MTH PS-3 AEM-7 Internals?

No one huh?  Just for knowledge, the PS-2 3V AEM-7 uses a flash coded board for bidirectional ditch lights.  Then two separate aux light boards to control all the features.  These boards are different front to rear.  MTH part site is wrong on boards.  Also different than the PS-1 version.

For PS-3 the outputs are weird, so there must be an aux board used too, but MTH sample did not have one.  But the engine requires more output than the PS-3 diesel has so it must have one.

MTH did not create the flash code for Bi ditch lights for PS-32.  So a PS-32 repair will not function completely correct either.  G

under Perry's toy exchange on Ebay, he has shells listed with some having pics inside. I bet they're older models though?

Hi George:

I have worked on more PS3 AEM7s than I care to admit.  I have parts coming out of my ears and can get you pictures.  I'll reach out to you by email and we can go from there.

Hi guys!

Thanks for the information; we appreciate it.

@GGG, can you share photos and possibly videos that you've done for our project!  You are the best!

Thanks,

- Mario

@Pantenary posted:

Hi George:

I have worked on more PS3 AEM7s than I care to admit.  I have parts coming out of my ears and can get you pictures.  I'll reach out to you by email and we can go from there.

Truthfully, that information is of interest to more than George, any reason you don't post it here?

Hi John:

I emailing George by email about sending him parts he might need.  I read it on the run, and I see now all he needed was pictures, any of which I have / take will get posted here.  Sorry for my misread.

Thanks Nate.

@Pantenary posted:

Hi John:

I emailing George by email about sending him parts he might need.  I read it on the run, and I see now all he needed was pictures, any of which I have / take will get posted here.  Sorry for my misread.

Hi Nate!

What George and I would be interested in are the white rooftop strobes, as George was asking on my behalf as he has expertly upgraded my PS2 AEM7 to PS3, but I believe the rooftop strobes are carryovers from the PS2 unit, which are amber. Do you have the white strobes available from a PS3 AEM7?  We’d like them if they are.

thanks!

- Mario

Thanks Joe, That is the PS-2 3V shell with the 2 aux light boards.  Need to see the guts of the PS-3.  The 2014 did not have operating panto, in 2018 they added that too.  So I am sure the 2018 has the panto board for PS-3, but the unit has more directional light outputs than a PS-3 diesel can support.  So there must be a second board too.  G

@GGG posted:

Thanks Joe, That is the PS-2 3V shell with the 2 aux light boards.  Need to see the guts of the PS-3.  The 2014 did not have operating panto, in 2018 they added that too.  So I am sure the 2018 has the panto board for PS-3, but the unit has more directional light outputs than a PS-3 diesel can support.  So there must be a second board too.  G

Nope. Contrary to the MTH catalog, no Aem7 ever had operating pans. I have models from both PS3 runs (wrong Amtrak blue and corrected) and the internals are the same.

There is a daughter board behind the cab with plugs for all the LEDs,including the strobes. On my phone now but promise some pics tmrw. I think I have a spare couple boards. Will advise tmrw.

I deleted the post with pics of the PS2 models. The pics weren't mine. I'm still not sure about posting stuff like that.

@Engineer-Joe posted:

I deleted the post with pics of the PS2 models. The pics weren't mine. I'm still not sure about posting stuff like that.

No worries. Thanks for your help!  I think they were up long enough to get what we needed.

We’re very excited about this build as the PS2 sounds just aren’t up to par with the PS3 sounds, and the PS3 are non-existent anywhere. Anyway, we did away with a lot of the extraneous lighting. For example, the center warning beacon above the windshield; this was an emergency light that was engineer controlled and also came on when if brake line air pressure was lost. It was also a strobe, not a solid light…

In addition, I don’t switch with my AEM7, so a lot of the rear lighting was shorn. Rear cab lights and figures, and strobes, for example.

Thanks again to @GGG for this work!

I’m stoked.  Here’s a “before” look at it.

thanks!

- Mario

So can you post that picture of the daughter board in the PS-3 engine please.  Would like to see it. G

How’s this?

LOL!

thanks for everyone’s help in this, this is exciting!

- Mario

Greetings all:

My apologies for the delay.

NOTE, I have three models from the first PS3 run (Amtrak with erroneously dark blue paint) and two models from the 2nd PS3 run, (with the proper Amtrak blue paint).  The internals are identical.

