Mysteries of the new LokSound in the Intermountain SD40-2

[AKNscale]

mu: yup, you are correct

I can't find my multimeter to verify the pads stating AUX3&4, but my guess is that they are actually 3&4 as the manual does include the head and taillight as 2 functions(for some reason I was thinking otherwise) for the Select decoders. Now why they would have preset LEDs to these functions that are in random places(such as AUX4) your guess is as good as mine. Tomorrow morning I will get some answers/clarification. I will also be looking to get some clarification on what the 4 round solder pads(that's what they look like) are there for.

[jdcolombo]

Ok, quick update: I've gotten all lights to light up, these are their factory settings...

-The front LED on the underside of the decoder lights up on F6 forward, and is set to AUX1 while the rear lights up on F6 in reverse and is set on AUX2(I think someone has already figured this)
-The front taped over LED lights up on F5, is preset for a beacon, and is AUX3(also figured)
-The rear taped over LED lights up on F13, is preset as a dimmable headlight, and is AUX4

The mysteries are being solved . . .

John C.

[AKNscale]

Ok, so today I talked to both Ron and Matt, which I must say are awesome to take the time out of their day to talk to me. Kudos to ESU for sure. Sadly, they are still waiting for a schematic from Germany for this decoder so that our questions can be answered as they are wondering the same as we are. One thing Matt was able to tell me is that this decoder was designed to not need any outside caps as the engineers in Germany claim that our locos are too small to fit. Matt and I both completely disagree with that, but until a schematic is provided, the mysteries continue. I will check in with them about the schematic again shortly. I will also add ditch lights to this model soon and of course record as I go.

Side note, Matt wanted me to let everyone know that the new drop in decoders will be here very soon(they should be in production now) and I believe he said they will be showing them at Springfield?

[peteski]

Ok, so today I talked to both Ron and Matt, which I must say are awesome to take the time out of their day to talk to me. Kudos to ESU for sure. Sadly, they are still waiting for a schematic from Germany for this decoder so that our questions can be answered as they are wondering the same as we are. One thing Matt was able to tell me is that this decoder was designed to not need any outside caps as the engineers in Germany claim that our locos are too small to fit. Matt and I both completely disagree with that, but until a schematic is provided, the mysteries continue. I will check in with them about the schematic again shortly. I will also add ditch lights to this model soon and of course record as I go.

Side note, Matt wanted me to let everyone know that the new drop in decoders will be here very soon(they should be in production now) and I believe he said they will be showing them at Springfield?

It would be nice to have a schematic diagram and the PC Board layout, but I won't hold my breath that it will be released to the public.  Matt might give you a hint as to where to hook up the negative lead of the capacitor. But that is something that could be determined in about 10 minute of examining the decoder board and using ohmmeter.

[jdcolombo]

AKNscale:

Thanks for taking the lead on this; I'm not terribly surprised that Matt doesn't know much more at this stage than we do - in fact, given your programming explorations today, I think we probably know more (though I'd still like to know about those Aux3, U+ and Aux4 pads). 

The really, really good news is about the new drop-in.   Can't wait to see THAT board!

As for the keep alive - well, the Germans' view is just wrong.  And that makes more imperative the second part of the mystery project, which is finding a way to add an external keep alive. 

Are you going to try an alternating ditch light project?  I thought I'd undertake this when mine arrives (probably Friday), but I'm happy to let you take the lead here.  Since I have no real use for an SD40-2 (too new for my 1957 NKP layout), I would do this only as "proof of concept" for other people.  But there's no need to duplicate efforts if you're going to do it anyway.

Peteski:
I've ordered an SD40-2 which should be delivered Friday.  I want to do some playing with it, but when I'm done I can either mail you the board for you to "inspect" or you can tell me what to look for and I can go poking around with my multimeter.  If I know what I'm looking for, I'm a pretty dogged sleuth . . .

John C.

