Atlas O Amtrak Dash 8 Blunami 4408 upgrade

I have some questions related to how to go about wiring up a Train America Studios AC smoke unit. This one has a thermistor to regulate temperature

I see that it has 3 wires going in, with 1 going off to a switch and coming back to a 3 wire connector (see attached pictures)

Anyone have suggestions on how to wire this up to be activated with a Blunami output? I figure it will need AC power to run it, but I have no clue now to wire it all up correctly without it going up in smoke.

I do want to keep the switch in place to be able to turn it off manually so I don't accidentally turn it on in the app with no smoke fluid in it

I happened to find exactly what I was looking for last night on OGR and it's a lot simpler than I thought to install it, even regulating it down to around 12 volts is easier than I thought, apparently a diode in series with the power feed will drop it down to 9 volts.

I might even be able to set it up to have variable output in the future, if I can find a simple solution to that to trigger the "chuff" output. Maybe 2 different relays that activate at different voltages and setting the function output to be an auto dimming headlight.

For now I will just run it without the chuff wire removed, the default fan mode should be sufficient smoke output for the time being

How about a video of it in action?

@bigtruckpete posted:

How about a video of it in action?

I'll do that once the install is finalized, I'm waiting on the arrival of the speaker, 4408, and some other components I need

@MichaelB posted:

After installing a 2200 in my Amfleet cab car,  ...snip...

Is that the Lionel version? If so, were you able to find the proper drop-equalizer trucks for it?

Sorry for the delay in updates, a lot has happened with the install and I have been working very hard on it. Got done with the sound system yesterday and also did some tests with it for the first time. Although I did not have any of the lighting hooked up yet

Today I finished wiring for the lighting, after some head scratching on how to get the class lights working properly. I found only the red bicolor LEDs would turn on when the function was active, but found a workaround to get both of them working in parrallel. A diode in series for the red LEDs, so the voltage drop is the same as the green LEDs. Which has worked great.

Below are some update pictures showing the install in its current configuration. It's a bit of a rats nest of wires, but everything all fits inside of the shell so it's not too bad yet


The black post will be used to secure a plexiglass piece for the relays. I could not think of a better place to put the relays and that area above the buck converter is just empty space, so might as well use that for something.

The frame had a giant open hole where the speaker driver lives, that was closed up with a piece of plexiglass that I glued into place. I also bought thin foam tape to seal the fuel tank to the frame very well. The speaker I used for the install is part number SC700208-1, which fits perfectly in the original speaker location in the fuel tank (even with the correct hole spacing too).

The speaker I chose isn't super strong on bass, but has the same frequency reponse as Soundtraxx's 50 X 90 mm speaker. It's also an 8 ohm 3 watt speaker, so it's perfect for the job.

And as per request of bigtruckpete, here's a demo of what I have right now. Sound and lighting.

The project is far from being 100% done, other than the obvious missing handrails (which will be added back on as the final step) I have the relays to install and also have some cosmetic issues to do (mainly put the missing brakewheel back on). But so far the project is going by very well

Very impressive work! Thanks for the video!

@MichaelB posted:

Sorry for the delay in updates, a lot has happened with the install and I have been working very hard on it. Got done with the sound system yesterday and also did some tests with it for the first time. Although I did not have any of the lighting hooked up yet

Today I finished wiring for the lighting, after some head scratching on how to get the class lights working properly. I found only the red bicolor LEDs would turn on when the function was active, but found a workaround to get both of them working in parrallel. A diode in series for the red LEDs, so the voltage drop is the same as the green LEDs. Which has worked great.

Below are some update pictures showing the install in its current configuration. It's a bit of a rats nest of wires, but everything all fits inside of the shell so it's not too bad yet


The black post will be used to secure a plexiglass piece for the relays. I could not think of a better place to put the relays and that area above the buck converter is just empty space, so might as well use that for something.

The frame had a giant open hole where the speaker driver lives, that was closed up with a piece of plexiglass that I glued into place. I also bought thin foam tape to seal the fuel tank to the frame very well. The speaker I used for the install is part number SC700208-1, which fits perfectly in the original speaker location in the fuel tank (even with the correct hole spacing too).