1.  When you pop the shell off, this is what you see first.  I do not like how MTH packs wires in here, so I pull the cabin and all the hot glue out and arrange it myself.  It is a tight fit, and you MUST be careful NOT to pinch wires between the body screw holes and the frame during re-assembly. That shiny body coating is inside as well (hence the green insulation at the pantograph hole) and is METAL.  Pinch the wires, you short the DCS board out and melt the entire unit down.  I have done that once, and it really stinks, (figuratively and quite literally).  Also when re-assembling, you must have the wire arranged roughly how I have it around the green pantograph hole, or the flywheel on the motor will rub them to the strand and cause no end of electrical issues.

Finally, as is plainly apparent, there is utterly no room for a  pantograph machine.  MTH wrongly advertised this in the catalog.  Remove the two black screws immediately behind the cabin to loosen it up.  CAREFULLY pull it up and out.  You probably will have to manually widen the body a little with your fingers to get it loose.

2.  Once the cabin is out, you can see the strobes in the shell corners (covered with stick black felt from the factory), number board and center marker LEDs.  The little spikes off the roof manage the wires quite well, but you need tweezers to get them in there.  The wires out of the cabin are the ditch light, cabin and nose marker LEDs.  All the LEDs terminate to a small color coded plugs into the LED "daughterboard" where I have the small zip ties.

3.  Another angle of the LEDs and daughterboard in place.

4.  The 'daugherboard' itself.  It is only color-coded plugs, no electronics.  The two red plugs at the top of the shot are ditch lights.  The blue plug to the left under the ditch lights is the headlight.  The red plug to the right under the ditch lights are the number boards.  The center plug (that is yellow, I have a white plug LED in there) is the interior light.  The white set under that are the strobes.  The yellow set at the bottom are the markers.  The bigger black plug to the left goes to the DCS PS3 board.

Any questions, please let me know.

Wow.

thanks!

Alright, that is not a daughter or aux light board rather just a patch panel.  Which MTH told me the sample had.  So how does this operate lighting wise?  Do both cab lights come on in neutral?  Do all roof beacons work?  I have seen conflicting videos on You tube.

The PS-2 3V version has the forward cab interior light come on in forward.  Forward ditch lights, and forward roof beacons.  The rear red warning lights come on.  When you switch directions the forward cab light goes out and the rear cab light comes on.  The roof beacons switch to rear along with rear ditch lights.  The front red warning lights come on.  This is done with 2 Aux light boards with logic circuit and such to operate all the lights changes off a single output.

In PS-3 there is only one Interior light output.  Rule 17 applies in forward but when in reverse the cab light comes on and with brighter intensity.  So are both cab lights on all the time in PS-3?

The ditch lights can run normal as PS-3 has front and rear ditch outputs.

The roof beacon run off Mars 1 and 2 output.  Alternately flashing and differently whether in motion or neutral.  So each forward beacon is independently controlled.  So Mars 1/2 are used on the front beacon.  So what drives the rear beacons?  No output for it?

The 2 or 3 red warning lights can be operated off the Directional lights.  When in forward, the HL comes on it can drive the rear red warning lights too.  When in Rev the rev light output can drive the front red warning lights.

LNB have and output and drive the 4 number board lights.

The marker lights also are on, constantly, but the PS-2 3V does not have markers or classification lights.  Not sure about PS-3, but did not see them in the videos.

So the bottom line is you have insufficient output to operate the unit like a PS-2 unless those patch panels have some circuits.  MTH said they didnot see any on the patch panel.  G

So can some one tell me how the PS-3 light features operate?  Cab lights and roof beacons specifically?  G

Greetings George:

Sorry for the delay in response.  I am not aware exactly how the circuitry works as I have not tested it to the board level.  The function of the cab lights and beacons are directly controlled and operated by the DCS PS3 board, and hence the remote or app.  When the unit is started up after being added to the remote, the the beacons and cab lights are on by default.  Also, once it starts to move forward, the headlight brightens and the cab light dims, (it may also shut off; I forget as I write this). Other than that, all the light features and behavior are controlled by the PS3 board / remote.

That is what I am talking about, how does it operate via DCS.  But please some specifics, do all 4 beacons come on? Or just the forward pair when going forward?  Do both cab lights come on in neutral?

When you change direction to reverse what do the roof beacons do?  What does the cab light do?  Do they stay on in motion?  Last question, if you turn off the LMK (Markers) via soft key does any light output change or nothing happen, since the traditional marker/classification lights are not installed.

