Command Upgrade to willams Chuffington.

Hi Shawn, The DC Commander will directly drive LED's. You should not need the load resistor.

I don't think you will need the caps on the motors. I am sure the other guys on the forum will confirm. I have not used them in the Williams GG1 I converted with the Cruise Commander and that operates perfectly.

Nick

Hi Shawn,

Did you check for clearance to determine if you could stand the Commander vertically with the heat sink on the chassis? That would provide room for the FatBoy or a Baby FatBoy.

Agree with Nick, drive the LED lamps directly. use a motor cap only if you get audio noise.

gunrunnerjohn would know the coupler size for you.

The shortest coil coupler that I see is this one.

6407950550     COIL COUPLER / MNT C / LENGTH: 1.40"

I have to issue a correction however.  You DO NEED THE SERIES RESISTOR, that's for current limiting.  You don't need a load resistor (or capacitor), those are required for the plain R2LC equipped stuff, but the cap is on-board the DC Commander.  Make sure the LED positive lead goes to ground, the lighting outputs in command mode are negative in respect to frame ground.

gunrunnerjohn posted:

The shortest coil coupler that I see is this one.

6407950550     COIL COUPLER / MNT C / LENGTH: 1.40"

I have to issue a correction however.  You DO NEED THE SERIES RESISTOR, that's for current limiting.  You don't need a load resistor (or capacitor), those are required for the plain R2LC equipped stuff, but the cap is on-board the DC Commander.  Make sure the LED positive lead goes to ground, the lighting outputs in command mode are negative in respect to frame ground.

 

Thanks John. Does that apply to LED headlight(s) using your Super Chuffer board? 

Nick

Exactly, the way it is setup. When in command mode the front light flickers until the throttle is past a certain point.

i used a 470 ohm resistor. The led is a little dim. But, in convention the DC commander will output more voltage.

To make the motors go, lionel and mth seem to rely on external sensors to determine rpm on the motors. Does the commander board need that kind of input or does it just go with estimating based on current of the motors while "coasting" , etc... ?

Nick12DMC posted:

Thanks John. Does that apply to LED headlight(s) using your Super Chuffer board? 

No, the S-C is designed to drive an LED headlight directly, that is spelled out in the installation instructions.

shawn posted:

Exactly, the way it is setup. When in command mode the front light flickers until the throttle is past a certain point.

i used a 470 ohm resistor. The led is a little dim. But, in convention the DC commander will output more voltage.

The DC Commander puts out half-wave track voltage in command and full-wave track voltage in conventional, that's a function of the R2(4)LC lighting outputs, it's the same for factory TMCC stuff.  The LED should be at pretty full brightness with a 470 ohm resistor and the diode.

Severn posted:

To make the motors go, lionel and mth seem to rely on external sensors to determine rpm on the motors. Does the commander board need that kind of input or does it just go with estimating based on current of the motors while "coasting" , etc... ?

No, the Cruise Commander (but not the DC Commander) uses back-EMF to sense motor speed for the cruise function.  At low speeds, they also count commutator pulses for better low speed performance.

The DC Commander or the Mini-Commander II just puts out a fixed voltage based on throttle setting, there is no sensing of the motor speed as neither of them offers cruise control.

Thanks John. S-C has arrived in the UK safe and sound. I haven't had chance to read the instructions yet... 

Waiting on the ERR kit , tether from MTH and order from Lionel for the smoke unit to arrive. 

Nick

I have a vague-ish feel for back emf -- in that a coasting motor is a generator.  But how do they count commutator pulses?  Oh hey look there is plenty on google on it, even patents.  Who knew.

I'm sure there's tons of stuff on back-EMF speed control.  Could be that Jon Z. at Lionel may be named on some of those patents, he's the guy that did the ERR Cruise Commander, and that uses both the back-EMF and counting the commutator pulses.

gunrunnerjohn posted:

The shortest coil coupler that I see is this one.

6407950550     COIL COUPLER / MNT C / LENGTH: 1.40"

There's a reason for the shortness of that particular coupler... it's for American Flyer S Gauge.

TRW

OOPS, I guess the 1.44" one is the shortest then.  I've used this one.

6208447550     COIL COUPLER / MNT C / LENGTH: 1.44"         $10.00

I wondered how I had missed the 1.40" one previously.

I did my own back emf using a raspberry pi,  motor power board and a circuit of my own design.  But it doesn't work very well.   Besides the fact that my circuit may be dodgy -- I think the PI isn't reactive enough through linux.  Yet it's a nice easy package to deal with.

