Replaced Pass Car Lights . . . now flickering?

Yep, the flicker is because you have no bulk storage, aka capacitors.  You need a capacitor in the DC power supply to ride out small interruptions in track power.  The LED's respond much more quickly than incandescent lights, so it's more apparent.

Thanks, John.

Seems I have stumbled into a known pitfall.

Am I wrong to feel I have been sandbagged? Seems to me the LED strips GGD are selling should have had this capacitor in the first place?

Don't think I would have gone this route at all had I known this huge downside to buying and installing these LED strips.

Any easy way out of this?

I'm somewhat stunned that the GGD LED kit doesn't have the capacitor, do you have pictures of what they look like?

If they don't have a capacitor, one could probably be added, but I'd have to see a strip to provide the exact details.

If the illustration below is your kit, I don't see any capacitor.  Truthfully, for $15/ea, I'd expect more than a 12" strip of four LED's and a bridge rectifier!

Yep. That's it. Of the eight I bought, one was DOA. That happens I know. Not upset 'bout that.

But not very happy about the flickering. I'll be going back to the incandescents!

Surely people are complaining about this?

Terry, I'd probably opt for adding the capacitors, it shouldn't be that hard, and they're cheap.  The LED lighting will really drop the power consumption of the cars, it's practically mandatory for running more than one large passenger train.

From the looks of the board, you should be able to add the capacitor right after the bridge rectifier.  If you take a closeup of the top and bottom of the board near the round rectifier, I can probably illustrate it for you.

Thanks John:

Image attached. What capacitor at Digikey?

100uf,220uf,330uf,470uf,or 1000uf,35 volts or more. Install in proper polarity across the + and - of the bridge rectifier. Backwards capacitors will explode. An 18 ohm resistor in series to one of the capacitor leads will reduce wheel arcing,sometimes experienced with high value capacitors. The bridge is the black round thing.

Dale H

Here's a good choice at Digikey: http://www.digikey.com/product...61S/P11239-ND/356142

See the illustration below.  The basic procedure is to drill two holes for the capacitor leads at the same spacing as the leads so the cap can be inserted through the board from the top. 

Scrape a small amount of the insulated coating off each trace next to the leads.  I use an X-Acto knife perpendicular to the board to just remove the solder mask.  Do NOT cut through the copper of the trace, you just want to expose it.

Insert the capacitor through the holes, fold the leads over to lay over the bare spots on the traces, and solder them on.

Note that the wide trace on that board is the positive side of the cap, the negative side of the cap is the one with the big stripe, that goes to the narrow trace.

That's the whole job.

If you have DCS, you might want to add a 22uh choke as well, I found a fairly inexpensive one at Jameco: http://www.jameco.com/1/1/4953...0-inductive-tol.html

For the choke, you'd drill a couple of holes next to either of the leads from the track, then cut the one trace and solder the choke to either side of the break, just like you did the capacitor, only you're using just one lead.

Isn't anyone concerned with the increase in voltage?

For future reference, you should check out the last link in GRJ's signature line.

Passenger Car LED Lighting Kit @ Henning's Trains

Many thanks for the instructions and the Digikey link, John.

Before ordering, is there a suitable choke at Digikey? Getting Jameco to ship to Canada might not be so easy. If DK have something suitable, that would be ideal and any price difference immaterial. What will the choke do, BTW? I do have DCS.

Great question. I assumed they were sized for peak voltage. You know what happens when you assume! Clicking on the pic and enlarging, Looks like 160 ohm? 6 band resistor, brown blue brown black purple yellow. 4 LEDs in series? (cannot see the whole board) would drop about 13 volts. 18 volts peaks at 24.  24-11 /.02 should be around 650 ohms. without a cap about 250 ohms.  160 ohms seems inadequate in any case. 

EDIT Looking more carefully each LED has a resistor. Should be at least 1K then. Must be reading the resistor wrong from the pic. 

Dale H

The capacitors cost so little that "I've been sandbagged and I'm going back to incandescents" seems to be a bit of an over-reaction.  This is a creative hobby, not a science. The toys are meant to be fiddled with, kit-bashed, modified and, most of all, used as a learning tool for many different skills.  The necessary caps cost about a quarter.  Seems like a small price to pay for the benefit of steady lights and some electronic education.

I am eager and grateful for the education, but my micro soldering skills are rather poor!

And I'm not so eager to let manufacturers off the hook for products lacking in basic design with the philosophy "let the buyer fix it!"

If those caps are so cheap and easy, and they are, even less excuse for not incorporating them into the design in the first place.

Without actually seeing the board, It's hard to know they're really wired and to properly size the resistors if they need to be changed. 

I asked a question about this circuit to Scott Mann in a post when the lighting strips were introduced way back on the Forum here. . He said they adhered to the KISS principle. It is not that the lighting strips do not work,it is that they flicker without a capacitor unless the rollers and contacts are very,very good. Sometimes KISS is too simple and this is the result. 

