Railpower 1300 testing

[up1950s]

Is the 1300 a current MRC product? If not, when was it discontinued?

The other question begging be asked is how many model railroaders do you personally know which were affected by the alleged problem with 1300 and Rapido decoders?  None, one, ten, 30, 100?

How many I know ? I only know who posts here . Jason knows enough to connect the dots .

[mmagliaro]

It is not that bad.  The (only) alleged problem might occur when decoders are used on DC with DC throttles that use pulsing DC (unfiltered output from a full-wave rectifier).  That is when the decoder's built-in filter capacitors will smooth out the pulsing DC to be closer to the peak voltage (rather than RMS voltage).

When decoders are used in their intended (DCC) environment, they are much safer.  DCC signal is a square wave which is highly unlikely to get neat the maximum rated voltage. Its peak and RMS values are the same.

Let's not overthink this or have knee jerk reaction to a problem that none of us heard of until the Rapido's MRC 1300 warning was somehow brought to our attention.  If you run your trains with DCC you have nothing to worry about.  How many people buy DCC equipment to exclisively run them in DC and also have a MRC 1300 throttle?  I think a very small number of modelers.

To me Rapido's problem is more an exception than a rule. And we don't even know if anything was proven scientifically by Rapido.  HAve they actually done some scientific tasts?  Do they know what component is failing?  What are their decoders specs for max safe operating voltage?

Even this thread, while informative and enligtening, doesn't prove that 1300 blows up decoders. ALl we know it is that 1300 seems to have slightly higher output voltage than few other DC throttles.  But we haven't even tested it powering a DCC-sound-equipped loco to see how much lower the output voltage is under more than 50mA of load.

Well, I haven't powered a decoder with it, but back from page 6....
"At 1/4 amp, full DC is 13.6v average, and the peaks are at 23 volts."

So at least we do know that it drops somewhat under a reasonable load, but is still rather high.

[Point353]

Is the 1300 a current MRC product? If not, when was it discontinued?

According to the MRC website, both the 1300 and 1370 are shown as out-of-stock but can be backordered.
http://www.modelrectifier.com/product-p/aa300.htm
http://www.modelrectifier.com/product-p/aa370.htm

The other question begging be asked is how many model railroaders do you personally know which were affected by the alleged problem with 1300 and Rapido decoders?  None, one, ten, 30, 100?

Jason from Rapido stated that they had at least 100 failed decoders, which they attribute to use of the 1300/1370.
https://www.therailwire.net/forum/index.php?topic=44120.msg566142#msg566142

[peteski]

According to the MRC website, both the 1300 and 1370 are shown as out-of-stock but can be backordered.
http://www.modelrectifier.com/product-p/aa300.htm
http://www.modelrectifier.com/product-p/aa370.htm
Jason from Rapido stated that they had at least 100 failed decoders, which they attribute to use of the 1300/1370.
https://www.therailwire.net/forum/index.php?topic=44120.msg566142#msg566142

If we are to play a detective, it would be *REALLY* useful to know what are the electrical specs for Rapido decoders and what components failed on them.  Something we will probably never know (unless one of the Railwire member has a Rapido model with a disposable ESU decoder and a MRC 1300 throttle), and is willing to kill the decoder and let me do a postmortem. 

All the ESU decoders I reverse-engineered had a what appears to be a Zener diode installed at the output of the bridge rectifier.  I'm not certain it is a Zener, but from my knowledge of circut design, a Zener diode makes sense.  I have not taken it out of the circuit to verify if it is a Zener, and what is its breakdown voltage.  Its purpose seems to be to shunt any excessive voltages. I wonder if the common failed component is that diode, and possibly one or more of the bridge rectifier diodes.

[mmagliaro]

If we are to play a detective, it would be *REALLY* useful to know what are the electrical specs for Rapido decoders and what components failed on them.  Something we will probably never know (unless one of the Railwire member has a Rapido model with a disposable ESU decoder and a MRC 1300 throttle), and is willing to kill the decoder and let me do a postmortem. 

All the ESU decoders I reverse-engineered had a what appears to be a Zener diode installed at the output of the bridge rectifier.  I'm not certain it is a Zener, but from my knowledge of circut design, a Zener diode makes sense.  I have not taken it out of the circuit to verify if it is a Zener, and what is its breakdown voltage.  Its purpose seems to be to shunt any excessive voltages. I wonder if the common failed component is that diode, and possibly one or more of the bridge rectifier diodes.

That's good information!  A diode is the sort of thing that *will* fail if the peak voltage exceeds its peak rating.
But a peak rating of 25v?  Other power packs are hitting peaks around 20v.  So is the difference in blowing a diode really only 5 volts?  Could be.  And remember, with minimal load on it, that power pack is peaking at 25v.  It wouldn't surprise me at all to see a very tiny load from a decoder when it first turns on.  And people running a DCC-equipped engine on a DC power pack are exactly the people who would be running power pack full-out because of the voltage drop from the decoder.  Many DCC engines will move very slowly at full throttle on a DC power pack.

