"Keep Alive" nonsense

Erik84750 Feb 27, 2023

  1. Erik84750

    Erik84750 TrainBoard Member

    334
    134
    12
    Someone please correct me:
    1. a capacitor discharges to 63% of its initlal voltage after a time T = R * C.
    2. a typical running HO locomotive uses roughly 250mA (no lights, no sound), @15V DCC, at least
    3. meaning a average resistance of 15/0.250=60 Ohms.
    4. typical capacitor values for "Keep Alive" circuits are 10.000µF at best given its size must be fitted somewhere
    5. T = 60 Ohms * 0.01F = 0.6 seconds for the voltage to drop to 63% of 15V (9.45V)
    These all are best case assumptions?
    Conclusion: a "Keep Alive" is an illusion.
    Right or wrong, and why wrong?
     
    BNSF FAN likes this.
  2. jbonkowski

    jbonkowski TrainBoard Member

    40
    27
    16
    Well, there are two types of keep alive caps, typically.

    There is the kind that sits external to the decoder and keeps both the decoder and motor going. These often have much larger values than in your calculation. Digitrax has products that are either 0.16 F or 0.33 F, which is over 5 seconds of keep alive time in your calculated scenario.

    The other kind of keep alive is integrated into the decoder board. These are much smaller values of capacitance, but they are not meant to keep the motor going. The purpose is to keep the decoder "booted up" during power interruption, mostly from dirty track but also non-powered turnout frogs, etc. The amount of current needed to keep the decoder going without the motor is much, much smaller. Even if one of these only keeps things on for 0.6 seconds, that's an eternity in electronics time, keeping the decoder "up" through dozens of events that would have caused a decoder "reboot". A motor flywheel is usually sufficient to keep the engine moving through the trouble area until power to the decoder is restored.

    Jim
     
  3. HemiAdda2d

    HemiAdda2d Staff Member TrainBoard Supporter

    22,014
    27,407
    253
    I'm watching this one, as I don't know how it works. I'm also not an electrical engineer, so there's that, too.
     
    Hardcoaler and BNSF FAN like this.
  4. BigJake

    BigJake TrainBoard Member

    3,259
    6,172
    70
    Not wrong, just incomplete...

    The purpose of the keep-alive is not to keep the train moving at speed, it is to keep the sound decoder from resetting, and interrupting/restarting the sound, which is much more noticeable/irritating to the observer. The locomotive's flywheel momentum can keep the motors spinning for a while anyway, reducing the loss of train speed.

    Also, the decoder can keep its wits down to a much lower voltage than that necessary to keep the train running at speed. Therefore, 63% of nominal tack voltage is not a realistic threshold for the electronics/audio to stop. More like 3V or less (e.g. 25% or less of nominal track voltage). Perhaps the audio volume drops a bit, but that's far less noticeable than stopping/restarting.
     
  5. Erik84750

    Erik84750 TrainBoard Member

    334
    134
    12
    Can you provide a reference please?
     
    BNSF FAN likes this.
  6. Erik84750

    Erik84750 TrainBoard Member

    334
    134
    12
    I stand corrected on my assumption in point 5:

    5. T = 60 Ohms * 0.01F = 0.6 seconds for the voltage to drop BY 63% FROM 15V (5.5V)

    By that time the motor is not capable of running anymore since engine friction will cause it to stall. But current still is flowing through the 60 Ohms (optimal case, in reality probably a lower value).

    The size of a 25V 10.000µF capacitor is about 18x40mm (https://tinyurl.com/ypzkjwjh). I would not know how to fit this comfortably in a standard HO locomotive.

    Objections?
     
    BNSF FAN likes this.
  7. Sumner

    Sumner TrainBoard Member

    2,798
    5,837
    63
    That is why I'm adding them to the locos that will get sound installs. Here is a video with a 940uF Run-N-Smooth PowerKeeper in a N scale loco...



    I have the parts to assemble a similar one and will give it a go. Some of the installers that I've watch their work have said they get similar results doing about the same with ones in this size.

    Sumner
     
  8. wvgca

    wvgca TrainBoard Member

    499
    305
    21
    well, your 'fact' number two is possibly wrong ... average power draw [for an ho loco] -may- be up to 250 for an old Athern , but is in all probability around 150ma [i checked a couple of newer ones myself], and the voltage is not 15 volts [to the motor], it's actually around eight or so [unless you run wide open] in DC to the motor ... remember it's not drawing fifteen volts to the motor , it's just what the track happens to have ..
     
    BNSF FAN likes this.
  9. BigJake

    BigJake TrainBoard Member

    3,259
    6,172
    70
    The 60 ohms "resistance" assumption is based on current draw at wide open throttle, and I assume the rotor is not locked.

    Back-EMF (the voltage generated by the spinning motor), and motor load & friction, establishes the motor speed for a given driving voltage. Back-EMF is proportional to the motor's rotor speed (RPM). At low RPM, BEMF reduces (to zero if rotor stalls), and current increases (if driving voltage is constant) to the limit determined by winding resistance and whatever BEMF voltage is present from the spinning rotor.

