Although Dave's original post references the ESU Loksound product, I was having issues with 4 TCS decoders with "keep alive" (KA22 and KA26) they simply stopped working as far as "keep alive" was concerned. I contacted TCS which of course asked me to send them in for repair or replacement, the tech mentioned that they have improved the product to address some keep alive issues.
Just an FYI to everyone, if you experience an issue like me, submit it to TCS for warranty and don't waste time trying to troubleshoot it, I should have done it a month ago but wasn't happy about unsoldering all the work I had done, and wanted to be sure it wasn't a bad solder/wire, etc.
"The true sign of intelligence is not knowledge but imagination."-Albert Einstein
http://gearedsteam.blogspot.com/
Some do have command timeouts where the decoder stops if it doesn;t see any DCC packets for a specified time interval. OOn Tsunami that's CV11, default is 0 which means it never times oout, but any non-zero value means if it doesn;t see packets addressed to it, it will shut down.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
A decoder will continue to do whatever its last command was when power is lost and the capacitor kicks in. It's still receiving power, just no more commands. If you were running at speed step 10 with the lights on and the signal is cut off, it will continue to run on speed step 10 with the lights on until the capacitor can no longer provide enough power to do so.
There is nothing in the decoder that will send a stop command if / when power is lost.
Mark.
¡ uʍop ǝpısdn sı ǝɹnʇɐuƃıs ʎɯ 'dlǝɥ
Geared Steam
I was lucky to stumble on that site as well. Another nice thing about it is that it seems to be fairly up-to-date
It certainly is frustrating to find dead-end links or information that is sorely behind the times!
I'm going to be reading more of the content there also.
Thanks, Ed
Ed
I have experienced the same thing at times, so I do run the track cleaning "extra" and seems to improve performance. I agree, good electrical pickup from the engine (multiple pickups) and clean track are still a requirement for Keep Alive to perform at it's best.
Thank you for that link, I only have a chance for a quick glance but I have bookmarked it, it looks to have some great information, and some unanswered questions, like, how long does it take for the caps to fully charge? it looks like several seconds to a minutes depending on track voltage and the capacitor value.
I'm discovering that stay-alive isn't the cure-all that many modelers, myself included, thought it was going to be.
Indeed it is helpful for short interruptions in track power but if the rail is dirty enough or power to the locomotive is otherwise compromised your engine will still stall.
I had just finished an install using a TCS KAT-22 in a Proto-2000 0-8-0 with all wheel pickup. Some earlier models only collected current from the 8 drivers.
This engine runs smooth-as-silk. As I was making a break-in run at slow speed I encountered a spot where I had recently ballasted and must not have gotten the rail as clean as I thought.
As soon as the engine got into about 6" of this track (maybe two feet all together) it stopped dead while both headlights continued to glow for another 20-25 seconds. I have them programmed to be dim except in the direction of movement.
Therefore if the DCC signal gets interrupted or scrambled, the decoder still seems to think "stop" until another command comes along.
Maybe there is a way around this? I do keep my rails as clean as possible and I don't expect to have long stretches of "dead rail" like the example above but still... if no DCC command can get to the decoder all the caps you can tow aren't going to help.
I have not checked BEMF CVs so I do not know if turning it on or off will help.
There's a good source of KA information here:
https://sites.google.com/site/markgurries/home/decoders/keep-alive-compatibility
Regards, Ed
Found the issue, a cold solder joint on the front truck wiper, which also had the rear truck wiper wire, it ran okay, just no more keep alive, until now. I will replace the wires completely, as the insulation is brittle, redo the wipers and install a Loksound 4 with keep alive, and LED's.
Thanks again for the help Randy, sometimes it's the simple things.
rrinker For the ones that have had failures - what's your track voltage? --Randy
For the ones that have had failures - what's your track voltage?
Digitrax Zephyr Extra is putting out 13.6. Havent had a chance to try your suggestion.
Hi Randy,
The passenger compartment floor is pretty much fixed in place. I'd have to re-do the lower headlight wiring to get the floor out.