EDIT:  So after re-reading your comments about the Germans and keep alive, I think that what this probably means is that the folks in Germany don't think there is room in an N scale diesel for something like the ESU PowerPack (they are right about that) and therefore designed the decoder with SOME keep alive on board (I still suspect those two dark yellow caps near the bottom center of the board) and no officially-sanctioned connections for external keep alive.  If those are two 20v 150uf tantalum chip caps for on-board keep alive, then I can see the point of the folks in Germany, since I usually don't use more than two 220uf caps for my after-market installations.  Still, folks might be interested in adding more, and it looks to me like you could stuff three or four more tantalum chip caps on the shelf under the decoder board in the rear.  Will have to see when I get mine.   People seem to experience fewer drop-outs once the wheels are thoroughly clean, so maybe the German engineers are more right than I would first admit.  Still, it's just not that big a problem to offer pads for external keep alive for those intrepid souls that want to go that route, and so I don't know why they wouldn't just do it.

Meanwhile, someone on another board has noticed that the SD40-2 doesn't have the "Prime Mover Startup Delay" typical of other LokSound Select decoders (this is the behavior where the engine throttles up to Notch 1 before moving.  It is certainly more prototypical in some cases, although it makes it impossible to consist a sound loco to a non-sound loco unless you turn this behavior off).  Prime Mover Startup Delay is usually controlled by bit 3 of CV 124, but the person in question read the value in CV124, and it appeared to have bit 3 set to "on" - yet no delay.  Every LokSound decoder I've installed has had this behavior (I have to turn it off on my road units, because I consist them with non-sound  units); it's very odd to me that the SD40-2 wouldn't have this "feature" - perhaps Intermountain didn't want it because it does befuddle some folks, but even if it were turned off at the factory, you'd think you could enable it by normal CV programming.  I'll have to check that one out, too, when I get my unit.  Just seems weird that ESU would eliminate that (purely programming, not a hard-wire-component issue) in this engine.

[peteski]

Peteski:
I've ordered an SD40-2 which should be delivered Friday.  I want to do some playing with it, but when I'm done I can either mail you the board for you to "inspect" or you can tell me what to look for and I can go poking around with my multimeter.  If I know what I'm looking for, I'm a pretty dogged sleuth . . .

John C.

EDIT:  So after re-reading your comments about the Germans and keep alive, I think that what this probably means is that the folks in Germany don't think there is room in an N scale diesel for something like the ESU PowerPack (they are right about that) and therefore designed the decoder with SOME keep alive on board (I still suspect those two dark yellow caps near the bottom center of the board) and no officially-sanctioned connections for external keep alive.  If those are two 20v 150uf tantalum chip caps for on-board keep alive, then I can see the point of the folks in Germany, since I usually don't use more than two 220uf caps for my after-market installations.  Still, folks might be interested in adding more, and it looks to me like you could stuff three or four more tantalum chip caps on the shelf under the decoder board in the rear.  Will have to see when I get mine.   People seem to experience fewer drop-outs once the wheels are thoroughly clean, so maybe the German engineers are more right than I would first admit.  Still, it's just not that big a problem to offer pads for external keep alive for those intrepid souls that want to go that route, and so I don't know why they wouldn't just do it.

I suppose I could direct you where to poke around. I already made some speculations on the fuzzy photo in the other thread.  I guess if I had a really sharp and higher-res photo of the board, that would make it even easier.  But having the board in-hand would be the quickest.

As far as those caps go, the photo is not clear enough to verify their values.  For all we know they might be 220 uF or even larger, but only rated for 6.3V (and used for filtering the low-voltage supply for the micorprocessor).

[AKNscale]

It would be nice to have a schematic diagram and the PC Board layout, but I won't hold my breath that it will be released to the public.  Matt might give you a hint as to where to hook up the negative lead of the capacitor. But that is something that could be determined in about 10 minute of examining the decoder board and using ohmmeter.

I was told that they'd email it to me so we'll see

JD: Yeah I can't wait for the new boards either!
Yes I will be doing the alternating ditch lights fairly shortly. I'm like you in that my timeframe is the '80s so no ditch lights, but, this will be ran at shows so I want to do it. And for S&G, I'll probably try to use the AUX4 pad for the alternate ditch light, to see what the result gives me.