The speaker I chose isn't super strong on bass, but has the same frequency reponse as Soundtraxx's 50 X 90 mm speaker. It's also an 8 ohm 3 watt speaker, so it's perfect for the job.

And as per request of bigtruckpete, here's a demo of what I have right now. Sound and lighting.

Update: I have decided to remove the switch for the rear class lights so the issue with the front green class lights is pretty much resolved

The project is far from being 100% done, other than the obvious missing handrails (which will be added back on as the final step) I have the relays to install and also have some cosmetic issues to do (mainly put the missing brakewheel back on). But so far the project is going by very well

Very nice……………actually fantastic

I have identified the problem with the green class lights getting brighter when the rear lights are switched off. I had both the green and red LEDs running off a single 1K ohm resistor, for the front and rear lights. I changed the setup so the green and red LEDs get their own 1K resistor.

Only issue now is the green class lights are much brighter than the red lights, but that will be fixed with a higher value resistor for the green LEDs. Looks like an extra 10K ohms in series for the green LEDs will dim them down to a much more reasonable level.

I am glad I caught this earlier as I think I would have ended up melting something, since some of the resistors were getting hotter than the rest.

Also I have ordered the relays for the couplers and smoke unit. I will use the original smoke unit switch on the frame to manually interupt the function ground to the relay. So I do not accidentally switch it on in the app with no smoke fluid or I can just completely turn it off if I don't want to run the smoke unit.

The run program switch will be used to select front and rear couplers by switching the function ground between 2 relays (one for the front coupler and one for the rear). As I think that switch is not designed to have like 2 Amps of power run through it, so the relays will handle all that power instead.

Nice work, so far.  If you're going to use relays to fire the couplers, feed track AC to the couplers instead of rectified DC, to preclude momentary DC supply sag.  I may design a relay-based coupler interface board before the year's out, as an alternative to my DC-based transistor board.

@KarlDL posted:

Nice work, so far.  If you're going to use relays to fire the couplers, feed track AC to the couplers instead of rectified DC, to preclude momentary DC supply sag.  I may design a relay-based coupler interface board before the year's out, as an alternative to my DC-based transistor board.

That's the plan, as I was told by Gunrunner that they're AC coils. Smoke unit is strictly AC too, so I will need a relay for it as well. I ended up ordering relays with 24 volt coil ratings, but Digikey stated they will operate at minimum 18 volts DC

@MichaelB posted:

That's the plan, as I was told by Gunrunner that they're AC coils. Smoke unit is strictly AC too, so I will need a relay for it as well. I ended up ordering relays with 24 volt coil ratings, but Digikey stated they will operate at minimum 18 volts DC

Blunami Fx outputs are 12v. As long as coil current is less than 100ma you are good.

I am using 5v coil current relays with 130 ohm resistors in series because I have dozens of them. They have a 30 ma coil current and two amp contact rating. They operate both couplers and smoke units no problem with track voltage. Smoke resistors are two 27 ohms in series. K.I.S.S.

Pete

Actuallly Bluami outputs are rated at track voltage minus 1.5 volts, I measured what the function common voltage is and it read 17.5 volts for me (I have my buck converter set to 18 volts DC). The 4408 also has its function outputs rated to 400mA each

@MichaelB posted:

Actuallly Bluami outputs are rated at track voltage minus 1.5 volts, I measured what the function common voltage is and it read 17.5 volts for me (I have my buck converter set to 18 volts DC). The 4408 also has its function outputs rated to 400mA each

Instructions indicate 12v bulbs should be used with the Fx outputs. 18v would instantly blow those bulbs out. My buck convertor is also set to 18v.

Pete

From what I can see, it seems the board has some kind of load sensing built into the functions. I have noticed in my earlier testing for my class lights that on initial power up the board illuminates the LEDs dimly for a split second, which suggests it's checking what's there with low current and adjusting itself to supply the correct power for the load.

In my earlier tests I had a red and green LED in parallel with one of the functions, I had a 1K ohm resistor in series too. The function would only allow the red LED to turn on when I activated it, suggesting it lowered the voltage to be around 2 volts. Which is enough for a red LED to light up,  but not enough for a green or white LED.

Upon adding a 1N4001 diode in series with the red LED, which has a 1.1 volt drop. It allowed the green and red LED to light up, since the voltage drop across both LEDs was now 3 volts. Which fully suggests my theory that the function grounds are not just direct contacts to ground, but can vary their power as needed.

