Have I burnt out the decoder ?

[Jimbo1707820979]

Rather carelessly, I ran my DCC Heljan Lion on an analogue track without checking that it had been programmed for DC running. (CV 29 ?)  It ran one circuit of the track rather sluggishly (not its usual performance) then stopped.   I have been unable to get any sort of peep out of it since, either on DCC or DC and the directional and cab lights do not come on.

Can a DCC loco be damaged in this way ?  Would be grateful for your advice.

[Howbi]

First step would be to place the loco on your program track and reset the decoder by writing 8 to CV8. This will set the loco code back to 03 but then you can test on your running track before resetting back to the preferred code.........HB

[Jimbo1707820979]

Thanks HB. I will have a go and report back. The Lion was bought "DCC on board" so I don't know what make it is. I will also try to read the make on my Elite.

Do you think DC running could blitz a decoder ?

[Rog RJ]

"Do you think DC running could blitz a decoder ?"

 

It's possible but equally possible that it would have burnt out on DCC.

[Chrissaf]

Applying pure DC to a decoder shouldn't damage it (even if CV29 is not set for DC running). The first thing the DCC signal hits is a bridge rectifier to convert the signal to DC anyway. But not all Analogue controllers put out pure DC**, some of the really old controllers can output some really noisy electrical rubbish and can exceed 18 volts in value. So it would somewhat depend upon which Analogue controller you were using as to whether there is likelihood of damage.

.

Note** I am discounting controllers that use PWM (Pulse Width Modulation) in this statement. Although PWM is not strictly pure DC. A modern PWM controller should still produce pulses of pure DC that are electrically clean.

.

[Jimbo1707820979]

Chris, I think you may have hit on the cause of the problem.   Though as a last resort I am going to try HB's recommendation when I get the Programming Track connected to the Elite.  The Lion 'expired' when running controlled by my H&M Duette.   Do you know if this controller puts out Pure DC ?

[Chrissaf]

As you didn't specify the exact Duette model you have. I have assumed the one that I have seen most often, namely this one:

.

/media/tinymce_upload/1544c50e6681f81a512a796b28f70d6f.jpg

.

Even if the one you have is not exactly the one above. What I have written about the one above will more than likely apply to yours as well, if designed and made around the same time period.

.

This H&M Duette has no semi-conductor voltage controller in it. Even the rectification from AC to DC? (cough) uses copper oxide or selenium metal plate rectifiers**. The output to track is controlled via a pair of large wire wound rheostats. Some mechanical switches give a very crude way of modifying the output waveform (one of the switches controls whether full wave or half wave rectification is used - half wave will give a very lumpy output waveform). You can see from the photograph that there are no smoothing capacitors inside the case. The output will thus be quite rough and be a wavy DC waveform. At best, I would call it pulsed DC and certainly very noisy electrically. The pulsed nature of the output is not PWM. With no semi-conductor voltage regulation, the output could potentially exceed 12 volts when only supplying a low current load. For any given speed control knob position the output voltage will drop as the current load increases. You have to remember that these Duettes were designed when locos drawing 0.3 to 0.7 amps was typically the norm. This is significantly higher than modern loco motors.

.

If the DCC decoder is indeed damaged, then I think that there is a high probability that the Duette is the cause. If lucky, it may be that the Duette output electrical noise has just corrupted the decoder configuration and the CV8 reset suggested by HB could recover it.

.

/media/tinymce_upload/b632ce94a92c20c36cf16626590fe453.jpg

.

Note**: Metal plate rectifiers were very common when valves were king. Now a largely obsolete technology, they can still be found in specialist power products (non railway controllers that is) that generate very high currents.

.

 

[Fishmanoz]

The thing about all DC controllers, whether using PWM technology or very basic rectification such as the Duette is that they produce an output with a non-zero average voltage which is controllable, unlike the input AC waveform which always has an average value of zero.  The output voltage will always have AC components as part of it, starting with the 50Hz input voltage and then harmonics thereof (100Hz, 150Hz etc) up to very high order harmonics in the case of PWM given it is a square wave (Fourier or frequency domain analysis tell me this, look them up if you want to know more).

 

When such waveforms are applied to the electric motor in a loco, the highly inductive load the motor provides is open circuit to all of the AC components in the voltage waveform, leaving only the average or DC value to drive the motor.  This is also the case with a decoder whose motor output is PWM too.  

 

So so what happens when such a waveform is applied to a decoder.  Two things occur.  First, the decoder rectifies and smooths the waveform through a bridge rectifier and smoothing capacitor arrangement to provide a DC supply to all of the decoder electronics including the motor drive.  The second thing is part of the input is diverted for analysis to first determine if it is DCC and, if so, interpretation of the DCC signal to determine the address and the instruction.  If the address matches the decoder address the instruction is decoded and fed to the motor or function outputs, otherwise it is ignored.  But if the decoder doesn’t recognise the information as DCC, it assumes it to be DC and applies the average value of it for motor control.

 

Where does all of this get us?  HB and Chris are hypothesizing that the mass of harmonics coming in have confused the decoder and somehow corrupted it’s interpretation of the signal such that it no longer knows whether what it is getting is Arthur (DC) or Martha (DCC) and needs to be reset to sort it out.  After which it will again do as above.

 

Or the alternative - the incoming mess with relatively high current capability has exceeded decoder power handling specifications somehow and it is now blown.  Attempting to reset the decoder should sort out which is the case.

 

PS.  Chris, may I suggest that it is better to think about average values rather than whether the output waveform can be characterized as DC-like or non-DC-like.

[Jimbo1707820979]

Thankyou, Gents, for taking the time to post all this very detailed information. Yes, Chris, that is indeed the model of Duette I have.

This leaves me with these conclusions : A. Not to run any locos with the Duette without - in the case of DCC-fitted locos - making sure they are set for analogue operation.       B. Preferably not to run them at all.   C. To try to rescue the Lion with the CV8 reset.                Kind regards and thanks again. Jimbo.                                                                                                               

[Fishmanoz]

Jimbo, the only difference you should see between have DC Running set on or off in CV29 is the loco will or won’t respond to DC, no more or less likelihood of damage. And your first action to track this down has to be the reset, after which DC Running will be on.

[Jimbo1707820979]

My attempts to reset the Lion's decoder to 8 failed completely. Also when trying to read the decoder, my Elite showed XXX. The loco appears to be quite dead so it seems the decoder has been blown. No response either to DC current or PP3 battery.

Afterwards, I tested a couple of DCC-fitted locos by switching them to DC "ON" and then later to DC "OFF". To my surprise this did not appear to make any difference - they both ran OK on the analogue track (Hornby controller not H&M !).   A bit puzzled by this.   However, the Little Controller's Hornby 0-6-0 "Thomas" which was bought new and DCC-fitted would only run in a weird way - 3 or 4 feet, then a pause, then forward again. And so on, each time.  I cleaned the wheels and lightly lubricated. Also checked and cleaned the track. Then I measured the BtB on the wheels with a digital caliper and found it only 14 mm. Could this affect the running ? Anyway, I have ordered a DCC Concepts brass wheel gauge but am hesitant about "spreading" the wheels with a small screwdriver as this seems rather brutal. Is a wheel puller necessary ?

Could there be another cause for Thomas's strange behaviour ?  Please excuse my intoducing other themes under this heading. 😳