Basic circuit questions

[garethashenden]

Electricity isn’t my strength and I have some questions that are probably pretty basic to some of you.
I’m currently installing a  Loksound 5 Micro in an Arnold SW1. I got it installed and it works,  it definitely need a capacitor. The diagram on ESU’s website says to include a resistor in the capacitor circuit, is that actually needed?

Now for the really basic questions. Does it matter which pole of an LED has the resistor attached or is its presence somewhere in the circuit good enough? Can I attach one resistor to the common lighting wire or do I need one for each headlight?

[C855B]

Yes, the resistor is absolutely necessary with the capacitor, otherwise it will want to "dump" its stored energy all at once, defeating the keep-alive function.

Resistor on the LED is fine on either pole (a/k/a "lead", pronounced "leed"). If the headlights or other lighting are never on at the same time you can use a common resistor, but if you set the CVs for dim-in-reverse, you won't get the lighting levels you expect. Use one resistor per LED to be safe.

[ednadolski]

For any interested folks:

https://www.tigoe.com/pcomp/code/circuits/understanding-electricity/
https://science.howstuffworks.com/environmental/energy/circuit1.htm

[Steveruger45]

Electricity isn’t my strength and I have some questions that are probably pretty basic to some of you.
I’m currently installing a  Loksound 5 Micro in an Arnold SW1. I got it installed and it works,  it definitely need a capacitor. The diagram on ESU’s website says to include a resistor in the capacitor circuit, is that actually needed?

Now for the really basic questions. Does it matter which pole of an LED has the resistor attached or is its presence somewhere in the circuit good enough? Can I attach one resistor to the common lighting wire or do I need one for each headlight?

Page 43 of the loksound 5 manual shows a 2200uF cap with not only a resistor but also a diode.
The resistor and diode would be needed for such a large capacitor.
It has been common practice to install stay alive capacitors without the resistor and diode Iro between about 400 - 600 uF  in the v4 micro and select.  These are not true keep alive or power packs and will not power the whole locomotive for any noticeable time.  They do help with keeping the sound going at a momentary power interruption.
LED’s.... in a series circuit it doesn’t matter on the sequence of the components, you can put your resistor on either positive or negative end of the LED.  The Positive and negative orientation of the LED is important in the circuit though, the LED is polarized, if you get it backwards it won’t work.

[peteski]

Yes, the resistor is absolutely necessary with the capacitor, otherwise it will want to "dump" its stored energy all at once, defeating the keep-alive function.

I wouldn't say "absolutely necessary", but it is good idea to limit the current.  Actually it is the initial capacitor charge current that needs limiting, not its discharge current.  The available discharge current of a keep-alive circuit should be as high as possible.

A capacitor that is not charged will appear a a dead short when the power is first applied to it.  The ideal capacitor that is, but even real-world capacitors will be close to a dead short. Then as the capacitor charges up, the current drops until it charges up to the same voltage as the DC power source. At that time no current will flow either way.  Because the initial current is so high, if we were to hook a large-value cap directly across the output of the decoder's bridge rectifier, that would place a strain on the rectifier (because the initial charge current would be so high).  That is why it makes sense to limit the charging current by installing a resistor in series with the keep-alive cap.

But when the track power is interrupted, and the capacitor begins to supply the power to the loads (decoder, motor and headlights), it has to provide all the current it can for those devices to stay running. The loads all have internal resistances which will only us as much current as they need.  If you limited the cap's discharge current too much (with a series connected resistor), that would result in voltage drop across that resistor. If the voltage drop was too high, then the load devices would start stop working.

Many  keep-alive circuits use both, a resistor and a reverse-connected diode in parallel with the resistor.  This way when the cap is charging, the diode is not conducting and the resistor is limiting the charging current. When the capacitor is called to supply power, the current flowing in the opposite direction will bypass the resistor, so a higher current can be supplied to the loads on the decoder.  This is routinely done on keep-alives with very high capacitance (thousands or hundreds of thousands of micro Farads).

Some more basic keep-alives which have lower value caps can work with just a properly chosen resistor in series with the cap.  Its value will be high enough to prevent excessive initial charging current, yet low enough not to drop too much voltage to supply some useful power to the decoder/motor/headlights when track power is lost.

Then sometimes a coil is used in series with the cap instead of a resistor. Coil's electric properties are opposite of a capacitor's (its resistance is high when the current is changing - like when a capacitor is first getting charged).  There are also some hybrid designs using coils, diodes, and resistors to control the current in keep-alives.