This wouldn't be anything new, vehicles with canbus on exterior lights also have the capability to check the load. By pulsing a low current through the light circuits, it sees what the load is, figuring out what's burnt out and what's normal. Normally it's for vehicles with incandecent lights, which need a lot of current to make them light up, so you wouldn't normally see the load checking. But add low power LED lights to the circuit, and then you will see the load checking function in action.

But I digress... Not trying to derail my own topic, just mentioning it's not out of the ordinary since it's a system that already exists in other applications

@MichaelB posted:

That's the plan, as I was told by Gunrunner that they're AC coils. Smoke unit is strictly AC too, so I will need a relay for it as well. I ended up ordering relays with 24 volt coil ratings, but Digikey stated they will operate at minimum 18 volts DC

Couplers on MTH/Weaver/Samhongsa engines are agnostic - AC or DC will activate them.  I believe that the electrocouplers provided by Sunset/3rd Rail behave similarly.  My Blu Shark and Blu Geep work just fine with DC.  Relays typically have a minimum activation voltage (often 3/4 of rating) and an even lower release voltage in their specifications.  If I design an electrocoupler relay board, I'll probably include a low-power linear regulator to provide 12V to the relay coils.

I was able to fix up a stanchion that was broken off from the part that was bent up, I think it developed a stress crack when I was taking the shell on and off. Turns out you can generally solder any metal handrails that were broken off, so long as solder sticks to them.

I previously broke off the handrails on accident on my Lionel UP ES44, the double height ones on top of the AC cabinet. I was able to repair the damaged railings by extracting the broken off pieces from the shell and soldering them back on, they took the solder very well (my guess is they're made from stamped tin) and are very strong. I have attached some pictures below to show the fix on that, if you look closely you will see the solder on the repaired areas

I did this same fix for the stanchion on the Atlas Dash 8 and it too is a very strong fix, a bit of silver touch up paint makes the repair unnoticeable. Below is a picture of the repair I just did, the one on the right has the repair

Took the engine to my local club to get its first shakedown and so far it's running great. Lighting looks great too, very happy with how that turned out. The green LEDs have the 10K ohm resistor in series, which dims them down to a pretty much perfect brightness

I plan to get one of my PS3 engines and use that to speed match it so each speed step is equal to around 1 smph. Found out how to do it, so should be pretty easy to speed match it.

Relays should be coming in today, so I should be able to get started on installing them tomorrow

I have got the relays prepared to be soldered in tomorrow, but I have a question regarding the TVS protection diode. Do I need to add the TVS diode for the smoke unit?

I will be adding a generic diode in series to the power wire, so the smoke unit runs at half wave AC. But also to prevent the heating element from burning up from just straight AC into the smoke unit

Relays have been wired in to the Blunami and switches, they're working great so far.

The blue wire, for the couplers, goes to function 8 on the 4408. The green wire, for the smoke unit, goes from the function 7. I also added a switch in series for the smoke unit trigger, so I can completely disable its relay if I need to.

I dont have the electronics they're switching hooked up to them yet, I will do that some time this week. I did solder some clamping diodes in parallel to each relay coil too.

The relays will be zip ties onto a plexiglass platform above the buck converter. That plexiglass will be screwed onto black post I installed earlier. I am in the process of creating a template for the plexiglass piece

If you plan to do more of these consider getting some Vector board or similar prototyping circuit board. Less stress on the relay pins and makes for neater installation. Measure the pitch on the pins. .1” is common as well as metric spacing.

Using appropriate JST connectors for the wiring further adds to the flexibility.

Pete

@Norton posted:

If you plan to do more of these consider getting some Vector board or similar prototyping circuit board. Less stress on the relay pins and makes for neater installation. Measure the pitch on the pins. .1” is common as well as metric spacing.

Using appropriate JST connectors for the wiring further adds to the flexibility.

Pete

I ended up super gluing the parallel diode to the body of the relay to relieve stress on the pins. But also added a bit of super glue to any exposed leads the heat shrink didn't cover.

I don't mind having the diode and relay permanently glued together now. If the relay fails, I would likely solder on a brand new diode, as I don't know if the diode on the old relay will be any good anymore

Got the plexiglass piece for the relays made and also mounted them up to it, wiring is just temporary and will be cleaned up a bit better in the final version.