This may be one of those odd engines with more discrete functions in PS-2 3V than PS-3.  Beside directional ditch, the 3V has directional cablights, and roof beacons.  The only thing it misses is Rule 17.    Thanks, G

When activated, the strobes always follow the remote's set direction of travel, regardless of if the unit is actually moving or not.  That is, if the unit is set forward by the remote, only the forward strobes will turn on, and vice-versa.  Also, each strobe blinks at an irregular interval, modeled perfectly after the real AEM7s.  This is as opposed to many F40PHs that had synchronized strobes blinking at regular intervals.  One characteristic of the strobe circuitry is that the as-delivered strobe LEDs must be used in the strobe plugs as-delivered for the effect to work as I described.  If you use other MTH LEDs from this unit or elsewhere (I tried TRAXX LEDs as I liked the bright white over the incandescent orange-ish) in either strobe plug, or use the as-delivered LEDs in the opposite strobe plugs, the strobe effect either acts oddly, or does not work at all.  It makes no electrical sense to me, as all the LEDs operate on the same voltage, (as tested with a bench supply), but I observed this strobe behavior in every PS3 AEM7 I have.  Also, strobe brightness is constant.

The cab light also follow the remote's set direction of travel as the strobes do, (IE, if the unit is on forward, only the forward cab light turns on).  When the unit begins to move, the cab light dims.  In neutral, the cab brightens to it full intensity.

The headlight behaves opposite to the cab light.  It dims when the unit is stopped, and brightens when the unit moves.  You cannot turn on the rear headlight when the unit is set to forward.  This behavior is not user-adjustable that I have seen.

The markers behave as the strobes do, only in reverse, (only rear markers come on when the unit is set to forward), regardless of if the unit is moving or not, and the brightness is constant.

Ditch lights have 3 settings, again based on the remote's set direction of travel, (but they are always off in neutral).  1. They can both always be on when the unit moves, 2. They are off except when the horn is activated, in which they blink for a few seconds after the horn is deactivated, or 3. Off completely.  They follow the direction of travel as the other effects do.

Again, in all these cases, you cannot turn on LEDs opposite to the direction of travel as I outlined above, by the remote.

So I don't see how it possibly can work. There is only one Cab Light output.  So how does it control both Cab lights in a directional manner.  Also interesting is the Rule 17 has the light go out in motion, not dim.  Headlight goes dim in neutral and bright in motion.  On the test set, Cab light does follow rule 17 in forward. But in reverse it comes on bright and stays on.  So how can that voltage/current change turn one dim, and the other on?

There is no alternating beacon output except Mars1 and 2 output.  They come on in forward and stayed on in reverse. At brighter intensity.  So unless there is some sensitivity level difference with the LEDs in parallel that does this on and off?  Is that even possible.  Interesting you state normal LEDs do not work as beacons.

The directional lights, red marks, and bi directional ditch all can be made to work.  So I am still confused on Cab and roof.  Did you ever inspect the traces on the patch boards to see how the beacons are controlled?  All 4 on 2 outputs or?   I am still baffled.  G

I failed to note that there are two of the 'patch panels' that I pictured above, one in each cab.

Those two patch panels plug into bulkhead connectors that mount to the frame.  The wiring for those plugs then integrate into the main harness that plugs into the PS3 board.

The patch panel itself has no electronics or passive componentry that I can see, only traces between the LED plugs.

The strobe behavior is utterly bizarre to me, and one can be forgiven for suspecting that I am either less than observant, or simply have no idea what I doing.  Let me provide a bit of context to all these observations.

First, I have a decent electronics background, having training in EE and several years in radio engineering.  I say that only to demonstrate familiarity with handling sensitive electronics and methodical board-level work.  I do not claim any 'expertise' with MTH models or circuitry; I describe only what I have repeatedly observed.  (Even so, I do know the AEM7, E44, E33 and TRAXX F140 models rather well, having performed a number of repairs, rebuilds and modifications to work effectively with a catenary-powered system).

The first run of PS3 AEM7s used all 'cool white' LEDs for the light effects, (the markers are white LEDs behind red jewels in the nose frame, so color temperature is obscured).  This is a perfect look for the strobes, but not really prototypical for the other effects; I thought little of it.  However, the second run of PS3 AEM7s used all 'warm-white' LEDs for the lighting effects.  This was perfect for all the light effects but the strobes, which now looked almost amber-ish.