To be fair, some of this was not his fault. A capacitor in this circuit if used and other circuits already then existing on the market using them, (even including some MTH lighting units) having capacitors messed up the MTH DCS signal.  At that time the choke fix was not well known so they opted to leave it out. Even now the 22uh choke is just a guess and perhaps not optimal. There is no engineering behind the value used. This really was and is a problem of DCS design,which was introduced prematurely and made the users Beta testers. It did not work and play well with existing  components in the field. 

3RD rail makes model trains and does a good job,in fact the best job IMO.  They are not electrical design manufacturers and are not responsible for deficiencies in a competitors system. 

Dale H

I"m curious how you came up with the "fact" that the 22uh choke doesn't solve the problem of interference with the DCS signal.  Since it's widely used, including by MTH, I'd think they would have tumbled to the fact that it didn't work if you were, in fact, correct.  What makes you think the value is "just a guess", do you have inside information that MTH and ASC techs don't have?  Have you computed the optimal value?

Looks like you might have an earlier version of the board if the new ones have the capacitor and inductor (choke)  This search result suggests MTH was including the inductor in their upgrade kits 10 years ago. The inductor was for the lighting circuit which has a capacitor.

Suitable inductor at Digikey for about 25 cents:

http://www.digikey.com/product...K-TR-RCCT-ND/3741659

If you don't want to mess with drilling/cutting/scraping the circuit board you can splice this inductor into either of the 2 wires going into your lighting board.  If doing so, protect the exposed connections with electrical tape, heat shrink tubing, whatever.

I'm not sure your question on exactly what the inductor does was answered.  An inductor has the electrical property of blocking high-frequency signals. So by placing an inductor before the capacitor, the high-frequency DCS signal does not reach the capacitor.  Since a capacitor absorbs or smooths out high-frequency signals, this would be rude behavior toward other devices on the track looking for that DCS signal.

Never said it was a fact or that it did not work. I said the value was not scientifically calculated,rather it was just arbitrarily used because someone tried it and it worked. The solution did not come from MTH but from a Forum member. Because it works  does not make it optimal nor does make it good engineering. No,I did not calculate the value but no one else did either. If someone did then it has not been presented,at least here that I have seen. I assume if I look at an electrical board of any sort,most all the values of components are calculated for optimal response. Guess work is not engineering.

Dale H

Well, I'm sure learning! But, rather than get involved in some 3-way battle about who did/didn't do what and why, all I want is flicker-free lighting in my "Pride-O-The-Line" 21" GGD cars! I just recently received these LED strips from GGD, so am perplexed at the idea that this problem has been fixed in newer strips???

And thanks for explaining about what the inductor does for the DCS. And for the link to suitable ones at Digikey.

I'll order the caps and chokes and see how I do at soldering them. Might try the Hennings strips too. My soldering really sucks! But, I've watched a few Youtube videos and can see some mistakes I've been making. So, I hope to do better.

Soldering is not that difficult once you get a little practice, you'll be surprised how fast you can pick it up.

Speaking of facts, he's a couple of facts for you.

FACT: MTH supplies the 22uh inductor with every PS/2 upgrade to isolate the CV power supply and prevent interference with the DCS signal.

FACT: Hundreds, if not thousands of people have installed the 22uh inductors to prevent interference with the DCS signal into Legacy/TMCC locomotives and many other track powered accessories that have capacitors and may or do interfere with the DCS signal.

CONJECTURE: MTH just watched this, or some other forum, and picked the 22uH value for their use out of thin air.  They didn't bother to actually think about it, or do any analysis to see if it actually was effective or the right value.  Did you actually have a contact at MTH that verified that they didn't actually research this issue and compute the correct value of the inductor? 

Can you point to the conversation or person that "discovered" the value of the filter and posted it for the very first time so that MTH could find it?  I think you're confusing this with the "magic light bulb" that Marty was instrumental in pioneering.

Truthfully, I'm not sure why you make such a big deal of this Dale, or for that matter, even brought it in the first place.  The fact is, the 22uH choke has been demonstrated countless of times to minimize interference with the DCS signal, and is a pretty universally recognized solution. 

Can we move on now?

GRJ stated it earlier but if you're new to making these mods it is critical to mind the polarity or + and - connections on that capacitor.  I recall several OGR threads where the capacitor was installed backwards and exploded when powered.

I don't know how many additional cars you need to convert, but definitely try out the Hennings strips.  In addition to having the inductor (for DCS compatibility) and the capacitor (for flicker-reduction), you can also adjust brightness.  Additionally, without getting into arcane circuit details, the Hennings approach is, in my estimation, about 5 times more efficient.  Yes, the big gain came from your decision to dump incandescents and go with LEDs.  So that's just an observation.

Also, you'll note that the Hennings method using the LED lighting strips places 3 LEDs per 2" section...so LEDs are spaced every 2/3".  As I understand it, you board has 4 LEDs spread over 12" so that every 3".  To each his own, but in my experience you'll get smoother, more-uniform, lighting with more LEDs spaced closer together than fewer albeit brighter LEDs spaced further apart.