The freight box floor pops out quite easily. I'll have to do what you suggested and cut part of the weight away. Removing the weight won't be a problem because the railtruck weighs 8.25 oz as is, but it will be messy.
Thanks
Dave
Edit:
Last night I ordered some tiny slide switches from Digi-Key which should do the job.
I'm just a dude with a bad back having a lot of fun with model trains, and finally building a layout!
Is it easy to take the roof or body off the chassis? If so, just put the switch inside. Or connect the cap and the decoder with a micro connector, no switch needed. If it's hard to get back inside once it's all closed up - you maybe could make an oversize hole in the floor, and glue a thinner piece of material across the hole, drilled to fit the switch.
Just to keep those interested up to date, I added more weight to the railtruck and ran it some more on my test track. It now runs very reliably, but there is still no keep alive action even with the lights and sound off, and with the capacitor having been given about 10 minutes to fully charge.
I have the components for DIY keep alives (including 25 volt capacitors) so I think the next step will be to replace the Loksound Power Pack with a home made one. That won't be difficult, but I want to include a shut off switch and I can't figure out where to put it so it doesn't show. The freight car floor is about 3/8" thick with the lead sheet weights that have been added, and the cab is even thicker.
I had a capacitor blow up on me. I made my own stay alive utilizing the cap / resistor / diode combo, and was using it to just keep power applied to a head-end passenger car ( engine at opposite end ).
My track voltage is set to 13.4 volts, so I assumed a 16 volt cap would be sufficient. After assembly, I placed it on the track and all seemed to function just fine for the initial five minutes. I left the train room to retrieve something out of the work shop and heard a loud bang from back in the train room and knew exactly what it was .... the capacitor exploded. Yellow shmutz all over the interior that took forever to get cleaned up.
Rebuilt the same circuit with a 25 volt cap, and, two years later it's still running.
Don't understand why the 16 volts cap failed, other than the fact it was a surplus component I retrieved from an old power supply and may have already been stressed.
I've heard of more than a few people reporting of track voltages as high as 18 to 21 volts !!! That's not good for ANY DCC component.
It's possible, the supercaps used in these keep alives are usually no more than 5 volts, sometimes only 2.5 volt, so they wire them in series to handle closer to track voltage. This will evenly divide the voltage, assuming the ESR of each cap is approximately equal. Operating any cap over its rated voltage will definitely reduce the life.span. Too much over voltage and they will literally explode. I suppose if you stuff the keep alive in the boiler of a steam loco, it would be a a boiler explosion. It takes a fair amount of abuse to explode them though, feeding 17V into a set rated for 16V won't explode You aren't getting track voltage to the caps though, they come after the bridge rectifier on the decoder. Even with a filter cap, the RMS voltage of a square wave is the same as the peak, so if you have 15V on the rails, after the bridge you have 15V - about 1.2 volts for the 2 diode drops. So to exceed 16V to the caps you would need over 17V on the rails. Or if TCS means not to exceed 16V on the rails, then they are using a capacitor combination good for at least 14.8 volts. Most systems put somewhere between 12-15 volts to the rails in the HO setting (if they have one), use of a large scale setting if one exists for HO and smaller scales is definitely not something you want to do. But the old Atlas Commander, using the recommended power supply, put way to much voltage ont he rails, a few of the other 'basic' systems do as well.
For the ones that have had failures - what's your track voltage? Use either a cheap meter (that has no RMS option) or a REALLY high end one that can handle more than just sine wave AC. One that uses RMS calculations based on sine wave AC will probably read too high. I have two high end Flukes (but old - I paid less for them than a modern high end Fluke) that both say can return proper AC voltage for sine, square, and triangular wave forms, but I currently don;t have any of my DCC equipment out to test it with and compare to a cheapy no-RMS Harbor Freight meter. Or of course something like the RRampmeter will work, the voltmeter in that is designed to work with DCC signals.
