Zimo MS500 in a LL SW-9

[RandyLeclair]

I'm in the midst of putting a Zimo MS500 sound decoder in a LL SW-9.  I'm looking into putting a couple keep-alive caps in there too, but does anyone have experience with this board to tell whether it needs some or a LOT of caps?  There's not a ton of room in there for much of anything and I'm trying to save as much frame as I can...

-Randy Leclair

[jdcolombo]

I don't have much experience with Zimo decoders, but my general experience with LokSound decoders that have no on-board keep alive is that 440uf is the minimum necessary to avoid sound dropouts.  This amount won't keep your motor running over dirty spots in the track; it assumes clean track and wheels, and is just enough to avoid sound dropouts and resets when there is a momentary loss of electrical contact, which happens from time to time even with clean rail and clean wheels.

More is always better.

John C.

[tehachapifan]

Glad to see this! I was just measuring the space of a LL switcher frame a couple days ago and realized the MS500 might fit up front, if the upward frame protrusions were ground down. I came to the conclusion it would just fit, but without any extra room at all. On yours, it appears you're keeping the stock lightboard too (I think I see it sitting atop the decoder?). Curious as to if the shell will seat all the way down with the lightboard there. Also, with the upward frame protrusions gone, the slots for the stock lightboard would have also been removed. Where and how did you route track power to the decoder?

[peteski]

In theory, any DCC decoder (unless otherwise stated by the manufacturer) does not require any additional capacitors to operate properly.  But having said that, especially in N scale where the models are relatively light with short wheelbase, electrical pickup is  prone to short drop-outs (even with clean wheels and track), so adding some external caps does improve reliability.

John,  ESU decoders do have built in what we call "keep-alive" capacitors.  Those 4 or 5 large multilayer ceramic caps are it.  While I have not unsoldered one (because they are fragile) to test it, going by their physical size, they are likely 47uF each.  So the on-board capacitance is around 200-250 uF.

By "keep-alive" I mean capacitors connected on into the raw rectified DCC voltage on the decoder.  All decoders also have smaller "keep-alive" caps connected on the output of the voltage regulator which supply reduced voltage (around 3V) power to the microcontroller section of the decoder.

I also just purchased some of the Zimo MS500s (along with ESY V5 nanos) to experiment with.  These MS500s sure give ESU nanos run for the money.  They are very small, and thinner than the bloated ESU V5 nanos.

Zimo decoders are in the same class as ESU as far as motor control and function effects go.  They do not have the same extensive customization capability that ESU has, but that for many modelers is over complicated to even touch. Sometimes too much of a good thing is just not needed.  Zimo also does not have the extensive selection of American sound projects that ESU has, and some projects are not  free, but overall I would put Zimo on par with ESU.

[jdcolombo]

Hi Pete.

Was that true (e.g., keep alive) of the old LokSound Select Micro?  It did have 3 tiny yellow caps on one side of the board toward one end, but I didn't know those were "keep alive" caps.  They sure didn't help dropouts any! The Loksound 5 DCC micro seems a bit more tolerant of momentary electrical loss, but these days I'm mostly working on steam locos, and I put a full-fledged commercial keep alive (the TCS KA-1) in those.

John C.

[RandyLeclair]

Thanks to all who replied.  John, I have a supply of 220 uf caps; should I wire a couple in series or parallel?  20v tantalums.  OR are there better options?  I really don't want to melt shells

Best -

Randy

[peteski]

Hi Pete.

Was that true (e.g., keep alive) of the old LokSound Select Micro?  It did have 3 tiny yellow caps on one side of the board toward one end, but I didn't know those were "keep alive" caps.  They sure didn't help dropouts any! The Loksound 5 DCC micro seems a bit more tolerant of momentary electrical loss, but these days I'm mostly working on steam locos, and I put a full-fledged commercial keep alive (the TCS KA-1) in those.

John C.

John,
I never bought one of those V5 micros because of the Next18 connector, and that awkward dongle for wired-in installs.



Looking at the photo, those 3 yellow/brown caps do look like what we call "keep alive" caps. If they are 47uF each, that is around 150uF total.

With that caps we add onto our N scale trains I hesitate to use the term "keep-alive".  Basically the caps already on the decoder board connected at the output of the bridge rectifier on sound decoders are filter capacitors. That is a standard design in DC power supplies.  Those caps are there to smooth out the rectified sine-wave in AC supplies, or in DCC, the spikes in the square-wave DCC track power.  They aren't really there to provide anything else but just to smooth out the DC voltage coming out of the rectifier.

