Jeff Smith Posted October 21, 2014 Share Posted October 21, 2014 I have an On30 layout fully wired for DCC but currently no DCC controller. I have several DCC equipped Dual Mode Bachmann locos plus a non-DCC Mogul. The Mogul works very well on DC (Gaugemaster) with excellent slow running etc. Yesterday I was experimenting with a DCC equipped 0-4-2 Porter. On DC this has a somewhat abrupt start - although you can see that power is on as the headlight comes on but the loco movement is delayed (this happens with all my DCC equipped locos on DC). Then I found it would stall on the point frogs. The pick-ups on all four wheels seem ok but it is a fairly rigid chassis so perhaps one wheel drops into the frog (Micro Engineering code 70 track with live switched frogs) but even so this should leave two opposite wheels on the track. Anyway my question is, could this stalling be improved by running with a DCC controller? My thinking is that the momentary loss or reduction of current could be causing the DCC chip to reset and delay recovery? Link to post Share on other sites More sharing options...
Phil S Posted October 21, 2014 Share Posted October 21, 2014 Whenever a dcc-fited loco is used on an analogue system - you must remember that the first part of the speed/voltage control is effectively 'dead' - although it may manifest itself as coach and other lighting (on-always types) coming on. once the track voltage reaches approx 5V (this can vary between decoders .... older types were made with processors which NEEDED 5V, and therefore dis not 'turn on' unitl about 7V was on the track ....newer types using 3V3 processors will probably work from about 5V .... the extra being caused by the losses in the bridge rectifier on each decoder's input. From 'that point', the processor will measure the increasing track voltage, and proprortionally, increase the drive to the motor power stage (pulse-width drive ... whilst the amplitude of the pulses is also rising as the ananlogue track voltage increases)). Once you have familiarised yourself with the 'point on the dial' where each individual locos starts up, (eg lights, sound start up) ... then a slow increase will produce a smooth start. This is likely to be affected by the decoder's setting for 'Start Voltage' when used on DCC ... but will need a programmer (and therefore a DCC controller) to adjust. Because Serial Control Data is used (circa 9kHz bit rate, but maybe a tenth of that, at most, for a complete instruction, and these are sent on a repeating cycle, whenever loss of data occurs, there will be a potential problem: As far as the control-logic goes on most decoders, this will simply cause the last instruction to remain in use ... BUT where the loss is of sufficient magnitude (duration or level) ... it is possible the the processor is forced to reset (most noticeable with sound locos!) ..... TO AVOID the risk of this, many decoders now have a 'Stay Alive' function involving large value capacitors which can maintain at least enough power for the processor, if not the motor, to continue 'as normal' over a short-term loss of power (whether dead-frog, or dirt on track). Then, in turn, to cope with an increasing number of sound fitted locos, and illuminated carriages, devices like the PS-X are(/can be) used to 'help' apply the power to the track without having a sudden surge of current which would cause the controller to rest on overload!! To 'summarise': 4-wheel locos: try to fit 'stay alive' decoders. On others, pickup from all wheels has been the common option .... nowadays?.... use both if possible, and space allows. 'Dead Frogs' are a problem of the past. track still benefits from cleaning. Link to post Share on other sites More sharing options...
Sunnysa Posted October 21, 2014 Share Posted October 21, 2014 To put it in simple language, a DCC decoder operating on DC soaks up the first 3-5 volts to wake it up before driving the motor. Nature of the beast I am afraid Cheers Ian. Link to post Share on other sites More sharing options...
Jeff Smith Posted October 23, 2014 Author Share Posted October 23, 2014 Thanks, this is all useful knowledge. I waited until I had done some more tests on the 0-4-2 before responding because I was concerned that one of the pickups was not making good contact, however I found it difficult to tell as the coupling rods are live...so without removing them it would be impossible to determine. They all look ok. Further work required. Link to post Share on other sites More sharing options...