Yep, I think back-EMF would require pretty fast real-time response.  My stuff is done with a simple processing loop with interrupts providing key timing as required.

what i did was read the voltage off the motor(s) when no current was going to them for a brief moment, pumped that through an a to d and read the values on the PI.  On a scope I at least think I see a "burble" when the motors are free spinning for a like say 20 ms.   But it also seemed rather course and unpredictable in the PIs linux and so I thought about moving that stuff to an arduino.   It doesn't help that i have no idea what I'm doing.

I could of course just buy some decent full on motor controllers with emf or feedback control... but where's the fun in that?

I believe the commutator breaks at low speed probably confuses the back-EMF reading.  I suspect that's why Jon Z. used the commutator pulses at low speed for better low speed control.  At higher speeds, you probably don't see the commutator pulses nearly as much.

Of course, Jon could really answer the questions since he actually designed it.

what I see are motors that -- and these are out out of an old williams -- the common "mabuchi" or similar -- that they worked fine as conventional that I could tell.  But on the bench with a rather poor mans set up -- they perform poorly at low speeds.    What I see is a big gap in the hz/mhz of the PWM pulses to get the things going.  I lose about half what the chip set will put in this regard before they go ... then with just slight speed ups they go way too fast or they may be going to fast. 

Basically the motor board may be under powered... I picked what I thought would work and I think they put out 3 amps continuous per channel.  But then I read even small motors have a big kick to get them going.   I could buy some other boards that have much more power per channel.   At the time what I bought seemed the best offering for an experiment.

Anyway the other thing I notice is the motors are not matched and run differently as different pulse frequencies -- or at least this obvious at slow speeds.

Although I guess the arduino or similar fully detected but run by software subsystem approach could work -- i imagine a nearly full blown hw solution might be ideal...  

At least in DCC land the motor controller is integrated into the sound boards... and the best of my knowledge really on PWM or possible other tricks (like the commutator pulse) to get RPM for the purposes of reasonably control.

They seem like pure HW to me when I look at them, which is a little unfair since I've but one of them -- so i'm over generalizing possibly.

There's of course 57 ways to skin a cat and using direct tach readings is an option ... but I wanted to make due with what I had on hand, not go out and buy a bunch of matched motors or all-in-one controller boards, etc...

sorry my response is somewhat garbled in places.  Basically the dcc folks seem to go to an all hw solution, and wonder if the embedded microcontroller is going to be enough as well.

The Led seems to be yellowish and sort of dim with 470ohm and diode in command mode.

Also, the led flickers up to a certain throttle point. Although, this could be signal? As, Tested this not open air!

Ok, I  know this isn't a great picture. But, In front of the left motor there is a plastic molded piece mounted to the frame.

Inside is a circuit board with a light pipe leading to the front headlight lense. I'm thinking of removing the LED enclosure

light, pipe and circuit board and mounting the fat boy speaker in that area. My only question is? When using LED's to replace a headlamp are you just glueing the led behind the head light lense.

I have tons of lionel led's in my service box. THe current setup is a waste of space.

I like the idea of removing that "Light Box".  Placing the LED behind the lens would require that the power wires have a quick disconnect for shell removal.

I would assume that the board is an AC-DC converter and voltage regulator to drive the light and would be powered by track power. It may come in handy for another project.

GRJ uses board header connector pins and snaps off pairs. I have used the method for passenger car lighting changes.

I found a review of the engine and grabbed a shot of it with the shell in stock. It provides a good view of the light box.

FWIW, in many diesel installations, I remove the plastic lens in the shell and have the LED project directly from the headlight opening.  This gives me the best looking headlight effect.  As far as the LED being "yellow", that's strictly the color temperature of the LED, it doesn't really change if the LED has less current.  To remedy that, pick a different color temperature LED.

As far as the flickering, if you're getting steady power like on a test bench with it wired directly, that's not the LED or the resistor, that's something to do with the TMCC signal.

Yeah, sort of figured the LED color thing. If you saw how I tested the engine...lol, that's why I figured it was most likely the signal. Nothing, wrong with asking someone with more experience...right

Hmm, Doens't the led look a little hokey - installed that way. Or are you suing a flat front led?

You can use a flat front LED, but I use the standard focused LED's.  I have never had a complaint after several hundred installations of how the headlights look.  I prefer the performance of the focused LED as it really throws a true headlight looking beam down the tracks.  For steamers, I typically use a 2700K amber LED, it gives you that "incandescent bulb" look, just a whole lot brighter.  Steamers typically have a headlight lens that stays.  For pilot mounted steamer headlights, I use a surface mount 3528 LED in the back of the enclosure as normal headlights don't fit in those.

For vintage diesels, I typically choose the 3000K-3100K warm white bulbs, and for modern stuff I may use the bright white LED's in the 4000K range.  It all depends on the effect you're trying to achieve.

Thanks, John - see you at york?

I don't know if I'll see you, but I'll be there.