ESU LokSound 5 58721, 58731, 58751, 58941 decoded

[skytop35]

Wonderful job Peteski, as always. 

I used this 58731 decoder in Kato’s GS4, using Kelley’s approach where the front smd led’s were replaced with bright white versions, and a rear led was installed in the tender.  I ventured into ESU’s flow-charting territory, and managed to configure the over-under front headlamps such that the first push of the ‘Headlight’ button turns on the lower lamp, the second push turns on the upper lamp, the third push activates the upper lamp as a gyro, and the fourth push turns both led’s off.  Sequential toggling of the one button is a rather elegant way to achieve numerous lighting effects without having to use several separate functions.  I have since implemented the same approach on a Kato E8 diesel with over/under headlamps.

I also used a 58731 to put sound in an N scale Model Power 4-6-2, and replaced the pathetically dim factory light bulb with an 0402 led. 

It’s a good little decoder for locos that don’t have any more than 3 lighting features, hence, well suited for steam.

I would very much like to learn how to do this. Can you post a screen shot and maybe a few pointers?

[Dwight in Toronto]

On another note, I just managed to drill & tap those frame halves, as per that other thread that folks were advising me on.  I’ll do a quick update in a bit, and will try to attach a photo (although previous attempts at posting pics have not turned out well).

[Dwight in Toronto]

I would very much like to learn how to do this. Can you post a screen shot and maybe a few pointers?

Skytop - I will try my very best to oblige, hopefully later today, but have to run out for a bit. 
I may need help posting pics, because previous attempts didn’t work out very well for me!

[skytop35]

Skytop - I will try my very best to oblige, hopefully later today, but have to run out for a bit. 
I may need help posting pics, because previous attempts didn’t work out very well for me!

No hurry at all. Although please start a separate new thread.

[peteski]

Lol … the name is DWIGHT … like in Eisenhower …

Oops!  Fixed.  My brain is not having much fun lately.  Hopefully it is just  stress, and nothing more serious.

[peteski]

Placeholder for the Loksound 5 wired nano.

[peteski]

ESU 58941 Loksound 5 decoder for "wide body" Kato locomotives

ESU58941_Sound_Decoder.png

This decoder consists of the ESU Loksound nano decoder with E24 pin interface daughterboard, installed on a Kato-outline motherboard. This arrangement provides access to all of the nano's functions.  There is also an on-motherboard amplifier circuit which converts AUX3 and 4 logic level outputs from the nano to full-power AUX functions.


ESU58941_Sound_Decoder_Pads.png

The solder pads are well labeled, and for clean installs ESU included a +U (12V) pads next to each full power AUX output pad.

While my worries are likely unfounded, I'm concerned about the way the bottom of the decoder is laid out in the area of the track power pickup strips.  The metal strips in the locomotive will be right against the bottom sides of the decoder where traces are located on the decoder board.  In addition tot he typical solder mask coating, ESU applied a layer of off-white coating which is supposed to act as insulator, but for extra protection Kapton tape could be applied to the orange-colored areas.  However the extra thickness of the tape could prevent solid contact between the metal pickup strips in the loco with the exposed gold plated pads on the decoder.  Use your own judgment as to whether to apply the Kapton tape.

Judging by the arrangement of the solder pads it appears that ESU dedicated the AUX 10 logic output for the ESU PowerPack control.  Remember to use LokProgrammer to assign AUX 10 as Power Pack Control. For 2-wire keep-alive installs just use the GND and +U pads.

I was also surprised that ESU actually installed a 0.01 µF capacitor in parallel with motor pads. It is there likely for RFI suppression. Usually decoder manufacturers recommend removing RFI suppression components (capacitors or inductors) from the motor circuit.  Obviously ESU engineers determined that using this capacitor will not negatively affect the motor operation.


ESU58941_Sound_Decoder_Resistors.png

ESU included 3900 ohm (3.9k) resistors on all the powered AUX outputs.  With a typical white LEDs (Vf=3V) and +U=12V, that will result with (12-3) / 3900 = 0.0023A or 2.3mA of current passing through the LED.  While this might result in desirable brightness of the LEDs, there are workarounds which will increase or decrease the current.