    Motor winding resistance can only be measured with the rotor locked (0 RPM), to eliminate the effects of BEMF in the measurement.

    Therefore the armature resistance is actually less than your calculation.
     
    Mark Ricci likes this.
  10. Massey

    Massey TrainBoard Member

    1,982
    6,258
    53
    60 Ohms seems a bit high for a decoder, even including the motor, but I cant speak for what the actual measurement really is.

    Now a capacitor will discharge up to but not always 63% in it's first cycle. Sometimes it's far less than 63%, but it will keep discharging until the charge has been depleted. So it's 63% of 100% the first cycle, then 63% of 37% the second and so on. So the keep alive time would depend mostly on the overall draw or load on the capacitor. A capacitor can keep a circuit alive for hours depending on the draw and capacity of the capacitor.
     
  11. Erik84750

    Erik84750 TrainBoard Member

    334
    134
    12
    Hi Sumner, that is a pretty convincing video. For N-scale. What is the size (both in dimensions and in Coulomb) of the capacitor in use?
     
    Mark Ricci likes this.
  12. Erik84750

    Erik84750 TrainBoard Member

    334
    134
    12
    If Keep-Alives only work properly in "newer" loco's than that fact should be stated as a prerequisite in the Keep-Alive specifications.
    15V is the actual nominal voltage in use for DCC. At any speed.
    You refer most probably to non-DCC use?
     
  13. Erik84750

    Erik84750 TrainBoard Member

    334
    134
    12
    Hi Andy, the 60 Ohm resistance is based on DCC current draw at average speed. This value surely changes depending on the speed.
    So it may be stated that current draw at higher speeds is higher, yet it would also allow a loco to drive "by" (or over, or.. call it "inertia") a spot where bad contact between trucks and rail is present. And at low speed surely less current used, less inertia, yet less draw from a Keep-Alive capacitor.
    So far so good.

    True, yet with the understatement that I use an "average" value dependent on the current consumption of a hypothetical loco.
    But my statement refers to the amount of time a Keep-Alive would actually be able to supply the energy required to bypass a dead spot on the rails taking into accout actual current usage at the time of the "dead spot".
    And so far I still am talking about milliseconds. Unless a cap is used that provides a headache trying to fit it into a locomotive?

    At this point I see a convincing video by Sumner, yet I wonder what size cap is used, and what the actual current draw is at the time of his video.
     
    Mark Ricci likes this.
  14. wvgca

    wvgca TrainBoard Member

    499
    305
    21
    keep alives are more efficient in lower draw locos, or as you phrase it, newer locos ..
    and the voltage figure, as previously stated, is what the motor sees, DC while the track may be DCC
    disclaimer: i have NOT bought any commercial stay alives, however i have built over fifty home made ones
     
  15. tjdreams

    tjdreams TrainBoard Member

    473
    607
    24
    Well they may or may not work in HO scale, But in Z scale they keep my loco moving for 2 or 3 seconds which is more than enough time for the wheels to pass over a dead spot on the tracks. And They keep my Sound Car playing for 30 + seconds after its been removed from the track.
    AZL R-70-15 Reefer Soundtraxx Tsunami Sound Car decoder Soundtraxx Current Keeper Railmaster Hobbies DS1240-Box speaker

     
    Kez and BigJake like this.
  16. BigJake

    BigJake TrainBoard Member

    3,259
    6,172
    70
    "Coulomb" is a unit of electrical charge, not capacitance.

    Capacitance is measured in units of farads, which are Coulombs per Volt (amount of charge stored per Volt of potential).
     
    Erik84750 and Mark Ricci like this.
  17. Sumner

    Sumner TrainBoard Member

    2,798
    5,837
    63
  18. Sumner

    Sumner TrainBoard Member

    2,798
    5,837
    63
  19. wvgca

    wvgca TrainBoard Member

    499
    305
    21
    noticed you went to 16 volt on the caps .. any particular reason, or you just had them laying around ?? they are rather pricey at $5.59 Cdn in quantity 10
     
  20. Sumner

    Sumner TrainBoard Member

    2,798
    5,837
    63
    That isn't my parts list but Iowa Scale Engineering's parts list/schematic.

    My thought was that they did that since the physical size of the caps is the same as 20v 220's that you see some using. I'd bought 220's due to the higher voltage rating after reading suggestions that 20v was safer (never actually used them yet). After seeing their schematic and seeing they had the protection circuit built in to limit any voltage spikes I've decided to go that route as I can get 940uf in the same physical space as I was going to get with two 220's (440uf) which is really important in N scale.

    I'm also running 12v power supplies for the DCC-EX command station and boosters which should help safety wise with the caps. I wouldn't go this route if I was running a higher voltage.

    For the two caps and other circuit components I'll have about $10 in the keep-alive per loco. Will probably end up thinking why didn't I spend $20 for theirs that are already assembled :(. In the 50 or so you've done have they been for N, HO or something else and what total capacitance have you been going with?

    Sumner
     
    Last edited: Mar 1, 2023
    Erik84750 likes this.

Share This Page