[nstars]

Meanwhile, someone on another board has noticed that the SD40-2 doesn't have the "Prime Mover Startup Delay" typical of other LokSound Select decoders (this is the behavior where the engine throttles up to Notch 1 before moving.  It is certainly more prototypical in some cases, although it makes it impossible to consist a sound loco to a non-sound loco unless you turn this behavior off).  Prime Mover Startup Delay is usually controlled by bit 3 of CV 124, but the person in question read the value in CV124, and it appeared to have bit 3 set to "on" - yet no delay.  Every LokSound decoder I've installed has had this behavior (I have to turn it off on my road units, because I consist them with non-sound  units); it's very odd to me that the SD40-2 wouldn't have this "feature" - perhaps Intermountain didn't want it because it does befuddle some folks, but even if it were turned off at the factory, you'd think you could enable it by normal CV programming.  I'll have to check that one out, too, when I get my unit.  Just seems weird that ESU would eliminate that (purely programming, not a hard-wire-component issue) in this engine.

If you use the Loksound Pilot decoder in the non sound unit then you can also get the delay in the non sound unit. I have assigned this delay to function buttten 8. So if I consist a non sound unit to a sound unit I have to switch on F8 on both units. There is only one issue with this, the value used for the delay in the Lokpilot has to be different from the value for the delay in the Loksound unit, but this is just a bit of trial and error. However, it would have been more logical if both would have been the same. With sme trial and error I have been able to get the delay the same for all units in the consist.

Marc

Marc

[jdcolombo]

I was told that they'd email it to me so we'll see

JD: Yeah I can't wait for the new boards either!
Yes I will be doing the alternating ditch lights fairly shortly. I'm like you in that my timeframe is the '80s so no ditch lights, but, this will be ran at shows so I want to do it. And for S&G, I'll probably try to use the AUX4 pad for the alternate ditch light, to see what the result gives me.

AK:
Might I suggest that instead of using the Aux4 pad, you cut the rear ditchlight and use it?

There's a reason for this suggestion - The headlight, rear light, Aux1 and Aux2 both have secondary references in the LokSound function programming.  If you look at the dropdown boxes for the function key assignments, you'll see something like "Aux1(1) and Aux1(2)"  - same for the headlight, rear light, and Aux2.

ESU does this so that you can use a logic override for differential behavior for these outputs.  Let's say, for example, that you want the headlight/rear light to dim when the engine is stopped but not when the engine is running.  What you can do is add a function assignment to F0 for "headlight(2)" that tells it to dim when the engine is not moving.  So you end up with headlight (1) on full when moving, and headlight (2) dim when not moving.

Now think about the ditch lights.  Suppose you want them on steady whenever the headlight is on, but want them to alternate when the horn button is pushed or the bell is turned on.  You can do the same thing.  You could assign Aux1(1) and Aux2(1) to come on when F0 is turned on (ditch lights on whenever headlight is on).  Then you could assign Aux1(2) and Aux2(2) to blink, with Aux1 set to blink out-of-phase, so that you get alternating ditch lights.  You could assign this to a separate function key, or assign it to F1 or F2 so that it happens automatically when the bell or horn is pressed.

The logic would be something like this:
Ditch lights on when F0 is on.
Ditch lights alternate when F0 is on AND F1 is on (or F2 or whatever).

Note that Aux 3 and Aux 4 do not have these secondary references in ESU's function mapping.  It doesn't matter if you don't want the ditch lights to do two different things (e.g., be on steady some times, but blink other times), but if you want to program a single physical output to do different things at different times, you need to use one of the outputs that has an alternate reference.

At least, I think this is the way it works - you might want to do some background research to confirm, since I've never actually done this myself.

John C.

[AKNscale]

This is true JD, I also wouldn't have to worry about the resistor if I do that. I was just trying not to run wires from the back, but I think you're right about it being the smarter choice. Plus, this model won't have rear ditch lights so there's no reason not to use AUX2 what it is normally setup for.

[jdcolombo]

More updates (I got my unit today).

1.  The Aux3 and Aux4 pads on the bottom FRONT (I thought they were at the rear, but they are at the front) are, indeed, just that: additional connections for Aux3 and Aux4 that mirror the "hidden" LEDs already on the board.  The U+ pad is common positive (blue wire on a wired decoder).  There is no resistor connected to these, so if you want to use an LED connected to the pads, you'll have to supply a dropping resistor.

2.  There is NO keep alive on board that I can see.  The caps that I thought might be keep alives are 6.3v 107j tantalum chip caps, so they aren't going to be keep alives (Peteski - thoughts?).

