Railpower 1300 testing

[up1950s]

Wow , not only did Max take this on , then found cause , but designed and built a fix . Max gets another badge to his hat , SPARKY . For sure if he was in the NMRA he would have been awarded MASTER MODEL RAILROADED MMR # 6?? with this latest effort . I bet Rapido is happy with all this ammo . Thank you Max .

[mmagliaro]

Wow , not only did Max take this on , then found cause , but designed and built a fix . Max gets another badge to his hat , SPARKY . For sure if he was in the NMRA he would have been awarded MASTER MODEL RAILROADED MMR # 6?? with this latest effort . I bet Rapido is happy with all this ammo . Thank you Max .

Thanks, man, but I don't deserve that much credit.  It was more curiosity than anything else.  I tended to believe that when Rapido said they had a lot of customers with blown decoders using that throttle, it was more likely that there was something at least different in that power pack than others, so I wanted to see if that was true.

And by the way, I am indeed an NMRA member.  I have oscillated in and out of the NMRA for some 30 years, but I have been "in" for the past 4.

[peteski]

Max,
try this circuit in your simulator.




The throttle is basically a common collector amplifier. Quoting the above link:
In electronics, a common collector amplifier (also known as an emitter follower) is one of three basic single-stage bipolar junction transistor (BJT) amplifier topologies, typically used as a voltage buffer.

As you see I added a 1k resistor in the base circuit with the Zener diode behind it.  Since the transistor has a very high gain, the base current should be so low that the added resistor will have insignificant effect. And this way the Zener current will also be greatly reduced. It will still clamp the base voltage (and the output voltage).  If possible I prefer to design circuits which do not generate much heat. And the above redesign should do the trick. You can try using even higher value resistance for R3 (to further minimize heating of the Zener diode).

[alhoop]

Why not:
Probably item referenced in reply #204.

MRC:
AT880 UNIVERSAL VOLTAGE REDUCER (AC/DC)
$10 or about $1 if you build your own.
Al

[peteski]

Why not:
Probably item referenced in reply #204.

MRC:
AT880 UNIVERSAL VOLTAGE REDUCER (AC/DC)
$10 or about $1 if you build your own.
Al

Sure, you can use it. But that will reduce the low end voltage too, so probably the first 1/4 of the speed knob rotation will not be useful (dead).  Max's modification only clamps the high end voltage peaks.

[mmagliaro]

Yes, the diode pack limits the whole range of control, which is not very desirable.

Peteski, your mod works beautifully.   As you surmised, and extra 1k of resistance in the base drive to that transistor isn't going to reduce the current enough to matter a whit.  With a 20v zener, the peaks are capped at 18v, and the RMS DC output is about 12.4.  What's more, the output peaks and RMS maximum are quite load independent.  At 40 mA load, the output is about 12.7V, and at 450mA load, it is 12.3V.  The maximum current through the zener is now more like 2 mA instead of 15 or 30 (what I was getting with the 100 or 220 ohm where I had it).

Very good observation, sir, and a better way to do this.

I will make the mod to the actual unit and photograph where the parts go so that people following along, who want to make this change, can see it (and so we can see how it actually behaves with the physical components.)

And one more thing... the irony of this is that the mod I made to my OWN throttle, the Peter Thorne High Capacity Throttle, was done YOUR way!  I don't know why I didn't think along the same lines when I did it to the 1300.  Here's the schematic to mine.   Note the zener bridging across the 330k ohm resistor (that resistor was there in the original design and I did not remove it, so that when the zener isn't conducting, that resistor still does its original job).  In this case, it is in series with a 4.7k resistor, so it has minimal current through it.  Because this "old" school design uses a 3-stage transistor amplifier, there is so much gain from those first two stages that you can do almost anything you want to the control circuitry in front of Q1 without affecting the linearity of the output.

Thorne's circuit also had an optional momentum and a remote "walk around" tethered control, which I omitted because I didn't need them.

[peteski]

Very good Max!  Wit some team-effort we now have a simple 2-component  modification which will do its job and run cooler too. 

[mmagliaro]

Very good Max!  Wit some team-effort we now have a simple 2-component  modification which will do its job and run cooler too. 

Now... should we send this suggestion to MRC?  It has been suggested to me.  I wonder if I should.  It is probably a moot point if they are no longer manufacturing the 1300/1370, and I'm not sure they would appreciate my attempts to "help".

[peteski]

Now... should we send this suggestion to MRC?  It has been suggested to me.  I wonder if I should.  It is probably a moot point if they are no longer manufacturing the 1300/1370, and I'm not sure they would appreciate my attempts to "help".

IMO, it is a moot point.

[Point353]

Now... should we send this suggestion to MRC?  It has been suggested to me.  I wonder if I should.  It is probably a moot point if they are no longer manufacturing the 1300/1370, and I'm not sure they would appreciate my attempts to "help".

Has MRC discontinued the 1300 and 1370?
They're still shown on the MRC site.

Also, MRC continues to assert that the issue is with the decoders in the Rapido locos, not with their power packs.

Has the failure mode of the Rapido decoders ever been established?
Rapido (Jason) has been silent in that regard ever since pointing the finger at MRC.

Do the Rapido decoders meet the minimum withstanding voltage requirement specified by the NMRA?
Is there some standard that the MRC power packs are not meeting?

[MK]

I will make the mod to the actual unit and photograph where the parts go so that people following along, who want to make this change, can see it (and so we can see how it actually behaves with the physical components.)

Thank yu ahead of time! 

[up1950s]

Now... should we send this suggestion to MRC?  It has been suggested to me.  I wonder if I should.  It is probably a moot point if they are no longer manufacturing the 1300/1370, and I'm not sure they would appreciate my attempts to "help".

Iffin you do , send it registered mail . This way the now can't deny knowledge , and what can they do to you ? Nada , maybe give you a free upgraded pack , any negative response would be foolish .

[peteski]

Max, while off-topic for this discussion, are you sure this schematic is accurate?
1. The negative output of the bridge rectifier is tied to one of the AC inputs.  That will burn up the lower left diode!!
2. The pulse injection switch: what is the purpose of shorting its output to ground when it is turned off?  Shouldn't it just be allowed to float?

[mmagliaro]

Ha ha ha!   Right you are, Pete!

The corrected diagram is below.

It has probably been that way for years in my circuit simulator.   And after I fixed it, it
had no effect on what the simulator did.  I can only surmise that it is because I put the bridge rectifier into the simulation as a single device (a "bridge rectifier").

As for that switch... the only reason that's in the simulation is because I want to be able to run it with and without
the pulses turned on.  But if I just disconnect a lead and let it "float", the simulator won't run because I have an unconnected node.  So I threw it in like that, harmessly going to ground, so it wouldn't affect anything.  I can then just touch the switch in the drawing and it will "flip" and I can run it the other way (the simulator user interface lets you have switches that actually "switch" like that, which is very handy).

And this, folks, is why after the simulator says one thing, you always have to build the circuit!

[peteski]

Ah, so simulators are just that: simulators. Not perfect. Still, it is weird that the simulation of a bridge rectifier is an accurate representation of its internal components.