I got the ridiculous brainwave that if I swapped the warm-white strobe LEDs in a newer unit with some of the 'cool-white' LEDs in an older unit, I could really have a prototypical lit AEM7, (especially now that they fixed the Amtrak-blue in the second run).  I tested the MTH LEDs, and they all were ~3v, so they should be entirely interchangeable.  I did not use any non-MTH supplied LEDs, but in any case, LEDs are LEDs, right?

I pulled the warm strobes and put in cool LEDs, and they did not light.  I swapped them in the two strobe plugs, and one LED blinked erratically.  Various combinations of LEDs yielded similar results.  I then went back to the original strobe LEDs, but only one of them lit, again erratically.  I finally swapped strobe plugs, such that the original strobe LEDs were in the plugs as delivered from the factory.  The strobe effect worked as they should.

I observed this strobe behavior in both runs of PS3 AEM7s, having attempted numerous 'LED swaps'.  I have no explanation or even an idea how/why this could be.  I only know that I tested this repeatedly in the same an different AEM7 models, and replicated the behavior numerous times.  I eventually abandoned all the LED swapping and returned all LEDs to their original plugs.  This took time, as I had not noted which strobe LED went in which strobe plug, (I never expected it would be important).  I clearly did (do) not understand how the circuitry was functioning, which made me uneasy about experimenting with it further, lest I wreck the PS3 board somehow.  I wouldn't know how, but since LEDs didn't seem to be LEDs anymore, I had no idea what else might be bizarre.

I can send you a video of lighting effects on my units.  You can also look at an AEM7 review that Eric's Trains did, which shows the effects as well.  (Ironically, his strobes are backwards as delivered).

In passing, I am also completely willing to ship you one of my PS3 AEM7s so you can add it to the DCS remote, play with it how you see fit, take it apart, or whatever might help.  All they do is collect dust these days anyway.  I'll pay postage out, you pay postage back.  I just need to come by a box.  Let me know, please use my profile email.

Thank you all for this attention to detail. It is so very interesting.

I do want to mention that I was able to get my AEM7 on the track today at the club, and a HUGE THANK YOU to @GGG for this thread and his PS3 install in my loco, which was the inspiration for this conversation.  

Please keep the conversation going.
thanks!

-Mario

PS- I also found it strange that Eric’s AEM7 strobes were backwards.

Nate I appreciate it.  Looking at the patch I can see pin 1 and 10 as common.  SO that should be DC ground for the LEDs.  Pin 2 thru 9 are the independent controls.  I imagine some doubled wires inputting from the wires going to the 40 pin.  All I can think off is they use some LED with higher voltage requirements and possible driving some with a voltage difference from two positive voltages possibly.  Appreciate the offer, but this is curiosity more than necessity.

I know what the test set LEDs do, and it is interesting you say the cab light just dim not go out in motion.

If you traced the outputs of the 40 pin using the diesel upgrade wiring instructions you might figure it out.  Like I said Ditch and LNB and the directional HL are straight forward.  I wonder if they use the Marker light output (LMK)  along with the Cab Light and MArs outputs to control the sequencing of the LED via a voltage drop.  Just speculation, but may be that they have reversed the led 2 pin molex or how they wired the patch panel to do this.  Maybe why some LED you tried did not work.  The Plug polarity was opposite.

This is true for PS-2 versus PS-3.  The marker LED are wired opposite and so are any other LED used in specialty PS-2.  PS-2 uses Constant + DC and pulse to ground.  PS-3 uses common DC ground and pulse +V for control.  G

Greetings George:

Well, the proverbial bulb went off in my head as I read your last post.  The reverse polarity LEDs are EXACTLY what the issue was, and even as I read your post I remember exactly how I found it, tested it and ultimately got around it.  My apologies; it has been a few years since I did all that work on PS3 AEM7s and my memory rather failed me earlier in this thread.  I also remember thinking it was a rather ridiculous an arrangement that was; you had to look close to see that the strobe wires were reversed in the plugs.  I suppose it makes sense and is an effective design, but I wasn't fond of it then, probably because I failed to notice it until I had wasted a bunch of time tracing / testing things through.

Also, you are correct about the cab lights; they do not dim but go out completely.  I looked at the AEM7, E44 and E33 this evening, and that behavior is consistent.  What I saw in all cases what the number board LED light bleeding into the cab, which made it look dim, (actually a pretty neat effect in the end).  So, put your mind at ease, I think we (you) figured out all the oddities with this unit. 

Did you have a photo of the front side of the patch panel with LED plugs installed?  Interesting methodology they used on this unit.  G