I also decided to make a plexiglass piece to put between the smoke unit and 4408, this should reduce how much heat gets to the 4408 by a lot.

It's just a piece of plexiglass cut to the shape and size I needed, with some aluminum foil glued to the smoke unit side (shiny side facing the smoke unit).

I ended up ordering the wrong TVS diodes. I needed bi-directional ones, but instead bought uni-directional ones. So I will place a parts order for the correct ones and hopefully they will arrive some time this week

Earlier I finished wiring up the relay outputs to their respective electronics to control. Smoke unit was really easy to wire up, couplers were more involved but not too difficult.

I was gonna buy bidirectional TVS diodes online but a friend told me I could just wire 2 up in series with the white stripe ends towards each other to get bidirectional ones.

I double and triple checked all my wiring to make sure I did it all correctly and it looks like I did. I also went and shortened down the wiring that was below the current keeper to more reasonable lengths

It's nearly 3 AM at the time of me writing this post, so testing the relays will have to come in the morning. When everyone else is awake in the house

I tested the smoke and coupler functions a few hours ago. Everything is working great and nothing put out the magic smoke, here are some videos showcasing everything working

Second video mainly is to show the smoke output from the stack. I put back the gasket material around the funnel to help guide the smoke out of the stack, making sure it has a bit of pressure on the funnel. In my earlier testing I saw some of the smoke was going in the shell from the small gaps next to the exit hole of the smoke unit, this doesn't appear to be happening with the gasket in place.

I would show final pictures of the electronics setup, but I got way ahead of myself and put all the handrails back on. So that might have to come another day. Although I think I got plenty of pictures showcasing how I laid everything out in the shell.

Last step I need to do is paint the replacement brake wheel I got, but I will need to make a trip to my local hobby shop to get silver spray paint. Could paint it up with my paint pens, but I don't think that will last or look that great.

I just finished up the cosmetic work on the engine and it looks just as good as the day I bought it. The white brake wheel is just a stand in, turns out I bought the wrong part from Atlas and I have now ordered the correct part. I'm really happy with the outcome of this project, never once did I think this engine would run again and that instead I would just hold onto it as a memento.

The hobby shop I took it to pretty much said it was junk and that I should just convert it to a dummy, I will not mention their name as I don't want to start any drama. But I have pretty much proven them wrong, even though it took me a couple of years to finally get it to this point.

Thank you everyone who followed along and give me help when I needed it, I think this project is officially done. I got to take it to my local club and give it a good shakedown for a couple of hours, I will post an update when that happens. A video will come later showing it in action, like I did with my Amfleet cab car.

Congratulations @MichaelB, you did a superb job! Very impressive! The engine runs and sounds great. I have been following it with interest as I would like to install one of these boards into one of my conventional engines. Thank you for detailing exactly how you did everything.

just curious, why did the hobby shop tell you to make it into a dummy? Did the original electronics totally stop working?

@Hudson J1e posted:

Congratulations @MichaelB, you did a superb job! Very impressive! The engine runs and sounds great. I have been following it with interest as I would like to install one of these boards into one of my conventional engines. Thank you for detailing exactly how you did everything.

just curious, why did the hobby shop tell you to make it into a dummy? Did the original electronics totally stop working?

Thanks, it was a bit daunting but overall a good learning experience

I believe the hobby shop doesn't understand the idea of sentimental value... Thinking it wasn't worth spending the time and money to get it fully working again. But I very much saw the value in bringing it back to full operation and I am very much glad I spend the time to do it.

At the end of the day it's not always about the cost of a project for me, but mainly what I will learn doing it. Here I learned a lot and will now be able to do more Blunami upgrades, when other engines with NLA electronics kick the bucket in my collection.

One thing I will recommend is test the stall current of the motors manually, as Ohms law doesn't always give a good idea of how much current motors will draw when completely stalled. I originally calculated a stall current on the motors of 6 Amps at 18 volts, as the resistance of the motors measured around 3 ohms on my multimeter. 6 Amps with those parallel wired motors really didn't seem right, especially since the datasheet on the motors said each one motor was rated for around 2.3 Amps stall current (they are Mabuchi motors, which I remember being very high quality and very well built from my work).

