More on this topic.  A fuse or circuit breaker is a bad idea in a motor circuit controlled by a MOSFET. 

I learned that many mobile decoders use MOSFETs (metal oxide semiconductor field effect transistors) as the output stage of the controller.   One of the common failure modes of MOSFETs is a large, fast voltage spike, which you will certainly get it you disconnect one leg of an inductive load from a controller!  So, above I asked "why not...?"  Bad idea! 

It turns out voltage transients are MOSFET killers, and arcing is a fact of life in electric model trains, so these decoders are working in a very hostile environment.  Good motor maintenance can reduce arcing at the motor.  Putting a capacitor across the motor terminals to suppress noise is a controversial topic, some motor controller vendors recommend it, others don't (don't know about DCC decoder vendors on this).

jv

Thanks Kevin.  Could you please tell me where you found the 22V v. 27V spec. for Soundtraxx?  I can't find it.

Thanks,

jv

Update: just received a reply to this question from Soundtraxx and they say that the maximum operating voltage is 27V.

Update 2: I just found the reference to 22V max in tsunamisteam_users_guide.pdf and Kevin is right according to this document, which states:

Quote
Error 10 - Over voltage fault
An Over-voltage fault indicates that the track voltage is greater than 22
volts. When this error occurs, motor control and sound will be shut off
and the error code will continue to flash until the voltage is brought below
22 Volts. Note: Tsunami’s absolute maximum voltage input is 27 volts!
Track voltages exceeding 27 volts may cause permanent damage to the
decoder.

I have asked Soundtraxx support for clarification on this.

Update 3: Kevin is right, 22V maximum operating, 27V absolute maximum.

An update on this for interested observers.

I'm returning the Digitrax Zephyr.  I like it, like the looks and the concept of a stationary cab with throttle know, brake lever, and all of those function buttons.  I just don't like that it's limited to 16 Volts and 3.5 Amps.   like the "jump ports" that allow DC controllers to be used as throttles.  I am not sure I couldn't have just added a booster to drive the track to higher voltages and currents, 'though I read the Zephyr could be used as a booster.  It would have been perfect if the Zephyr had a "G-scale" switch that allowed it to drive 24V at 5 or 8 or 10 Amps, perhaps with an external power supply.  Oh well.  I'm about to order the Super Chief 8 Amp starter set - quite a bit more money, but should suffice forever, for me.  I've also ordered a 30V 10A laboratory power supply to use instead of buying the PS2012, which at about $150 isn't a very good deal, and it's rather inflexible on output voltage, offering only 3 discrete steps of 13.8, 18, and 23V.  The lab supply has fine and coarse voltage and current adjustment knobs, a digital display for each of Volts and Amps, a built-in thermally controlled fan, and was half the price!

On decoders, I received the Tsunami TSU-1000.  It was $93.  It is a 1 Amp decoder, so marginal in the BBH 4-6-0.  Norm at Protocraft has offered to convert it for 3.5A operation for $70, exactly the difference between his price for converted Tsunamis and what I paid for mine.  Still thinking about it.  Meanwhile, I've found a G-Scale solution for mobile plus sound that's under $100: the Digitrax DH465 4 Amp HO/O gauge mobile decoder, coupled with their Sound Bug SFX006.  Street price combined this is $91.  Not sure I will try it yet, I like the Soundtraxx sound samples I've listened to, and I don't like that I'd have to invest in more gear to download sound files, and that the s/w is Windows-based (I'm a Mac user).  But we'll see.

By the way, Kevin, the Tsunami 1000 is limited to a maximum track voltage of 27V, not 22V.  Perhaps this is a revised specification?  Don't know, but that's what I see on the web site.

Thanks, guys, for the replies.  Joe S., I suspected that after reading Joe Z's reply, since a dropout wouldn't cause a negative excursion, unless there was some charge stored somewhere on the loco, and yes, the spinning motor which is nearly entirely inductive would do that; you can't change the current through and inductor instantaneously, as they say.  It could be solved with the addition of a Schottky diode and load capacitor across the motor, but I'll have to investigate a dual-polarity solution.

Quote from: Joe Satnik on March 17, 2015, 11:19:18 AM
4.  A thin layer of Wahl clipper oil to help minimize arcing between the track and wheels.

Is this the same thing as the conductive oil Bachmann sells?  I'd like to treat my new Bachmann steel track with something to prevent corrosion, and a conductive, light oil would be ideal.