Resistor on the LED is fine on either pole (a/k/a "lead", pronounced "leed"). If the headlights or other lighting are never on at the same time you can use a common resistor, but if you set the CVs for dim-in-reverse, you won't get the lighting levels you expect. Use one resistor per LED to be safe.

Easy thing to remember is the rule that the location of each component (their order) is irrelevant in a series-connected (or daisy-chained) circuit. 

[Point353]

If the voltage drop was too high, then the load devices would start working.

Care to elaborate?

[peteski]

Care to elaborate?

LOL, I meant "stop". I fixed my post.  I hope that makes more sense now.

If the charged capacitor is the source of power, and it powers the rest of the locomotive (the electrical load) through a resistor connected in series, if that resistor's resistance it too high, most of the voltage will be across the resistor, and the decoder (the electric load which needs the power) will only see a small part of the voltage, and the current will also be limited, based on the combined resistance of the series resistor plus whatever the load's resistance is.

[Point353]

LOL, I meant "stop". I fixed my post.  I hope that makes more sense now.

Infinitely.

[garethashenden]

Ok, reviving this thread from the dead as I have more questions.

I have the capacitor wired in and everything works well, it definitely helps keep the locomotive moving and the sound on. The only problem is that I can't read any CV values from the decoder when its wired in. I think its pulling too much power for the LokProgrammer to handle. Disconnecting the capacitor solves the problem. So do I need to add both the resistor and diode or will just one of them solve the problem? The diode in ESU's manual, 1N4007, is rather large at 4mm long and 2mm diameter. Is there something smaller that will do the job?

Thanks!

[peteski]

Ok, reviving this thread from the dead as I have more questions.

I have the capacitor wired in and everything works well, it definitely helps keep the locomotive moving and the sound on. The only problem is that I can't read any CV values from the decoder when its wired in. I think its pulling too much power for the LokProgrammer to handle. Disconnecting the capacitor solves the problem. So do I need to add both the resistor and diode or will just one of them solve the problem? The diode in ESU's manual, 1N4007, is rather large at 4mm long and 2mm diameter. Is there something smaller that will do the job?

Thanks!

Diode and resistor circuit is recommended (and required for higher capacitances (like 500uF and up).  Diode/resistor work together.  Resistor limits the capacitor's charging current, while the diode allows higher current to flow from the capacitors to the decoder when it is supplying power (while track power is interrupted). Installing in inductor (coil) in series with the capacitor also often takes care of the programming problem. That solution has been mentioned in past posts here.

1N4007 is a very popular and inexpensive rectifier diode, but a bit of an overkill for N scale.  There are plenty of physically smaller (SMD) diodes out there.  I prefer Shottky Rectifier Diodes, because they have smaller voltage drop than Silicon diodes (like 1N4007). 

Here are some Digikey part numbers:
Diodes: 1727-3882-1-ND,  SS16HECT-ND
Inductor: 587-1628-1-ND

[RBrodzinsky]

Whether you add the diode or resistor, it does not solve the LokProgrammer issue. ESU decoders cannot be programmed on the LokProgrammer with any stay alive circuitry, other than the ESU Power Pack, attached.  That said, Nolan Deardorf just showed a circuit on the ESU Facebook group that can allow LokProgrammer programming. Uses a solid state relay and one of the powered Aux outputs to switch the U+ feed to the capacitor off/on

[peteski]

Whether you add the diode or resistor, it does not solve the LokProgrammer issue. ESU decoders cannot be programmed on the LokProgrammer with any stay alive circuitry, other than the ESU Power Pack, attached.  That said, Nolan Deardorf just showed a circuit on the ESU Facebook group that can allow LokProgrammer programming. Uses a solid state relay and one of the powered Aux outputs to switch the U+ feed to the capacitor off/on

Solid state relay? In N scale space is a premium.  Adding a coil often resolves the problem (and those coils are small).

[RBrodzinsky]

https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/tlp3406s-tp-e/6200250

I probably used the wrong terminology.

He posted this diagram

[ Guests cannot view attachments ]

[peteski]

That is a interesting device.  Photo-relay. Looks like "solid-state relay" would be a correct name for it.  I'm not used to seeing solid state relays that small.

That will work as long as that AUX output is turned on during normal operations, and off during programming.

[Point353]

Should there also be a small value resistor connected between the capacitor and the photo-relay to limit the current to the 1.5A maximum specified in the datasheet?