Will work next on seeing if I can find a common negative connection; if I can do that, you could add external keep alive, BUT - I think you'd have to mill the frame to do it.  Not enough room under the board to fit a tantalum chip cap of sufficient capacitance - I don't think.  Will experiment some more.

3.  It's going to be interesting to see what the "generic" board looks like.  This thing has lots of pickup pads and could obviously fit other locomotives, but the length might preclude putting a speaker on the top of the frame and limit one to a fuel-tank-mounted speaker, like what Intermountain has done.  I just don't know - there doesn't seem to be a way to shorten this to much less than 60mm; if you put something that length in a GP9, you're not going to have enough room for a speaker unless you can really offset the board to one end.  Guess we'll just have to wait & see.

John C.

[peteski]

More updates (I got my unit today).

1.  The Aux3 and Aux4 pads on the bottom FRONT (I thought they were at the rear, but they are at the front) are, indeed, just that: additional connections for Aux3 and Aux4 that mirror the "hidden" LEDs already on the board.  The U+ pad is common positive (blue wire on a wired decoder).  There is no resistor connected to these, so if you want to use an LED connected to the pads, you'll have to supply a dropping resistor.

2.  There is NO keep alive on board that I can see.  The caps that I thought might be keep alives are 6.3v 107j tantalum chip caps, so they aren't going to be keep alives (Peteski - thoughts?).

I suspected that those caps were not what you expected.

Can you take clear close-up photos of both sides and post them here?  That'll be a good start of the reverse engineering process. 

[jdcolombo]

I suspected that those caps were not what you expected.

Can you take clear close-up photos of both sides and post them here?  That'll be a good start of the reverse engineering process. 

Top:



Bottom:



Front of board is to the left in both photos.

The originals are on photobucket:

http://i613.photobucket.com/albums/tt217/jdcolombo/Intermountain%20SD40-2%20LokSound/IMG_4845_zps8dx7ntfe.jpg

http://i613.photobucket.com/albums/tt217/jdcolombo/Intermountain%20SD40-2%20LokSound/IMG_4849_zpsti7dc0vb.jpg

John C.

[RBrodzinsky]

If you use the Loksound Pilot decoder in the non sound unit then you can also get the delay in the non sound unit. I have assigned this delay to function buttten 8. So if I consist a non sound unit to a sound unit I have to switch on F8 on both units. There is only one issue with this, the value used for the delay in the Lokpilot has to be different from the value for the delay in the Loksound unit, but this is just a bit of trial and error. However, it would have been more logical if both would have been the same. With sme trial and error I have been able to get the delay the same for all units in the consist.

Marc

Marc

Marc - how are you doing this?  Are you simply calibrating acceleration so the non-sound unit takes a long time to start rolling (and turning acceleration on/off with F8), or is there some other delay setting you have discovered? Besides the fact that the table for CV124 in the LokSound Select manual has the bits labeled incorrectly (the first bit (value=1) is labeled as Bit 1, not Bit 0), there is no bit in the LokPilot for startup delay (why would there be, there are no engine notches).  I can fully understand having to program different acceleration parameters, as each unit (and decoder) may be slightly different.

[jdcolombo]

Marc - how are you doing this?  Are you simply calibrating acceleration so the non-sound unit takes a long time to start rolling (and turning acceleration on/off with F8), or is there some other delay setting you have discovered? Besides the fact that the table for CV124 in the LokSound Select manual has the bits labeled incorrectly (the first bit (value=1) is labeled as Bit 1, not Bit 0), there is no bit in the LokPilot for startup delay (why would there be, there are no engine notches).  I can fully understand having to program different acceleration parameters, as each unit (and decoder) may be slightly different.

From Matt Herman on the Yahoo LokSound list:

The second part of the question about “startup delay” is available now in both the LokPilot V4 and the LokPilot Standard. It is a logic feature called “Virtual Drive Sound”. You must enable this logic feature in the function mapping on the same button as your sound locos. (Usually F8 in North America and Australia) From there you can adjust CV 252 to the amount of time needed to match the motion of the sound loco. If you are using the LokProgrammer, this is the last item in the “Driving Characteristics” menu if a LokPilot File is open.

John C.