Jeff Smith Posted November 7, 2014 Author Share Posted November 7, 2014 Now have hooked up my new Digitrax Zephyr and tried the 0-4-2. Bachmann chips are factory-set to (or perhaps only capable of) 28 steps so I have reset the controller to 28 steps (default is 128). I have yet to programme the chip so am running on register 03. Anyway the loco still stalls on the point frogs and sometimes on plain track; cleaning the track improved it but still not good. A test run with the 2-4-4 Forney was much better, probably because of better pick-ups and more wheels on the track and more weight. Inspection of the 0-4-2 indicates that all four pick-ups are in contact with the wheels but the chassis seems pretty rigid and although the track is level I am convinced that most of the time it is only on two opposite wheels. When stalled (light off) pressing on the top of the loco usually restores the power. It is a fairly light loco. Both locos seem noisier running on DCC than on DC - is this normal? PS - have removed the wheel keeper plate and re-adjusted the pick-ups. Initially the light just was flashing with no drive but eventually got going. A few high speed runs seemed to improve it - maybe it needs to be run in..... Link to post Share on other sites More sharing options...
Phil S Posted November 7, 2014 Share Posted November 7, 2014 Not knowing the age or type of (Bachmann) decoder makes a definitve comment more difficult. Early Bachmann decoders defaulted to the 'old style' drive at low frequency, whilst later models from almost all manufacturers now default to high-frequency drive - often called Silent or similar, as it is from 16kHz or above (-32kHz?) - and out of the range of many peoples hearing - it also provides an inherantly smoother drive than the 50Hz pulse rate that originated from the time of Zero-1 (mains frequency based) diigtal controllers, and 'diigitisation' of existing analogue pulse-width controllers 8-(. (Some of the first UK Bachmann decoders also only supported basic addressing (1-99/127), as did the original Hornby basic decoder, despite extended addressing having been widely adopted by then .... neither firm made the best of their entry into the UK DCC market!!) On some Bachmann models, a slip of paper provided with the loco or decoder advised about setting it to the high frequency drive. If you have misplaced the Bachmann instructions, you may be able to download them from the Bachmann website. Current decoders have a Back EMF system which monitors the voltage generated by the turning motor during brief intervals in the drive pulse - and uses it to adjust the drive power.... HOWEVER this can require 'tuning' to get the best result - and avoid 'leapfrogging' at lower speeds. ALSO I have often found a 3rd suppression capacitor mounted from the motor to the chassis/motor case (beneath the motor) ... which can be removed when removing the other, more obvious, capacitors - removing this can improve the running too. Your best best may be: STAY ALIVE decoders help maintain running over intermittant pickup - especially from rigid 4-wheel chassis - and the Hattons version is very economical - Capapcitors bought separately with a suitable 25V minimum working voltage. Others may come with the cpaacitor already connected. Other types of decoder can have a 'stay alive' module fitted (easily made at home for those with a fine tipped soldering iron and acceptance of its removal of the decoder from warranty! HOWEVER: But there is an old wise 'chinese proverb' - stated at all places where dcc decoders are fitted: FIRST it must run well on analogue, before you can expect it to run well on DCC. Finally: It might be that, when current is being taken whilst a loco is running, that the track voltage drops - especially in the feed to the live frog - and this is causing or contributing to the 'brown out' of the decoder {The Stay-Alive may help here]. BUT Ideally, track voltage should be tested at maximum current, and in all parts of the layout: A 'Coin Test' to check the short-circuit protection capability is a good indicator here too ....does a coin or metal object placed across the track at the live frog cause an immediate protective cuttout of the controller (or other protective device such as a PSX) Link to post Share on other sites More sharing options...