I found an old DC transformer in my closet and used that to test stall current by turning up its voltage to the max, which was around 15.5 volts. Then I manually grabbed the flywheels to smulate a complete stall, also having my multimeter wired in series between the motor leads and the DC transfomer. The motors drawed around 1.85 Amps when stalled, transformer was rated for 7 Amps total for the output and I didn't hear its overcurrent protection activate. Both motors had a lot of torque while I was holding them too and the torque felt equal in each hand. So I knew I would be more than safe to run the motors with a 4408 after my manual test.

I ended up taking the shell off one more time as I realized I didn't check if the diode I put in series got warm while the smoke unit was heating up for the first time. In my earlier testing I used a 1N4001, which is a 1 amp diode, in series to see if the smoke unit worked good with a series diode. I found it got warm during a cold start of the smoke unit and then cooled off when the smoke unit got to temperature. The smoke unit seemed to take a rather long time to get to a high enough temperature to start making smoke, so I figured that diode was not quite sufficient enough for that use.

I bought a 2 amp rectifying diode in the parts order for my relay circuit. This 2 amp diode  doesn't warm up on a cold start of the smoke unit and the smoke unit starts outputting smoke a lot faster now. Anyway, while I had the shell off I took pictures of the final electronics setup under the shell. I also took pictures of the class light setup.

I reused the old PCBs the original bicolor LEDs were soldered to, but this time made an effort to position the LEDs so they're as close to parallel with the lenses as possible. The front LEDs had a lot of light shining back into the shell, so I added some heat shrink around the lenses and that reduced it down a lot.

The bicolor LEDs used here are 3 pin common anode green and red LEDs. The black and red wires of the PCB are the LED grounds. Black represents the green LEDs and red represents the red LEDs. In total there's 4 wires going to each end of the engine for lighting. Common positive, headlight ground, green LEDs ground, and red LEDs ground. I wasn't able to reuse that black plastic cover for the rear class lights, as the setup was way too big to fit within the cover. The switch for the rear lights is there to completely disable them if I am running my Legacy Cabbage with it

I made the connectors so the pins are the identital between them. So if I accidentally connect the lights backwards, I will get the lighting swapped and not possibly mess up LEDs or the Blunami from having grounds and positive power swapped. The connectors are marked with which is front and rear lighting, so it would be hard for me to accidentally swap them around. The ditch lights each have a warm white 0402 LED put right up against the lense, so they're about as bright as the headlight.

This engine is pretty heavy for its size, thanks to that thick diecast frame...

But I figured a little extra weight in some empty spaces would not hurt, but also making the weight distribution exactly 50/50 would make it perform a lot better. So this is the final result of weight...

Added a 1 oz weight where the volume pot went, as well as 1 oz of weight in the air conditioning cabinet . In the rear I was able to fit two 1 oz weights, where the 9 volt battery holder used to be. All of which resulted in a nearly perfect 50/50 weight distrubtion, it was 0.3 of an oz lighter at the rear. But I wanted it perfect so I took 2 washers from my weights collection and glued them on the inside of the shell, above the trucks.

That got the weight distribution to 50/50, 3 lbs 1.5 oz on each truck. The scale was a bit off on the whole engine weight as it was leaning forward slightly on the scale, due to the underside of the fuel tank not being completely flat.

Truthfully, the best weight distribution is normally favor the rear truck with the weight.

Took the engine to have it's first official test and it performed perfectly

Smoke worked good and the engine operated great. I didn't build a lashup with the cab car, as I was just mainly testing the engine out and didn't want to mess with that yet. Ran it for around 6 hours in total and had 0 issues, smoke unit heat was good too

I’ll say it again. Absolutely fantastic work!

I came to realize today I can have a 2 stage smoke unit if I change my function wiring a bit. In the real world HEP mode would have the RPMs locked to a minimum setting of notch 6, which would result in more smoke coming out of the exhaust.

What I might do later is set up the front marker lights and rear green class lights to be connected to the rear headlight function, opening up 1 more function slot for me to use a relay to switch the chuff pin to track ground and therefore set the fan to 100%. That 100% fan function slot would be also mapped to the HEP mode button, so it will set the fan to 100% when the HEP mode is activated.