Thanks,

jv

Thanks Kevin.  Fifty bucks - ouch!  I was thinking of trying to make one.  I thought I'd use a small R/C plane servo, making a linkage that can be forced manually without damage; use a single function DCC decoder, and build a circuit that outputs a 1 ms, 5V pulse with function off, and a 2.5 ms pulse with function on.  I can just use a slow-motion servo, as are used for flaps or landing gear, or move my pulses slowly from one 1 to 2.5 ms slowly (it's a train of pulses nominally 20ms apart).  Would be nice if the decoder supplied 5V, else I'll need a rectifier and regulator.  Even so, should be easy.

Greetings,

I am the recent new, proud owner of an Anniversary D&RGW "Bumble Bee" 4'6-0 locomotive.  I'm using it on a relatively large, indoor, steel sectional track layout (over 100' of steel track).  I am powering it with an MRC Trainpower 6200 which is rated, in Mode I 0 to 18.5VDC and 18VAC for accessories to 60A max.  I'm not using any accessories.  I've seen much higher voltages, though, at full throttle, up to 24V or so.

When driving the train forward, the light on the tender flickers.  Does anyone know why this might occur?  I was thinking that the only way that's possible is if the power pack is controlling output using pulse width modulation, and the output is making negative excursions. 

Thanks,

jv

I searched in large and couldn't find anything on this topic.  

What I'd like is a slow motion, DCC controlled switch machine for the Bachmann steel turnouts.  Ideally it would be designed specifically for these turnouts and snap right on.  

I don't have an "under the layout" option, because my "layout" is sectional track on the floor.  

I'm in the process of reading up and researching G scale track layout, wiring, DCC, and more, so please excuse me for my dumb questions.  I don't think what I seek is available, but maybe the experts here can guide me.

Thanks,

jv

Thanks again, Kevin.  I may have been hasty here, having purchased the TSU-1000, and the Digitrax Zephyr Xtra controller.  The controller documentation says the output is limited to 13 Volts- it's a bit vague and contradictory - but it seems it's rated for a maximum of 16. 8 VDC input ('do not exceed 16.5VDC on the next line), and the controller drops about 0.5VDC - so max to the track about 16V.   With the supplied DC power supply it will be more like 13.3VDC.  This might not be enough for the 4-6-0, plus I'm already marginal for current handling on the TSU-1000.  Yikes!  Might want to rethink this. 

Anyway, yes, I agree with your sentiment that there's comfort and more enjoyment in the confidence that you have design margin in the installation. 

Wow, beautiful work.  Looks real.

Here's an interesting update.  I now own two (2) Anniversary 4-6-0's, one a new chassis that I put under my PRR loco, and a new Denver & Rio Grande Western "Bumble Bee" 4-6-0.  The former is the one for which I have never seen more than one Amp current, hauling eight cars plus the tender.  The most current is seen dragging this through a succession of left/right/left sections of steel track, set up that way because I ran out of straight sections (more are on the way).  The latter draws over one Amp at several points along the way. 

I was wondering why these two seemingly identical rigs are so different in operation.  I have a few hours on the former, and near zero on the latter; maybe the latter just needs to break in a bit?  Or could it be lubrication, or lack thereof?  Or maybe just a difference in motors?

Kevin, thank you so much for taking the time and trouble to lay it all out for me, and to share your obvious deep knowledge of this topic!  Wow.

I have questions!  So, I take it that the 4-6-0 and much of the cars are taken from narrow gauge prototypes?  What about the PRR passenger coaches, are they also based on narrow gauge?  Was there a PRR narrow gauge passenger line in reality?

I guess a rail gauge of 1.6" would be perfect!  And, 2.5" for the standard gauge models.  Has anyone ever done this?  Just fanciful thinking.  I know there were 2.5" gauge trains before the WWII, but it's rather obscure at this time.  I haven't ever seen 1.6" gauge.

I wonder what the scale of my Bachmann figures is - the fireman, conductor, etc.?

Thanks for the detailed reply, Kevin.  I checked out the decoder vendors you cited, thanks.  The prices are high, like on the order of two or more times higher than the Soundtraxx unit; admittedly, they handle more motor current, but otherwise no particular advantage.  In fact, I prefer the Soundtraxx.  

