"Keep Alive" Capacitor Installation

[randgust]

Victor, I think so far I've built.... maybe 75 Kato 11-105 chassis-based locomotives for customers since I introduced the kit in 2006, sold as built-ups in various configurations.   Including one for Mr. Kato's personal inspection.

This was the first one I ever attempted to put a decoder in myself.   I know the genius of the pickup system and the DON'T MESS WITH IT rule, as it's as good as it gets.   Oh, and I do two things - polish the pickup truck treads with a felt wheel in a dremel, and use Conductalube on the end-axle cups.   They tend to run just about flawlessly.   My 'test track' for them includes a back-to-back crossover of Atlas C80 #6 switches with that long insulated frog that alternately insulates one truck on each side so it can't make it unless all 8 wheels are picking up properly.   It worked perfectly before I converted it, and after I converted it back.   I even have a YouTube video of it after removing the decoder.

Lee Weldon tinkered with setting about every CV there was (including the BEMF setting) with no particular success.

Hence my complete frustration with both this decoder, DCC, and TCS.   It's the major reason why I will not go to DCC - if I can't get my 'critters' to work as well as they do in DC, I'm not doing it.

But I am more that willing to try a different non TCS decoder, "Keep Alive" add-on, or whatever else is out there that will fit.   The "1.0" TCS decoder that a couple of my customers put in the Climax reportedly worked as intended.  The "improved" one I got seemed to have this slow-speed hiccupping issue on the original and the replacement, which is why I wanted their tech support to investigate.   I've been disgusted enough with the entire episode I removed all the DCC equipment and stored it, but I'll try again... 

[victor miranda]

Hi Randgust,

I am assuming you attached the decoder wires somewhere.
where did you pick?

I'd be real tempted to put a thin piece of PCB where the motor and flaps meet.
( my memory is that that chassis has no wires to the motor)

and solder the motor to one side of the pcb. the leads to the pickup on the other.
man, it has been a while.

I guess you could do much the same with a small amount of tape.

victor

[brill27mcb]

Perhaps "too small" may really mean "too little current draw."

Rich K.

[peteski]

There are plenty of small decoders out there. While I am totally disenchanted with the Digitrax command station and throttle design, I have no problem using their decoders for motor control (but their lighting effects with LEDs is another story in the older decoders).  Try DZ126T (T for tiny).  See how it handles the critter motor control.

[victor miranda]

Perhaps "too small" may really mean "too little current draw."

Rich K.

hmmmmm.

hi brill27mcb
I accept and agree that the tech most likely was thinking too little current draw.

please believe me when I say in this next part that I still agree,
we just have to take another step.

the Z2 decoder is intended to fit z scale locos and motors.
and the Kato 11-105 chassis is in no way special.
it is many electrical aspects like a z-scale diesel.

it should be able to control current down to 10 ma.
.... the decoder may be defective.

I got the impression that more than one decoder was giving similar results.

victor

[peteski]

The way I understand the DCC decoder motor drivers (with BEMF disabled) is that the load on the motor output will not matter.  The motor is driven with PWM pulses which are full 12V (or whatever the DCC track voltage is minus losses on the bridge rectifier and driver transistors).  The width of the pulses determines the average voltage seen by the motor.  If anything if the motor was very efficient (like a coreless motor in FEF-3 for example) then the loco woudl run too fast even at speed step 1.  It should not buck. It should also fully stop at speed step 0 (as there should be no 12V pulses present of any width).

If you thing that the motor does not put enough load on the decoder take a 100 ohm resistor and install it in parallel with the motor to increase the total load on the motor output of the decoder.  Since this is a short duration experiment and the resistor will not be enclosed, you can use a 1/4 or 1/2 W resistor (it will get hot). If you have a larger (2W resistor), that would be better. If the load on the decoder is the factor, this should make a difference.

EDIT: One more thing comes to mind:  I have some critter chassis but they aren't handy so I'm not sure if what I will mention applies. Do they have any caps and/or chokes (coils) in around the motor?  If there are any - remove them as they might be affecting the decoder's circuitry.

[victor miranda]

Forumers,
I took a little time to set a critter on rails.
it is an old 11-105 with the 100 ohm resistor

.... it will not do a tie crawling slow.
at about the slowest it will go it runs at 15ma and 4 volts
so while not a lot of power the decoder should handle it.

I may have seen what Randgust has as a problem.

while I was attempting slow on at least two ocasions the model oscillated
a little one way and then back. a total of maybe a 16th of an inch
until I gave it more power then it would take off.
I think a flywheel may be in order.

it may be a harmonic issue.

victor

[peteski]

Forumers,
I took a little time to set a critter on rails.
it is an old 11-105 with the 100 ohm resistor

The resistor should be in parallel (not in-series) with the motor, simply to increase the load on the decoder's motor output wasting it as heat  (since Tony thought the motor might create too light of a load).

[victor miranda]

it is an old first version.
 the resistor is there to save the motor from over current.

experiments may be on the morrow.... after sleeeep

[victor miranda]

I did a quick solder in and kapton tape job.
I put the tape between the motor tabs and the chassis thumbs
and soldered the wires.

the z2 decoder does not do DC well with this chassis.
the critter did the same bouncing/oscillating I described.
until I got to 3/4 throttle and then it zoomed.

so I tried DCC

well... it will not creep.  it does run smoothly once it goes.
Later today, I'll see if I can ressurect some kind of jmri based programmer for my
dust covered pr3.  the old laptop had that ....

right now I am thinking I want a headlamp on it
so I can see if the power to the chassis is still working.