[peteski]

That's good information!  A diode is the sort of thing that *will* fail if the peak voltage exceeds its peak rating.
But a peak rating of 25v?  Other power packs are hitting peaks around 20v.  So is the difference in blowing a diode really only 5 volts?  Could be.  And remember, with minimal load on it, that power pack is peaking at 25v.  It wouldn't surprise me at all to see a very tiny load from a decoder when it first turns on.  And people running a DCC-equipped engine on a DC power pack are exactly the people who would be running power pack full-out because of the voltage drop from the decoder.  Many DCC engines will move very slowly at full throttle on a DC power pack.

Well, that is not exactly what I had in mind. Here is a sample of how the power section of ESU decoders is laid out.



Assuming that it is a Zener diode (I did not actually take it out of the circuit and test verify it) then if the voltage coming out of the bridge rectifier exceeds the Zener voltage, the Zener diode's resistance will decrease and it will start shunting the voltage. As the voltage gets higher, the diode's resistance will get lower and lower. This means that the current in that circuit (through the rectifier diodes and Zener diode) will increase. At some point the current might get high enough to destroy the diodes. Either just the Zener diode or also some of the diodes in the rectifier circuit. Mind you, this is just a speculation on my part. The failure mode could theoretically be a diode with either open or short circuit.

The rectifier diodes used by ESU have low forward voltage drop (around 0.2V under light load) so those are most likely Shottky diodes. Those are often rated for lower reverse voltages (like 40V or even lower).  But I think that over-current rather than excessive voltage would be more likely reason for a failure.

[OldEastRR]

I don't understand why some model mfgrs are creating some products like it was still the last century. MRC with powerpacks, Digitrax with decoders, Code 80 rail .... locos with wiper pick-ups, and other stuff. It's like if Ford brought out a 2019 car that ran on leaded gas and had a distributor. 

[MK]

Sometimes it's hard to change a production line, especially if it's overseas.

[nkalanaga]

Also, every customer doesn't want, or need, the latest-and-greatest.  The main reason Ford doesn't build cars burning leaded gas is that the government banned leaded gas.  Do we need someone to regulate what model railroaders can buy, or use?

[peteski]

Do we need someone to regulate what model railroaders can buy, or use?

Well, yes.  Not sure if I would call it "regulating", but just defining standards.  Do you call that "regulating"?   If standards weren't clearly defined then N scale track from all manufacturers would not be 9mm, the couplers would not be at the same height, and not all models would run on 0-12V DC.  See what I mean?

The operating voltage of N scale DC locos is defined (regulated?) as 0-12V DC.  If someone makes a power pack which produced 0-17V or 0-22V then who is wrong?  The manufacturers of the 0-12V locos or of the throttle that produces excessive voltage at its output?

[Maletrain]

The operating voltage of N scale DC locos is defined (regulated?) as 0-12V DC.  If someone makes a power pack which produced 0-17V or 0-22V then who is wrong?  The manufacturers of the 0-12V locos or of the throttle that produces excessive voltage at its output?

Peteski, I don't understand how to reconcile your comment here with your earlier response to my suggestion that maybe the standards need to be rethought to deal with peak (and or spike voltage.  Since the 12 volt DC standard is presumed to be the RMS max from some sort of rectified AC that has higher peaks, and there are manufacturers that use intentionally peaked (pulsed) power to make motors operate smoothly at low rpm, it seems like there is really no constraint on how high the peaks might become, particularly under such a variety of circumstances as plain DC, plain DCC and dual mode decoders, all with or without sound, "running" from zero to max speed with sound on soft or loud.  If the intent of the current standard for decoders was to assume that the peak voltages were due to sine waveforms with the RMS giving 12 volts, then specifying a standard for the electronics minimum voltage tolerance in the 20-something volt range was probably intended to provide a margin for the electronics rating above the actual peak voltage it would experience.  If the peaks are instead raised to the rating of the electrical components, then shorter life and some prompt failures are to be expected.  Would you want to drive your 2-ton truck over a bridge that had an actual failure estimated to occur with a 2 ton load?  (Bridges are more usually designed with a factor of 3 safety margin, based on experiences with bridge failures over the centuries.)

[DKS]

There must be some language clarification here. There is no regulation of the model railroad industry other than Underwriters Laboratories certification of electrical devices, and all UL does is ensure a product is safe to use, not that it meets any specific performance standards. Any actual legal regulation of model railroad products would be impractical, pointless, and probably destructive.

The NMRA is an organization that establishes recommendations for "industry standards." Manufacturers are not bound in any way to follow NMRA recommendations; it is strictly voluntary. There are some manufacturers who do not follow NMRA recommendations, for whatever reasons, and so compliance is more of a marketing gimmick than anything else.