But at some point someone figured out that if the capacitance of those filter caps was increased, they would  have enough power reserve to keep supplying power to the decoder circuits during very brief power interruptions.  That was a welcome feature with sound decoders (to prevent the electronic circuit from resetting on such brief  power drops).

Then, as larger and larger capacity compact-size capacitors (Super Caps) became easily accessible and inexpensive, someone figured out that increasing the capacitance of those filter  caps from hundreds or thousands uF to hundreds of thousands of uF those filter caps can provide enough energy to power the decoder, motor and functions for seconds, easily traveling through longer-duration track power dropouts. But those large capacitance caps needed additional circuitry to limit their charging current.  DCC manufacturers started producing add-on circuit boards with the capacitors and the current limiting circuitry to tack them onto the on-decoder filter caps.  Those come with different names, depending in the manufacturers. Some call  them "keep-alive"  others call them "power pack".

With the caps we add to our decoders in N scale (with the total capacitance of <1000uF) I have hard time calling it  keep-alive.  It is more like "large filter caps". To me a true keep-alive circuit should have enough capacitance to supply power to the entire model for at least a second.  But the name stuck, so I guess we will call those few add-on caps we cram into our N scale models "keep-alives".

[peteski]

Thanks to all who replied.  John, I have a supply of 220 uf caps; should I wire a couple in series or parallel?  20v tantalums.  OR are there better options?  I really don't want to melt shells

Best -

Randy

You need to connect them in parallel to get the sum of the capacitance of all the connected caps.  20V rating is IMO barely adequate. I would rather use tantalum caps rated for 25V (for the extra safety margin).
I also highly recommend a burn-in procedure whenever using tantalum caps to prevent any shell meltdowns.  The info is a bit scattered around the forum.

First, read this post and the remainder of that thread, then this entire thread.  Also read this post and the remainder of that thread.

[RandyLeclair]

Where and how did you route track power to the decoder?

I drilled a #60 hole in each frame half, used a bit of nokorode flux in there, and attached the leads with a dab of solder from a very hot gun.  I put a bit of heat into the frame for good adhesion (in other words, I left the tip on the frame for 10 seconds or so with my cheapie 22w Weller, which had been plugged in for over 5 minutes).

Hope this helps.

-R

[RandyLeclair]

You need to connect them in parallel to get the sum of the capacitance of all the connected caps.  20V rating is IMO barely adequate. I would rather use tantalum caps rated for 25V (for the extra safety margin).
I also highly recommend a burn-in procedure whenever using tantalum caps to prevent any shell meltdowns.  The info is a bit scattered around the forum.

AVX is in my backyard (they're two towns over), so I guess I'll root for my home-team tantalums, and bump up some on the voltage.  No flaming caps, please!  Thanks for the posts.

-R

[conrad]

Randy,

Looks like it will be a great install. 

Are you using Oliver Zoffi's (ZIMO) free EMD NW2 sound file or ZIMO's free EMD 12-567 sound file?

Conrad

[peteski]

AVX is in my backyard (they're two towns over), so I guess I'll root for my home-team tantalums, and bump up some on the voltage.  No flaming caps, please!  Thanks for the posts.

-R

You're welcome -- just trying to help in preventing a disaster.  The posts I linked to mention that the polymer tantalum caps are less prone to "spectacular thermal failures" than standard tantalum caps.  I'm thinking of going that route in my next capacitor install.

[tehachapifan]

Randy,

Looks like it will be a great install. 

Are you using Oliver Zoffi's (ZIMO) free EMD NW2 sound file or ZIMO's free EMD 12-567 sound file?

Conrad

FWIW, I used the 12-567 file (not the OZ NW2 file) in a Kato NW2 install and absolutely love it! It easily rivals any of my ESU files. The horn is set to a preset length per blast, but I learned a way to modify that so now I can have a short blast for a nice grade crossing sequence. I was a little less than impressed with the OZ NW2 sound sample and took a risk on ordering the 12-567 file (which has no sample) and was not disappointed in the least. Great sound file!

[RandyLeclair]

12-567 sounds pretty good to my ears.  How did you mod the horn?  That's about the only flaw I've heard so far.

-R

[nstars]

12-567 sounds pretty good to my ears.  How did you mod the horn?  That's about the only flaw I've heard so far.

-R

I’m also looking into some Zimo decoders and changing the sound is in some cases quite easy. Just download the full featured Zip file and use the Zimo ZSP software to adapt the sound. Right clicking on the horn sound file gives you the option to choose the section of the file which will be the loop section. Check the loop box should make this possible.

Concerning stay alive I would recommend to read the Zimo decoder manual as Zimo has a good description how to add a stay alive to almost all decoders.

Marc