I should say at this point that I'm an electrical engineer, and have started thinking about this topic in engineering terms, specifically, what my requirements and desired system operational parameters are, and what behavior I want, both nominally and in extreme conditions.  

First, 'though I did do a brief test to see what is the stall current of my Bachmann Anniversary chassis, generation 5, I did not want to test it to destruction, and did it without building any elaborate test stand.  But, yes, I saw current over 3 Amps, and that was enough for me at that moment.  The voltage at that time was about 14V.  

This morning it occurred to me the idea of putting a decoder in a locomotive that can source the motor's stall current is to sacrifice the motor when the stall condition occurs; and likewise, putting in a decoder that will smoke when the stall condition occurs is to sacrifice the decoder.  Neither alternative represents a good design to me.

Stalling the motor is not normal operation!  If this happens it's because something has gone wrong in operation, or the combination of available traction and motor power is inappropriate.  

My design requirement is to have a motor that can pull within its safe operating region the load I want to subject it to, and to employ a decoder that can supply sufficient current to support this, including any normal operating mode transients.  I would further protect the motor, decoder, wiring, and contacts with some kind of circuit protection device.  

So, my BBH Anniversary loco never draws 1 Amp or more in normal operation on my layout, which is indoors and level (no grades).  I think a Soundtraxx TSU-1000 will do just fine, and I might even add a 1 A circuit breaker to protect the decoder.

I don't know yet if the decoder can sustain a stall current condition long enough for a commonly available 1 Amp breaker to trip; and, I don't know if the breaker will interfere with back EMF.

Sound reasonable?

Update: after researching circuit breakers, I will probably go with a fuse, like an ATOF automotive, 32V rating, low current fuse: much lower resistance, more choices of blow speed, much more easily sourced, and much cheaper.

Thanks, that's a good question, and it nags me a bit, too.  I can't find, or figure out if there exists, a G-Scale decoder that runs the motor, functions, and produces sound as the Soundtraxx decoder does.  I find the Digitrax documentation and website to be a bit obtuse - there seems to be an assumption that people going there already know the structure of the product line, but I don't.  So, does anyone make a G-Scale equivalent of the Soundtraxx TSU-1000? 

Thanks,

jv

I am looking around at DCC systems, and decoders in particular, for my Bachmann Big Hauler Railroad.  I really like the Soundtraxx decoder, the TSU-1000, P/N 826116, for my 4-6-0 locomotives.  

I just measured over 3 Amps stall current for my new Anniversary loco chassis!  This decoder is rated at 1 Amp.  However, I never see more than about 800 mA in normal operation.

I thought for a while about building a current follower, but then a simpler solution occurred to me.  Why not put a 1A circuit breaker in the motor circuit?    BAD IDEA!  See reply umber 12, below.

A 1A, resettable C.B. can be had for about $7., quantity one.  It can handle up to 35 Volts, and trips in from 4.5 to 28 seconds.  Sounds ideal, to me; I'm assuming the TSU-1000 has about 50% margin and that the damage would be thermally-related, so 4.5 to 28s at stall currents should protect the decoder.

Has anyone tried this?

Apologies if this has been covered before, my search mojo is waning lately.

Running the Big Hauler with my grandson Sunday, just lounging on the floor and eye-balling them, it seems to me the scale of these pieces vary.  I'm sure there are experts here who know, and I would love to learn what we have.

My inventory:

Big Hauler PRR set, circa ~1994
4-6-0 loco, gen. 3, with tender
passenger coach
observation car w/ drumhead

(Big Hauler Emmett Kelly, Jr., set, about the same vintage, but not the subject of this post)

Purchased separately:
Item No. 93114 'G' Four Wheel Caboose (Pennsylvania)
Item No. 98214 'G' 3 Bay Hopper Car (Pennsylvania)
Item No. 93814 'G' 8-Wheel Center Cupola Wood Caboose (Pennsylvania)
Item No. 97197 'G' Combine w/Lights (PRR Tuscan Red)
Item No. 93253 Doublemint Gum Bill Board Reefer
Item No. 93938 Open Streetcar United Traction Co.
Item No. 993433 Single Dome Tank Car (Dad's Rootbeer)

It seems the passenger cars are a different scale - maybe 1:24 or even 1:29?  The reefer also looks small.  Or maybe the cabooses and engine are bigger.  

I assumed that all were 1:22.5, but now it looks odd to me.  Maybe it's just my lack of knowledge of real railroads.

Thanks,

jv