To lower the current (increase the resistance), just add another series-connected resistor to the AUX output pad.  The total resistance will be 3.9k plus whatever additional resistors value is.  For example, adding a 1k resistor will result in total resistance of 4.9k.

To lower the resistance, there are couple of ways to do it.  Due to the small size of the resistors, good soldering skills, steady hands, and soldering iron with fine soldering tip are recommended.  I also should mention that any modifications to the decoder void the warranty.

To totally bypass the on-board resistor simply solder a short length of small wire (shunt) across the resistor's end terminals.  The resistor can remain on the decoder.  Just place a piece of pre-tinned bare copper wire on top of the resistor  and with a tiny bit of solder tack the wire to the tops of the resistor's metal terminals.  Adding tiny amount of flux to the resistor's ends before soldering is helpful in getting a quick and clean solder joint.  I would use a longer piece of wire to be able to hold on to it while soldering, then trim the excess length after the wire is soldered.  Very thin bare wire will work well (like a single strand from a piece of insulated stranded wire.  Now an external resistor of any value can be added in series with the AUX solder pad.

For a cleanest and most compact install, another resistor can be piggybacked over the existing resistor.  For a total new resistance of 1k (Actually 918 ohms) piggyback a 1.2k resistor.
Few other examples:  piggyback 1.8k resistor for a new total resistance of 1.2k. Piggyback 2.2k resistor for a new total resistance of 1.4k. These examples use standard resistor values.  If someone wants to get other total resistance value, there are resistors available with a wider range of resistance values.  There are also online parallel resistor value calculators available on the Internet.  All the on-board resistors are SMD 0603 size, so the piggyback resistor should be the same size.


ESU58941_Sound_Decoder_Dimensions.png

Dimensions of the decoder are provided in case someone would like to install the decoder in a model not designed to accept this board.

While I have not tested each install, I was told that these will fit (thickness wise) in pretty much all Kato locos which use that outline light board.  E-units, F-units, F40PH, SD45, etc.
UPDATE: In this post Kelly warns that Kato PA1 and PB1 locos do not have enough vertical clearance for these decoders.

CAUTION:  Like with similar drop-in decoders from other manufacturers, it is vital to insulate the motor lead strips to prevent them from touching the metal track pickup strips on the top of the chassis. If not properly insulated, the motor driver on the decoder will likely burn up (regardless of any over-current protection built-in on the decoder).   I always found this to be a poor part of the Kato's design.

I'm surprised that ESU does not provide better technical documentation for their excellent decoders, compelling  me to create these write-ups. Their documentation is not only sorely lacking technical details, it is also incomplete (at least the English version, since I have not looked at the German documentation). It is as if they expect their customers to be quite familiar with electronic circuitry (a skill that most model railroaders do not posses).

[Steveruger45]

Thank you Pete.  Brilliant.

[peteski]

ESU 58751 LokSound decoder for legacy Atlas, IM, and other narrow-hood models

This V5 decoder is a slightly reworked version of a 73100 V4 decoder.  Most of the circuitry appears to be identical between those decoders.  Electrically, the power sections (power supplies, motor driver, audio amplifier, voltage regulator, and other ancillary circuits) appear to be identical to 73100.  The microcontroller chip, and related circuitry around it are the changed components. A new feature on this decoder is addition of solder pads for connecting ESU's Power Pack 3-wire keep-alive module.  Those pads also make it easier to add any standard 2-wire keep-alive.


ESU73100_58751_compare.png

Here is a comparison of 73100 and 58751 showing just how similar they are.



ESU_58751_Sound_decoder_dimensions.png

Dimensions of this decoder are also same as that of 73100


ESU58751_Sound_decoder.png

For those who are interested,  I'm including a partial schematic diagram of the decoder, along with access points to various voltages.  It is included it for informational purpose only, since this decoder already includes solder pads for connecting Keep-Alive module (or individual capacitors).


ESU58751_Sound_decoder_layout.png

Here is the component's layout.


ESU58751_Sound_decoder_pads.png

Identification of decoder's track and motor pads.


