urbana, IL
mlehmanIf you add the caps that further dampens things and will take the flicker out, as it will feed the LEDs with the correct polarity when the circuit reverses each time.
Is that all I would need is adding a capacitor?
-Bob
Life is what happens while you are making other plans!
It could also be that, without any caps, the car loses power for a second which causes the flicker. I know none of my $80 Superliners have caps in them, and it's really obnoxious for such a premium product.
Bob,
The 3-unit LED segments are rated at 12v pretty universally (and similarly any unit of 3 x however many you're likely to squeeze in in most scales), so that's why they work for you like that.
What's going on is that the LEDs are seeing the current constantly reversing, so they are actually flickering on and off, since LED require they be fed with a specific polarity and they don't light when they see reversed polarity. What explains why you see the flicker only sometimes is that it's happening so fast your eye typically won't perceive it, something like how your eye reacts to a movie as a moving picture.
It's possible that you could exceed the tolerance for reverse current that all diodes, inc. LEDs, have as a limit with such an unregualted set-up, but then again, what works, works. Adding the resistors helps limit that possibility. If you add the caps that further dampens things and will take the flicker out, as it will feed the LEDs with the correct polarity when the circuit reverses each time.
I am not sure what the difference is in my lighting for passenger cars on my DCC layout, but I installed a 6 and 9 LED strip in the Walther's cars and just plugged them into the contacts at the end of the car (which is connected to the wheel pickups). I had to add some resistors finally because the light level was much too high. I now use 2.4k resistors. I don't have any caps or diodes, but the lights do flicker occasionally.
Mike and Randy:
Thanks for the advice on the caps.
hon30critterFWIW, I'm not going to replace the 5 volt caps that I used in my caboose lighting circuits. Mark R. has been using them for 10+ years with no problems so I'm not worried about them exploding any time soon, if ever.
What works, works. I wouldn't take any model apart for something like that. It's not a threat to anything but itself and even then a failure seems very unlikely as your experience has shown. It's not something I'd worry about either.
25V caps should be fine for HO/N DC or DCC. DC may be more an issue than DCC - some DCC packs use some pretty high voltage pulses for pulse power. Way back when, I put a circuit in a loco for a fellow club member that relied on this to get the ditch lights flashing before the loco was moving. It worked on one main, controlled by an MRC CM20, but did not work on the other main, controlled by a Hogger. (or maybe that was the other way around - it was over 20 years ago). The pulses were present even when the DC component was stiull not enough to turn over the motor, but a capacitor circuit charges to the peak and I'm pretty sure it was around if not over 25V. DCC, no problem, HO track voltages are typically around 15V, so a 25V cap is plenty there.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
The 2200 uF 25 volt caps are about 1/2" x 1" so they are not enormous. I have 10 on hand so I will use those before buying anything else. I can't remember what I intended to use them for but I'm sure I had a dire need for them at the time. You should see how much stuff that I thought I needed but haven't used yet! I think I'm too much in love with DigiKey's fast and cheap delivery system.
FWIW, I'm not going to replace the 5 volt caps that I used in my caboose lighting circuits. Mark R. has been using them for 10+ years with no problems so I'm not worried about them exploding any time soon, if ever.
Dave,
Shouldn't be unless you're like some folks and set you command station as if you're running G scale, jacking things to 18 v or something; wouldn't be much head room there. The only 25 v cap I have on hand is what I think came along as supplied by Soundtraxx with the Tsunamis. If it's good enough for Durango, it should be good on the track for other things at standard DCC voltage settings.
2200 uF, though? Way overkill, unless it's already paid for. I'm curious, how big is it? You could do like the WiThrottle folks did: put a power button on it and a bright LED in the end and the car could double as a flashlight
mlehmanTypical run time with the 470 uF is at least a minute.
Would there be a problem using a 25 volt 2200uF capacitor? I have some on hand.
Thanks
Mark,
No, they're really not. I've experimented with Photoshop some, as well as different camera settings on my Canon SX100 and can suppress it or color-shift it to a certain degree. It's probably something I'll get back to eventually, but I suspect there's really no simple fix.
The human eye and brain interpret things differently than the camera does. There seems to be difficulty in enough ambient light without the blue lighting for both human operators and the camera. I know some find it off-putting, but there seems to be no ready substitute that looks closer to "natural" that doesn't have even more issues.
The blue wavelength does vary depending on the LED source. In addition to the overall ambience, I also use the blue to activate my "stars" painted on the backdrop and a few other feature that depend on the UV part of their emissions.
Given I grew up in the 70s, the UV doesn't bother me, in fact has a certain familiar coziness to it that I actually enjoy. Similarly, for that same reason it may be why it irritates some
Mike - Not meaning to veer off topic, but I've always been curious ....
Are the blue lights on your layout as vivid in person as they are in the pictures ?
Mark.
¡ uʍop ǝpısdn sı ǝɹnʇɐuƃıs ʎɯ 'dlǝɥ
Most of the caps are 35 v, but I may have a few that are 25 v. Never had any trouble with them in this circuit when preperly installed. IIRC, I did get the cap polarity reversed on one and I think I had a FAIL there, but that one's on me.
Length of run depends on how many LEDs in the car. Typical run time with the 470 uF is at least a minute. For anti-flicker, I've used as low as 47 uF successfully, which are smaller, but there's shorter run time. I know I have a few cars that must run for 5 minutes or so, because I'll get them off the track to get some maintenance and by the time I get them to the workbench several minutes later they're still glowing.
I've run as many as 4 x 3-LED segments in 80' cars, plus a pair of markers, although the markers typically get less resistance off the same VR/cap combination.
Short NG cars typically have room for just 2 segments, or even just one like this install in a brass Key San Juan parlor car.
In this case, there's also a lighted drumhead.
Mike:
I found your diagram a few days ago, but thanks for posting it again.
A few questions if you will:
What's the voltage on the capacitor?
How long do the lights stay on?
How many LEDs can you run off of the circuit?
Found a copy of my circuit plan done as a crude graphic and some more handsome older stuff...
Here's a Before shot of the BLI CZ dome-obs.
And After it was dimmed.
This is a typical install in a heavyweight sleeper.
Component failure seems less likely an issue than potentially shorting the 12+v voltage regulator feed to something bare in the circuit that should not see more than 5 v. An install in brass or other metal car has special potential for this sort of failure. Good install practices can limit it from happening, but between that and test probing things can still happen.
Assuming failure typically occurs inside a closed car, at least any failure shrapnel is contained
I wouldn't worry about that circuit if I had it. It works. But it's the sort of thing that's mostly needed because the super-caps can be squeezed into places you can't put the big klutzy components like my 35v caps. You can stuff a lot inside a HOn3 bathroom, though, and the components I use are probably cheaper than super caps (or have they come down enough where that doesn't matter?)
hon30critter Hi Mark: I guess the real question is: Will the 5 volt capacitor blow up if either the resistor or the LEDs fail? Thanks Dave
Hi Mark:
I guess the real question is: Will the 5 volt capacitor blow up if either the resistor or the LEDs fail?
Of course it will. No different than any other circuit if a component fails - something is going to give .... I've seen two decoders literally catch on fire due to a component failure.
hon30critter Hi Mark: I decided to err on the side of caution and withdraw the circuit at least for now. I hope you are not offended. I don't want to be the source of potential problems. Maybe I'm being paranoid. CSX Robert's caution about using a 5 volt capacitor in a circuit that is being fed 12+ volts seems to make sense. If a higher voltage cap can be used then perhaps that's the best way to go. For now, I'm going to sit back and see what others, including you, have to say. Regards, Dave
I decided to err on the side of caution and withdraw the circuit at least for now. I hope you are not offended. I don't want to be the source of potential problems. Maybe I'm being paranoid.
CSX Robert's caution about using a 5 volt capacitor in a circuit that is being fed 12+ volts seems to make sense. If a higher voltage cap can be used then perhaps that's the best way to go.
For now, I'm going to sit back and see what others, including you, have to say.
Regards,
As simple as my circuit is (was), it is no different than any other circuit .... all the components are configured to work properly with each other. Alter one component or delete one component, it either won't work or it will destroy itself.
No different than if the coil of a speaker goes bad and shorts itself (the odds of which are about as good as an LED burning out), you'll blow the audio amp on your sound decoder.
If you play those kind of odds as a deterent, you probably shouldn't drive your car either as you might blow a tire and get into an accident !
Mark R. My caboose lights stay on for well over one minute.
FWIW, I had the same results with Mark's circuit.
CSX Robert hon30critter Hi RicZ: I have used this circuit successfully in a fleet of cabooses. It has a stay alive capacitor in it so the lights don't flicker if the power is momentarily interupted: ... Note that this circuit uses three separate LEDs. It will not work with a 12 volt 3 LED strip. Circuit designed by Mark R. Dave What I don't like about this circuit is it uses the voltage drop of the LED's to limit the voltage to the super cap. What this means is if the LED's all fail, or you have a bad connection and it comes loose, the cap will receive full track voltage (minus the diode drops through the rectifier), not good for a 5 volt super cap. Also, you're not making the best use of the cap because it will never charge above the voltage drop of the LED's, so you will get very little keep-alive capability. I would think you would get better results from a similar size regular cap charged to track voltage, even though the capacitance rating would be much smaller.
hon30critter Hi RicZ: I have used this circuit successfully in a fleet of cabooses. It has a stay alive capacitor in it so the lights don't flicker if the power is momentarily interupted: ... Note that this circuit uses three separate LEDs. It will not work with a 12 volt 3 LED strip. Circuit designed by Mark R. Dave
Hi RicZ:
I have used this circuit successfully in a fleet of cabooses. It has a stay alive capacitor in it so the lights don't flicker if the power is momentarily interupted:
...
Note that this circuit uses three separate LEDs. It will not work with a 12 volt 3 LED strip.
Circuit designed by Mark R.
What I don't like about this circuit is it uses the voltage drop of the LED's to limit the voltage to the super cap. What this means is if the LED's all fail, or you have a bad connection and it comes loose, the cap will receive full track voltage (minus the diode drops through the rectifier), not good for a 5 volt super cap. Also, you're not making the best use of the cap because it will never charge above the voltage drop of the LED's, so you will get very little keep-alive capability. I would think you would get better results from a similar size regular cap charged to track voltage, even though the capacitance rating would be much smaller.
Have you tried this circuit ? Dave is successfully using it and I have as well for probably 10 years or more. You say there will be very little keep alive capacity .... that tells me you HAVEN'T actually tried it. My caboose lights stay on for well over one minute.
hon30critter Hi CSX Robert: I'm no electronics expert but I do understand your point about the potential problem with the 5 volt cap. I'll put a caveat in my post. Thanks Dave
Hi CSX Robert:
I'm no electronics expert but I do understand your point about the potential problem with the 5 volt cap. I'll put a caveat in my post.
Ummm .... I wouldn't be so sure. The single resistor is factored to the voltage required by a single LED (in parallel, the voltage remains constant) which is well below the 5 volt rating of the super cap.
Three LEDs in series will require more voltage and hence a lower value resistor which will feed way more than the rated voltage to that 5 volt cap.
Be sure to shield yourself when you bench test it.
I'm no electronics expert but I do understand your point about the potential problem with the 5 volt cap. I just deleted the circuit from my post.
CSX RobertWhat I don't like about this circuit is it uses the voltage drop of the LED's to limit the voltage to the super cap. What this means is if the LED's all fail, or you have a bad connection and it comes loose, the cap will receive full track voltage (minus the diode drops through the rectifier), not good for a 5 volt super cap. Also, you're not making the best use of the cap because it will never charge above the voltage drop of the LED's, so you will get very little keep-alive capability. I would think you would get better results from a similar size regular cap charged to track voltage, even though the capacitance rating would be much smaller.
Yeah, that's the main difference between the two circuits, besides my using the output to feed the strip light LED segments. I typically use a standard electrolytic cap that's rated for 470 uF at 35v. Plenty of voltage overhead and the lights will stay on for a minute or more, depending on you resistor and how many LED segments.
Yes, I prefer a belt and suspenders approach and would use a regulator circuit, or, even better, since there should be plenty of room inside a passenger car, use multiple supercaps like used in the stay-alives for decoders such that the cap combo can handle full track voltage and then you wouldn't have to worry about it.