problem getting a DZ126IN to be recognised. (Solved)

[nickelplate759]

I installed a DZ126IN in a Minitrix ICE 3 (there's a socket, so it's easy).
After installation (as well as before) it runs fine on DC - or as fine as the Digitrax controller can, which is somewhat less smoothly than with no controller.  So far so good.

When I start up DecoderPro3 to try and program the thing, I keep getting "not found" errors.   But not EVERY time.  every now and again it will succeed in reading a CV or two.

Any suggestions on how to debug this?

George

[peteski]

If it reads CVs sometimes (and the decoder can be controlled under DCC) that would indicate that the decoder is properly installed and working. The easiest thing is to check for dirty track/wheels.  That could cause intermittent problems.

One thing I have observed with Digitrax decoders is that they do not play well with the RFI (Radio Frequency Interference) circuit which is installed in the motor circuit of all European models.  Usually it consists of 2 caps and 2 choked (small wire coils).  Removing the capacitors and replacing the coils with just  a wire jumper takes care of the problem.  But I'm not sure how that RFI circuit would mess up reading out CVs.

[C855B]

... I'm not sure how that RFI circuit would mess up reading out CVs.

By rounding off the square waves. Square waves are composed of a progression of odd-integer harmonics of the fundamental frequency. By the time you get to the 3rd or 5th harmonic you should be well out of the bandpass of the LC filters.

[nickelplate759]

Hmm, it does have filters on the power supply side.
I tried attaching power directly with alligator clips, and it isn't any better, so cleaning the wheels (more) probably won't help.
I'm debating removing the filters.

Another option - should I try a different model of 6-pin decoder (maybe the one from TCS)?

[peteski]

Hmm, it does have filters on the power supply side.
I tried attaching power directly with alligator clips, and it isn't any better, so cleaning the wheels (more) probably won't help.
I'm debating removing the filters.

Another option - should I try a different model of 6-pin decoder (maybe the one from TCS)?

Yes, trying to use another brand decoder is a good idea.

I also never seem the RFI circuitry (in 6-pin NEM connector equipped locos)  on the track side of the circuit. It is usually on the motor side of the decoder (between the M+ and M- output of the decoder and the motor brush contacts).  If it really is in the track circuit (between the wheels and the decoder's track inputs) then it could affect the programming track readability and DCC operation.

C855B - when programming on the programming track the decoder just pulses the motor outputs to create a load on the programming track when acknowledging a programming packet from the decoder. I didn't think the programming track circuitry would be so sensitive for the RFI circuitry low-pass filtering on the motor to affect the programming. In my experience that RFI circuitry only affect the BEMF functionality of a decoder resulting in erratic speed control.  I don't believe that BEMF is enabled while the decoder simply pulses the motor circuit during programming track operations.

[nickelplate759]

Well, I got back to this train again.
There's a filter (coil) in each of the  circuits from the decoder to the motor (not between track and decoder).   There's also a capacitor connecting each motor lead to one side of track power (both motor leads to the same side of the track power, it hapens to be the lower right side in this picture).   There's also a black rectangle with a lead on each side, in the red circle in the picture, that I can't figure out.  one end is connected to a motor lead - I'm not sure about the other end.

Here's what it looks like:
[ Guests cannot view attachments ]


The decoder sometimes comes to life after it's been sitting on the (powered) track for a while.

I gather I should replace the filters with solid wire (22 AWG I happen to have handy), and just remove the 2 caps.  What about the mystery device in the red circle?

[peteski]

The mystery device looks like solid-state self-resetting circuit breaker (protects the motor from overcurrent).  It is also marked "F1" with "F" for "fuse".
Some Digitrax decoders (or all of them) do not like some of the RFI suppression components in the motor circuit.   But that usually caused unexpected speed behavior whiel the loco is running. I have not seen the RFI circuit cause intermittent programming failure like you describe.

Yes,  removing those 2 caps and shunting the chokes (coils) with pieces of wire will eliminate them from the motor circuit.  You can leave the chokes in, just solder pieces of wire between the solder pads on the bottom of the board to shunt them.  That way you can easily undo the modification.  Same with the caps - just unsolder them and then slide them over and solder one side to one of the pats. That way they are removed from the circuit, but still there if needed.

[nickelplate759]

I saw the F1, but didn't think of a circuit breaker as a fuse (silly software person that I am).
Can/should I replace that with a shunt as well?

[peteski]

I saw the F1, but didn't think of a circuit breaker as a fuse (silly software person that I am).
Can/should I replace that with a shunt as well?

I guess you can, but why not leave it alone?  It should not affect the motor's functionality or confuse the decoder.

[nickelplate759]

I shunted around the chokes (what I erroneously labeled "filters"), and removed the  caps and ferrite beads that were hidden under the circuit board that the motor leads were looped through.  Left the "fuse"/circuit-breaker dingus.   It works!

Thanks Pete!

On to fixing the wheels (not a DCC problem).

[peteski]

I shunted around the chokes (what I erroneously labeled "filters"), and removed the  caps and ferrite beads that were hidden under the circuit board that the motor leads were looped through.  Left the "fuse"/circuit-breaker dingus.   It works!

Thanks Pete!

On to fixing the wheels (not a DCC problem).

Excellent!  The ferrite beads could have been left alone (then have negligible inductance).

As for the naming, inductors, coils, chokes, is are basically coils of wire usually used in filter circuits, so you were not in error when you called them filters.  Well, the actual filter is the combination of the inductor and the capacitor.  It is a low-pass filter. It blocks high frequency noise, generated at the motor's commutator, from spreading to the rest of the circuit, and from being transmitted from the locos wiring as electromagnetic waves (radio-frequency noise).  Most European-manufactured decoders work fine with this type of filter in the motor circuit (which is required in European countries). But since there is nor laws dictating use of the filters in USA, American decoder manufacturers' decoders often do not work correctly when the filter is included in the motor circuit.