Randgust,  I am thinking defective decoder or a pick-up issue.
how do you get slow speed from this thing?
cause it didn't do slow before I added the decoder.

victor

 

[peteski]

Victor,
have you tried to install a parallel 100 ohm resistor across the decoder output along with the motor?  If the motor already has a series-connected resistor, leave that in place too.  So the extra 100 ohm resistor will be connected directly to the orange and gray wires.  Then the motor circuit will also be connected to the orange and gray wires. I am still wondering if increasing the load on the output of the decoder might make a difference.  I doubt it but it is worth a try.

[victor miranda]

Hi peteski,

I am very unlikely to solder in a resistor to ballast the motor.

I was trying to parallel Randy's chassis work.
the chassis when in DC mode acts like what he describes.
when in DCC.... well, it seems smooth.

From all I can tell randgust is pretty good at getting things going.
so when he... had an issue, I thought is was worth looking into.

I think the decoder may have a problem.
I would try very hard to be confident about the pick-ups before
I state the docoder has a failure.
when he says, I get 5 smph as an ordinary speed
then I'll consider alternate circuits.

victor

[randgust]

Because the basic chassis ran (before and after) DCC conversion just fine, any non-decoder (pickup/mechanical) issues are a blind alley here - like I said, I've kinda lost count how many I've done.   I can't remember the last time I saw one of the 3V ones; Kato has been 12V for many years.   The little circuit board in the center, however, makes a very handy spot to route decoder wires to if you grind out a couple bridge contacts.

And, remember that I returned the first Z2 as defective, got it replaced, and the second Z2 did the same thing the first one did.  And got no useful advice from tech support.    What it looked like to me WAS the BEMF - the way it was chugging, but turning on and off that variable did basically nothing.  The second concept (that I basically disproved) was pure erratic contact that was resetting the entire unit at low speed.   I'm also suspect that the 'new improved' Z2 screwed up something, because customers had done several of these with no complaints - then I got a 'new one' that had full analog/DCC capability, and well, not so good.

What I've been looking for - and you guys have supplied me with - are alternatives to that decoder that are just as small, or the alternative being that yeah, you really do have to do some kind of 'keep alive' on the tiniest of tiny chassis here.   I can afford to experiment, Max, as I have....hmmm. six of the chassis in inventory right now.

Here's the YouTube video of the stock Kato motor, including contact tests through the turnouts.  As you can see, pretty much dwarfed by a 40' boxcar.  Decoder size is a significant problem; the Z2 is mounted in the cab up under the roof, tucked inbetween the boiler and the water tank.  And that was a tight fit.   This is the unit on conventional DC:

/>And the same chassis converted over to a 3V micro gearhead (which is by far my recommendation now) running on DC with a 100-ohm drop resistor only.

/>

[jdcolombo]

What I've been looking for - and you guys have supplied me with - are alternatives to that decoder that are just as small, or the alternative being that yeah, you really do have to do some kind of 'keep alive' on the tiniest of tiny chassis here.   I can afford to experiment, Max, as I have....hmmm. six of the chassis in inventory right now.

In addition to the Zimo MX621, my other "Go To" decoders that are similar in size are the ESU LokPilot micro (10.5mm long, by 8.1mm wide, by 2.8mm thick) and the Lenz Silver or Gold mini (11mm x 9mm x 2.8mm).   CT Elektronik has decoder that is even smaller, the DX76zD at 6.9 x 6.1 x 1.7mm, but I don't have as much experience with CT's products (though I am told by others that their motor decoder circuits are the equivalent of other European brands).  I did try one in my FEF, and found it worked about as well as the ESU, but not quite as well as the Zimo.  The differences weren't huge, but the Zimo nosed out the other two in the FEF.

If you can afford to experiment, you might get one of each of these, try them out, and see what works best.   In my Kato FEF, I found that all three of the decoders I tried (the Zimo, ESU and CT Elektronik; I didn't try a Lenz because I didn't have one on hand) needed a bit of manual adjustment to the BEMF parameters to get the best slow-speed performance.   That's one of the things I dislike about the TCS decoders: you can't adjust the BEMF parameters manually.  When they work, they work very well; when they don't, they don't work at all.

John C.

[victor miranda]

Hi Randgust,

hmmmm.
I have a lot of respect for what you do.
I like your posts and I read your thoughts as carefully as I can.
when I grow up I hope I can be as good.
I know I will be a successful man then

take my word for this,
I am careful when I try to give you advice.

this was back a bit in this thread. I quote it for convience.
"The most annoying part was that it ran smooth as silk under normal DC
 (same track, same locomotive), but the decoder (either under analog mode or DCC)
the low-speed stop/go/stop/go hiccupping at minimum speed wouldn't stop."


from here I am assuming you have the problem with only this one chassis.

For the decoder to act the way you desribe, it has to see a signal.
this point has not got much room for debate.

so that leaves how does the signal get generated?

at slow speeds, somehow, a close to "DCC signal" train of pulses is created.

I do not know how.
I have guesses about it,
I suspect the gears,axles,and pick-up strips and at some slow speed
is lifting and dropping a wheel or causing the truck to lift a wheel
just enough to change the resistance of the current to the decoder
that the decoder sees a 'DCC signal.'

a slightly eccentric axle point could do this
a weak chassis pick-up strip could do this

There nothing about n-scale that causes more frustration
than the electrical pick-up problems.

victor