This system has its limitations and drawbacks. For example, when the NMRA states that there must be "not less than X volts" at the rails, it creates a problem: what if there is more than X volts delivered to the rails? The Railpower 1300 may fall within NMRA guidelines, but still pose problems for newer technology that's sensitive to over-voltage pulses, spikes, noise, or whatever.

Thus we have our current situation, and there's likely no simple solution. New guidelines or revisions are difficult to get added to the current NMRA recommendations, and the latency of design-to-market for manufacturers means there will be those who fall outside of the boundaries.

And so we must rely on the modeling community to help manufacturers like Rapido find the happy medium--if possible.
 

[jagged ben]

Generally speaking, if a component has a voltage rating then it should withstand that RMS voltage even if wave peaks are significantly higher.  It other words, decoders with a 12V rating shouldn't be failing from being used with a PWM throttle  that delivers high wave peaks but still maxes at 12V RMS.  Now, if a certian throttle pack has random intermittent voltage spikes that bring the RMS voltage higher for meaningful amounts of time, that's different.   It's also different if there are subtle long term effects from using packs with some waveform.   Unless I missed it,   I don't believe any evidence has been presented that conclusively demonstrates that the latter effects are or aren't happening.

[Maletrain]

Generally speaking, if a component has a voltage rating then it should withstand that RMS voltage even if wave peaks are significantly higher.  It other words, decoders with a 12V rating shouldn't be failing from being used with a PWM throttle  that delivers high wave peaks but still maxes at 12V RMS.

I don't think that makes any sense, once you drop the assumption that the wave form is sinusoidal and the frequency is from rectified 60 hertz AC power.  Obviously, one could design a wave with a periodicity of 1 cycle per second with a peak voltage of 1000 volts, but still keep the RMS to 12 volts.  A square wave with a pulse width of 0.012 seconds would do that.  So, should every tiny electronic device that can handle 12 volts RMS with a sine wave at 120 cps still be able to handle that? Probably not.

There are now "DC" power packs that do essentially what a DCC decoder does in the way of sensing BEMF and pulsing the track power to get smooth motor operation at slow rpm.  Is that power on the tracks always going to be compatible with every dual mode decoder that somebody might decide to run on DCC?  Probably not, unless the people who make the power packs and the people who make the decoders agree on limits to the track voltage that make sense in terms of the time parameters relevant to the decoder circuits.

[jagged ben]

I don't think that makes any sense, once you drop the assumption that the wave form is sinusoidal and the frequency is from rectified 60 hertz AC power.  Obviously, one could design a wave with a periodicity of 1 cycle per second with a peak voltage of 1000 volts, but still keep the RMS to 12 volts.  A square wave with a pulse width of 0.012 seconds would do that.  So, should every tiny electronic device that can handle 12 volts RMS with a sine wave at 120 cps still be able to handle that? Probably not.

There are now "DC" power packs that do essentially what a DCC decoder does in the way of sensing BEMF and pulsing the track power to get smooth motor operation at slow rpm.  Is that power on the tracks always going to be compatible with every dual mode decoder that somebody might decide to run on DCC?  Probably not, unless the people who make the power packs and the people who make the decoders agree on limits to the track voltage that make sense in terms of the time parameters relevant to the decoder circuits.

Ok, let me rephrase.

If a powerpack is delivering 25V peaks on a waveform that has an RMS voltage well under 21V, and that powerpack is being used to run a dual-mode decoder with a stated rating of 21V, and the decoder gets damaged, then I squarely blame the decoder manufacturer for overstating the voltage rating of the decoder.  I do not blame the powerpack manufacturer for putting 15VDC on their powerpack if the RMS voltage stays below that, and it is a regular waveform with a peak amplitude less than twice the nominal max voltage.  To be clear, in the context of previous posts in this thread, I'm talking about 25V peaks in a repeating waveform, without transient irregular 'spikes' that go higher.  And unless I missed it, no evidence has been presented in this thread that the powerpack in question is delivering such spikes.  It's delivering a regular waveform with an average voltage that matches the spec on the outside.

That's my opinion. 

My reasoning is that model railroaders, and consumers in general, cannot be expected to test their equipment with oscilloscopes before operation, but testing with a quality volt meter is a reasonable expectation.  Also, the difference between 21V and 25V is not large enough that it shouldn't be accounted for in safety factors and manufacturing tolerance, anyway.  That is, if the decoder package says 21V then I would expect that a substantial fraction of those decoders should survive some amount of operation at 25V in any case.  Not that I would make a warranty claim if I knew that a decoder had been so operated, but they should give themselves some headroom on their rating.   Tolerating waveform peaks up to twice the average voltage is an admittedly arbitrary number in this context, but eminently reasonable given common waveforms one finds in powerlines and electronics of all sorts.