Nigelcliffe Posted November 7, 2014 Share Posted November 7, 2014 Now have hooked up my new Digitrax Zephyr and tried the 0-4-2. Bachmann chips are factory-set to (or perhaps only capable of) 28 steps so I have reset the controller to 28 steps (default is 128). The setting is "28/128" or "14", there is no setting for "28" alone in the chip. The decision on whether it is 28 or 128 is made by the command station to send appropriate speed packets. So, you can revert to 128 if you wish, and may get fractionally better takeup from stop. I have yet to programme the chip so am running on register 03. Anyway the loco still stalls on the point frogs and sometimes on plain track; cleaning the track improved it but still not good. A test run with the 2-4-4 Forney was much better, probably because of better pick-ups and more wheels on the track and more weight. Inspection of the 0-4-2 indicates that all four pick-ups are in contact with the wheels but the chassis seems pretty rigid and although the track is level I am convinced that most of the time it is only on two opposite wheels. When stalled (light off) pressing on the top of the loco usually restores the power. It is a fairly light loco. DCC can't clean the wheels for you, or magically install extra pickups. Things need cleaning (cloth with spirit on it - avoid abrasives ) Both locos seem noisier running on DCC than on DC - is this normal? Probably less than optimal settings for the motor control parameters in the decoder. As I said earlier, Bachmann US Outline On30 factory sound locos are probably fitted with a Soundtraxx decoder; the Soundtraxx website has documentation on their decoders, and in that is advice on which CVs adjust motor control. There are usually a handful of CVs, and all interact with each other, so changes need to be done very systematically, taking notes as you go on better/worse. One thing the Soundtraxx should have is the brake function key. The idea of this is to set the CV4 (deceleration) to a large value, so if the speed control is set to zero, the loco will normally take ages to decelerate (as if driver shut regulator, but didn't apply brakes). Then, a function key (default is F11) is used to activate the brakes, with a smaller deceleration value to slow the loco. This behaviour should also be linked to the sounds generated (so coasts with rod-clank with regulator shut, and then brake noises appear when brake key is used ). PS - have removed the wheel keeper plate and re-adjusted the pick-ups. Initially the light just was flashing with no drive but eventually got going. A few high speed runs seemed to improve it - maybe it needs to be run in..... Shows that cleaning/adjusting is necessary. Some running in will help in "cleaning". - Nigel Link to post Share on other sites More sharing options...
Jeff Smith Posted November 7, 2014 Author Share Posted November 7, 2014 Thanks both Phil and Nigel for your advice. As you can tell I am new to DCC and have a steep learning curve to get up to speed, as it were. Incidentally none of the locos are sound fitted. If subsequently fitting sound do I add a sound decoder or replace the existing non-sound decoder? Or is it even possible that a sound decoder is already fitted so just needs a speaker? Link to post Share on other sites More sharing options...
Nigelcliffe Posted November 7, 2014 Share Posted November 7, 2014 Thanks both Phil and Nigel for your advice. As you can tell I am new to DCC and have a steep learning curve to get up to speed, as it were. Incidentally none of the locos are sound fitted. If subsequently fitting sound do I add a sound decoder or replace the existing non-sound decoder? Or is it even possible that a sound decoder is already fitted so just needs a speaker? In which case, my comments about Soundtraxx is almost certainly wrong. Sorry, I understood the locos to be the DCC-fitted-with-sound versions. No, it won't be "just add a speaker". A sound decoder is a fair expense and, in most cases, will replace the decoder already fitted; A Soundtraxx is perhaps $75-$100 upwards, something better such as a LokSound or Zimo would be more. (Another option would be TCS, I've not heard the TCS sound decoders yet). There is a piggy-back sound decoder available from Digitrax (the SoundBug) where you keep the existing motor decoder, but this is often a bit of a compromise in performance (possibly the newest version of the Digitrax chips makes it less of a compromise). - Nigel Link to post Share on other sites More sharing options...
GoingUnderground Posted November 7, 2014 Share Posted November 7, 2014 ......A Soundtraxx is perhaps $75-$100 upwards, something better such as a LokSound or Zimo would be more. (Another option would be TCS, I've not heard the TCS sound decoders yet).... ESU market the Loksound Select in North America and I believe its street price is in the USD75-USD100 range, plus tax. It is based on the LokSound V4 design, BUT, it can only be loaded with sound files specially created for the Select. It does not have Railcom, but as you have Digitrax that probably won't be a problem, and only works with DCC. REad about the Select here http://www.esu.eu/en/products/loksound/loksound-select/ and you can listen to the sounds available for the Select on the ESU web site. http://www.esu.eu/en/downloads/sounds/loksound-select/ Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.