ESU58751_Sound_decoder_functions.png

This decoder includes full-power (not logic level) outputs for AUX3 and AUX4 with no resistors.  But no solder pads are provided for AUX1 and AUX2 outputs which control the on-board LEDs. To add external LED to those outputs, the on-board LEDs would have to be unsoldred, and the LEDs cathode pads would be used as the output pad.  Again, no resistors on the cathode side (in-line resistors are on the anode side).

The Power Pack Control (PPC) solder pad is used to control the ESU's Power Pack Keep-alive unit when the decoder is being programmed on the programming track. While in other N scale decoders (like the nano) the PPC pad has to be selected from one of the available logic-level AUX outputs, on this decoder the PPC pad is dedicated to that function.  It does not show up in the Physical Outputs section in LokProgrammer. Hardware-wise it's just an undocumented logic-level AUX output. Again, the rather sparse info in the ESU manuals does not really provide much detail, probably because it is always enabled.  If not being used there is no problem leaving it disconnected - the decoder will not be damaged.

The ESU Power Pack connects as follows: "-" is the common pad (black wire), "PPC"  pad is the control signal (white wire), and "+U" is the positive voltage (red wire). If just attaching a large value capacitor, it connects to the "-" and "+U" pads.


Instead of calling the outputs "functions" like most DCC manufacturers do, ESU calls them "AUX" outputs.  This is likely due to the fact that all these outputs can easily be mapped to any DCC function.  The output mapping feature on the ESU decoders is much more flexible than on most typical DCC decoders from American manufacturers.  I highly recommend thoroughly reading through the ESU decoder and LokProgrammer manuals to get familiar with the AUX output mappings.  While the mapping can be configured through individually programming multiple of CVs using DCC system's programming track, this task is made *MUCH* easier using the ESU's LokProgrammer interface and software, or at least JMRI DecoderPRO.

[cbroughton67]

@peteski Any chance you've "decoded" the ESU board that they produce for Atlas' N-scale S2 and now S4 locomotives? I think they could benefit from a keep-alive cap or two.

Chris

[peteski]

@peteski Any chance you've "decoded" the ESU board that they produce for Atlas' N-scale S2 and now S4 locomotives? I think they could benefit from a keep-alive cap or two.

Chris

I did take a look at it several years ago when I first bought the model, but that was before I was doing these write-ups.  I guess I can do it too. When ready I'll post it in the  in the LokSound V4 thread.

[GGNInNScale]

Hello   Really nice work.  Thanks.  I have been converting from NCE/TCS/DigiTrax and Lenz(Atlas old) to LokSound where possible.  Just tried a Hornby/Arnold U28C LokSound conversion, but the smallest decoders are about 0.7mm too wide even after shaving things down... oh well.

[dvescovoi]

What is the value of the VCC coming out of the nano E24 connector pin 18? to ground (pin 22)?
Can I assume 5.4v?

[peteski]

What is the value of the VCC coming out of the nano E24 connector pin 18? to ground (pin 22)?
Can I assume 5.4v?

Yes, that is 5.0V (actually I measured 5.04V). It seems that on some decoders ESU uses 5.4V and on some other ones 5.0V for Vcc.  But no idea how much current it supplies, so I would be careful in how it is used.   Too bad ESU doesn't provide more comprehensive technical specs.

I suspect ESU provided Vcc on E24 connector to be used for the logic circuitry on one of their larger decoders which utilizes the nano board.

Write-up of the nano (including my findings about various E24 signals) is on my long to-do list.

[dvescovoi]

Great information. Very useful. Agree a more comprehensive technical manual would be great. Too bad we have to resort to reverse engineering everything.
I was looking to using "inputs" and the 58941 (nano based) does allow A10 I/O configuration to be configured as an input ... although it seems to be called wheel sensor in the programmer and E24 pin 19.
As A11 and A12 are also logic level I/O I assume they also go directly to the micro. I am guessing these two pins also do (or plan to do) double duty as the interface for some of the other exotic EU protocols(SUSI). I am wondering if A11 and A12 can also be used as "sensor1 and sensor2" inputs? I see sporadic mention throughout the programmer and other earlier decoder documentation of sensor  inputs.