Also, the Hennings circuit has an IC regulator chip that steadies the current into the LEDs which effects constant-brightness.  This may be semantics but simply adding a capacitor to an LED circuit indeed reduces flicker but does not provide constant-brightness.  So when you hit a piece or dirty track or the rollers skip over a turn-out/switch the capacitor is called into action.  Without the regulator chip the brightness of the LEDs instantly ramps down as the capacitor releases its energy.  You may or may not be able to see this "flicker" so it's a your-mileage-may-vary scenario.  Use a larger capacitor if you need more flicker reduction.

John

I probably have been on this Forum longer than you,not sure. I used to make automation  circuits for people with DCS. Way back I did ask the people over at MTH about the interference problem when DCS was introduced. After many no answers and crickets chirping, they denied problems existed. I can not remember back that far but the choke solution was posted by a Forum member, one of the Beta testers. Some of these people sadly are no longer with us.  I can not remember for sure,not that it matters.

If I repair a circuit and there is a 47 K resistor and I replace it with a 56 K one,most likely it will still work but it may not be optimal. It may not matter in some instances.  Usually there is a reason for a value and engineer puts in most anything. Tolerance varies. This is the difference between engineering and guess work. Speaker baffle sizes for example need to be calculated with formulas,not guess work. The sound will come out either way but it will not be optimal if not calculated. To some,engineering makes a difference. I used to work for engineers and they usually were pretty precise about things.

MTH may give a 22 uh choke with a kit but they do not supply TVS protection nor recommend it for conventional operators who take their starter sets and engines out of the box. Either they don't care or they are intellectually lazy. I do not look to them as authorities on the subject of electronics. If you want to go into ancient history they were very late to the party with the PS1 issue. It was only pressure from a coalition of customers which made them acknowledge it and do something.  And it was a 3rd party who gave us the BCR and a solution. From their history I dont think they know their own product and simply do not trust them.

I never said the 22uh choke didnt work.

I never said the 22uh choke didnt work.

i never said the 22uh choke didnt work.

Please take the time to read my post.

I said if someone calculated the value it has not been shown. I have a drawer full of 22 uh chokes for people who use DCS. I made a bunch of LED circuits for cars and cabooses years ago for people who did. These are still functioning years later and on some display layouts which run constantly.   I also sent 4 amps to people to put in Lionel engines who ran both systems.  The choke was recommend by a Forum member long before MTH acknowledged the problem existed. Simultaneously, a Bell labs engineer I consulted also recommended a choke in the lighting system due to the nature of the DCS signal. He could only guess at a value also but said it was not critical. There was nothing published but he suspected the nature of the signal by what the system did. The system itself he thought was nothing special.   I dont use DCS myself. I would not know how to calculate a value because the system is proprietary. I do or did know some people in some places but I wont mention names.

As far as the magic bulb only one person I know actually calculated the value,that was a female electrician who had a son that was an EE if my memory serves me from reading the post. Another 3rd party solution.

I think we should drop this so as not to bore readers to tears.

Dale H

This is interesting. I'm not bored .

so is this thread...

https://ogrforum.com/t...-dcs-signal-strength

Tony, if you want to send me one of these strips, I will mod it with the cap and choke and send it back.  That way you'll have a pattern of exactly what you need to use them.

Thanks for the link. Very enlightening explanation by F Maguire.

His final quote sums things up.

"The significant thing is that we are in a position to analyze the DCS transmission path in much the same manner as the branching layout of the master TV antenna system.  The Barry Rules express much of this to a certain extent.  Still, it can help at times to borrow a more structured engineering approach.

If for no other purpose to generate more "Rules" (or less "magic"). 

For this we would need formulas for transmission equations and impedance matching.  The are a bit hard to find, so I'll post them as I have time for it."

Dale H

Thanks for proving my point Dale.  That was in reference to the "magic lightbulbs" and "engineered filters" for DCS signal strength, something I previously pointed out.  The filters to simply block the DCS signal weren't mentioned.  The whole discussion revolved around treating the track as a transmission line, which is key to signal propagation of DCS signals along the track.  They're not trying to discuss blocking the signal, quite the opposite in fact.

I suggest you go back and read the thread again.

Tony? Is this offer directed at me? I'm Terry.

I take it that it is. If so, thanks for the offer but I've ordered the caps. and the chokes from Digikey and will follow your instructions. I have 7 strips to mod but will need another 6 to do my 18" cars too. I'll try the Hennings strips rather than buy additional GGD strips only to have to mod them.

I run both DCS and Legacy, depending on what I want to run on a given day. So, I need the chokes.

As for my soldering skills, they can only improve and I need to be able to do this stuff. Seems I need a better iron for starters.

I'm good at sweat soldering water lines BTW!!!   Haven't broken out the propane for the trains yet though!

John

I read the thread and don't have them confused. I did not bring up the subject of light bulbs, you did. F Macguire  explains things pretty well as far as the signal. Mr Macguire did not get his information from MTH,he figured it out himself,her had to. This is the problem with a super secret proprietary system not properly engineered or tested in the field. The customer is left with the burden of troubleshooting. It would have been nice if the system was engineered more thoroughly in the first place and not left to third parties and the users. