I have a couple of questions regarding the use of capacitors in sound decoder installations. What does the capacitor between the sound decoder and speaker actually do? And I’ve noticed that most of the provided capacitors are the non-polarised electrolytic type - is it possible to substitute these with a space-friendly (e.g. ceramic disc) capacitor of equal value?
If your are talking about SoundTraxx, the older DSD types require a 33ufd bi-polar capacitor to keep DC from getting to the speaker. Don’t substitute a smaller value. You will lose lower frequencies. This has been discussed before.
The Tsunami do not require the bi-polar capacitor but have a polarized stay alive 250ufd capacitor which is a totally different issue. Many new decoder users frequently confuse the two issues.
Take time to download and read the SoundTraxx documents.
If the capacitor is in series with one speaker wire, then it is keeping the DC out of the speaker. A few people have increased the value of the capacitor for lower frequencies but the physical size is a little larger.
Working with electronics over the years, I have seen solid state audio amps use non polarized capacitors between the IC and speaker but polarity had to be observed. I have built some IC powered audio amps and used non polarized caps. I suspect SoundTraxx went this route as some people would not understand the polarity requirement. Non-polarized capacitors do not care, read fool proof.
If you have more questions about SoundTraxx products, there is a Yahoo SoundTraxx group that that I belong to that specializes in SoundTraxx products with two or three online DCC suppliers in the group.
I have an SFX0416, and, as I recall, it’s not on the speaker leads at all. The capacitor on the speaker leads is meant to filter out audio noise. But, this is a “cleaner” design. Here, the capacitor is to smooth out the power to the decoder. If there is a short “dropout” in power, quickly restored, then the capacitor helps the decoder continue smoothly without interruption. It’s like a flywheel, only for electric power rather than motor motion.
Take a look at how the capacitor is wired. Is it really in line with the speaker?
On my Soundbug the capacitor is not in series with the speaker. I’m pretty sure it’s a keep-alive capacitor, although without cutting off the shrink wrap of the decoder I can’t tell 100%. However, AJ designed these things right and they don’t need a special booster on the program track nor do they cause a current inrush problem in the event of a track short.
They are polarized on threh Digitrax decoders so they are used for energy storage (i.e. filtering) and not as a bandpass filter or DC blocking capacitor, which would require a nonpolarized capacitor.
I know that this is an older post, but I am still interested in the answer;
Can you use a ceramic capacitor instead of an electrolytic capacitor as a keep alive without compromising the keep alive effect? Provided it is the same capacity, like a couple of ceramic caps in parallel?
The smaller form factor of ceramic caps would make the install in N scale engines a bit easier than finding the room for an electrolytic capacitor.
I believe this is the keep alive cap. Do you realise this is a sound only decoder? If you want to use it you need a motor decoder already installed. Joe
I’ve ordered some SMD Tantalum caps pretty cheap on eBay to see how they compare to the “regular” electrolytic caps that come with the decoder. The difference in size is enough for N scale to look into this I’d say.
What value capacitors did you get? That’s the key to if they will work or not. There are two ratings for a capcitor, voltage and capacitance. The voltage is the maximum it can handle without failing - and capacitors have a tendency to fail spectacularly when the ratings are exceeds - as in, explode. The capacitence determines how much energy it cna store. This is in direct properotion to both the size and chemistry used.
A 1uF or even a 10uF tantalum capcitor is not going to provide much if any keep alive capability, it just doesn;t store enough energy. To get to the second or two range you need something much larger than tantalum capcitors usually are available in, in the neighborhood of 4700uF.
I’ve ordered 100uF / 16V SMD Tantalum caps, for 25 cents a piece.
I hope the 16V aren’t cutting it too close though. Might have to do some testing before I drop one into an engine. I did have caps blow up on me before. It’s… interesting…
That is close but should work. I use the NCE Power Cab and the voltage for my Tsunami is about 12.2 vdc at the stay alive cap.
Remember, the stay alive with the Tsunami is 220 ufd and is only for the sound. Not enough capacitance for the motor.
With HO, some go as high as 4,700 ufd.
Keeping track, wheels and wipers clean will make the biggest difference along with powered frogs. Cleaning the rail along the inside top of the rail is another little trick.
One decoder company is selling a keep alive module which I understand works. Cannot remember which forum I read some reports in.
Electrolytic caps use the minus lead of the cap as an indicator.
Tantalum caps use the positive lead as an indicator. Be careful.
A cap blowing up is usually because of reverse polarity unless the voltage is really quite high.
TCS has several sizes of keep-alive modules, plus decoders with them already built in. Marcus updated his web site to show using the TCS unit instead of the discrete capacitor.
16V should be fine unless you up the track voltage. With a nominal HO setting of 15V to the track, and the cap connects after the rectifier on the decoder, so it shouldn’t get more than about 13V
A large cap cn cause an inrush problem, howver if it gets hooked up with a resistor and diode like Marcus used to show, and as it’s shown in the Loksound user manuals, it generally alleviates the problem. The more the better, but in N scale that can be a problem.