"Keep Alive" Capacitor Installation

[peteski]

If you are looking for another decoder, I already mentioned Digitrax DZ126T. I think it is a bit smaller than TCS Z2 and the street price is around $20 (at M B Klein).  Have you tried to increase the load on the decoder output yet by hooking up a 100 ohm resistor in parallel (not series) with the misbehaving (12V) motor?

EDIT:
I just had a good laugh!

This thread was started by BCR751 (Doug) with no specific problem or model being mentioned.  Asking for keep-alive info.  He has not posted any further info/questions/answers.

But then the thread was taken over by randgust's problem with a decoder mounted in a Kato critter.  At this point I don't even thing that keep-alive thing is even part of the equation.  We are fighting totally different fire here.

Re-reading the thread I am no not sure anymore: Is the decoder-equipped critter having problems running smoothly in DCC or under DC control?

[randgust]

There's apparently no head-banging emoticon.   I'll summarize....

What got me here was an entire year of fighting a TCS Z2 installation in a Kato critter - custom-built for a client that insisted on DCC.

Built and tested chassis and locomotive on DC, ran great.  Converted to DCC, sent to Lee Weldon for testing and programming - failed.
Came back to me.  TCS replaced the decoder under warranty.   Went back to Lee a second time - failed, due to the low-speed surging.
Involved TCS again, support suggested a lot of things to try including the BEMF CV settings.
Came back to me.  Extensive testing.  Went to impatient customer - he tried extensive reprogramming of all the CV's with no luck. 
Refused payment and asked for a refund, got it from me.   Got it back.
More extensive testing in DCC by myself using a Digitrax Zephyr and correspondence with TCS.   Exhausted everything, they finally decided it was just 'too light', also recommended a KEEP ALIVE in the locomotive so that the signal interruption would be minimized, only nothing was small enough to fit.

Abandoned project entirely, admitted failure, removed decoder, went back to DC mode.   Locomotive again performed excellent.  Resold locomotive to another customer, he's very happy, no problems.  Decoder now sitting in scrap box.   Lots of other critter chassis available for testing.

So yeah, this does belong here when TCS recommended I try a 'keep alive' yet there aren't any small enough to fit in it, at least then.  I think the Z2 simply doesn't work and I'm ready to restart with a different decoder, and in this thread I've had a lot of excellent suggestions.   I think it is flickering in and out of DC/DCC analog mode and a 'keep alive' would help, but I got so frustrated with TCS I'm more than ready to try something else instead.    When I was looking at the keep alive function I'd have to tow around a 40' boxcar just to hold it.   If there's a tiny substitute 'keep alive' home-brew solution I'm just as interested in that.   No thread drift.

[jdcolombo]

Well, to get back on the keep-alive track, the 220uf, 16v tantalum chip caps I use are 7.3mm long, by 4.3mm wide, by 3-4 (depends on manufacturer)mm high.  It looks to me like there is space in the cab of the critter to put at least two of these wired in parallel; could be room for a 4-pack.   Two wired in parallel would certainly end momentary drop-out problems; 4 would probably give a couple of seconds of running time without power, depending on the motor current draw.

The main problem would be wiring a keep-alive to a standard (non-sound) decoder.   Lenz has pads on its Gold and Silver mini series for doing this; I'm not sure about Zimo or ESU (the LokSound has pads for a keep-alive; not sure about the LokPilot).  The theory of the wiring is pretty simple - the positive side of the cap gets wired to the blue wire (which is the positive common supply for the decoder); it's the negative side that sometimes creates issues, because it needs to be wired into the decoder circuit between the negative side of the voltage rectifier and the actual power-consuming (motor and digital electronics) circuits.   The links that I provided in my original message show some wiring possibilities if a separate negative pad for a keep-alive isn't provided on the decoder, but I can't imagine trying to solder something like this on a decoder the size of a Zimo or ESU LokPilot.   

But again - I strongly suspect that if you use a different decoder, you won't see the problems you saw with the Z2 and the "keep alive" issue will be moot.  I will be interested to see if I'm right (perhaps I'm unfairly tarring the Z2 with poor performance in this application).   If I'm wrong, well, it won't be the first or last time!

John C. 

[peteski]

I have some Z2s and critters - I might have to investigate this some day.

I have installed this decoder in Atlas 4-4-0 and didn't notice any problems. Roger Holmes (member here) has a similar model.  These are tiny locomotives with very small motors (a-la the critter).

As I mentioned, there are 2 keep alive circuit locations on the decoders.  Even if there was a room for it, even a 500 uF cap installed on the primary keep alive side (where it would supply power to both motor and the microprocessor secondary supply) would really make much difference.  But that is something that could be easily tested and verified.

I have the necessary parts to do it - it is the time and amount of in-progress projects which will hold me back form doing such test anytime soon.