I've read on different forums about people having stay alive modules quit working on them over periods of time - TCS mostly. I've noted on the TCS web site that they recommend a maximum track voltage of 16 volts.
I've seen many layouts where their track voltage is easily at or over the rating limit. Will capacitors break down (stress) if operated too close to their tolerance level, ultimately making them non-functional ?
Interesting Randy, thank you
I have two TCS that were working, now suddenly stopped. I've checked connections and nothing has really changed at all, one day they worked, the next day, nothing.
I'm about to try a Loksound for the first time with the same power pack in a brass shay, wish me luck.
I'll give this suggestion a try on the TCS as well.
Thanks Randy and Mark.
Randy, I'll give your suggestions a try when I have a new decoder to play with. I'm not willing to dig into an existing installed decoder's wiring. The quality of pick up was an issue at first but the railtruck will now run back and forth on my 6 ft test track with only very occassional stalling. The front truck pick ups are easily bent out of shape and the SPUD pick ups are rather flimsy so getting consistent power to the decoder is a challenge.
Mark, I have set CV 113 to 255 and there was no change. I will try it with the lights off and giving the capacitor more time to charge. I know the duration of the power supplied from the Power Pack keep alive is very short but I have used them in three other locomotives and they seem to do the job of dealing with momentary power loss quite well.
Regards
CV113 should be set to 255 for the maximum amount of stay alive. I've used a couple of the ESU versions and wasn't overly impressed with them - althought the single capacitor makes them great for tight spaces.
Try it with the sound and lights off. I could get maybe a couple seconds under motor power only, but turn on the sound and lights and there was very little extra time once lifted from the rails - even after allowing it to charge for ten minutes. Two different installs, same results both times.
Try this - power it from the track via clip leads right to the red and black wires. Avoid the wheels and pickup system entirely. Let it run for a little while, the disconnect one of the clip leads and see if it keeps going or stops instantly. If it keeps going as expected, the problem is that the pickup is so poor there's never time for the power pack to fully charge. If it still dies right away as if it wasn;t even connected, now you know it is the decoder or the power pack. Hook the power pack to a different Loksound and see if that works, that will tell you if it's a bad power pack or a bad decoder. Even if there's no room in a different loco to mount the power pack, just test at the bench. There's not much that can go wrong, it's just a bunch of supercaps with a control circuit (the 3rd wire). If the control circuit isn't working, it won't charge and/or feed back power during interruptions, but it could also be the decoder not signaling the control line. Swapping the power pack around will isolate it to one or the other.
Even if you had a capacitor meter, most common ones can't read values as high as those of super caps. And measuring capacitance doesn;t really do much. The sign of capacitor failure, other than physical (bulging, leaking electrolyte), is generally a change in the ESR (equivalent series resistance) which is a totally different type of meter than your typical multimeter. And not worth purchasing unless you are really into electronics and plan to troubleshoot and repair lots of equipment where capacitors may have gone bad. I don't even have one. You can buy several extra Power packs for the cost of a decent ESR meter.
Hi folks:
I'm working on an HO scale rail truck. It has a Loksound Micro decoder and Power Pack keep alive. It has eight wheel pick up. All the wheels are brass. It is using an older Tenshodo SPUD drive in the rear and a plastic truck in the front with Kadee coupler box springs for pick up. None of the parts were previously used. I'm using an NCE Power Cab.
The problem is that I can't seem to get the Power Pack to have any influence on the motor. I have checked the wiring for the right connections and conductivity more than once. I have played with the duration of the Power Pack's output (CV 113) to no effect. I have reset the decoder including lifting the unit off the tracks.
To give you a few details about what has been happening, I have had trouble with stalling which could be caused by pick up problems. I tested the railtruck before installing the Power Pack and then after. There was no difference in performance. The frequency of stalls is decreasing the more the railtruck is run, but there is no indication of any keep alive function. If I pull the plug on the Power Cab the railtruck quits immediately.
Is there a way to test the Power Pack independantly?
Any other suggestions?
Thanks as always,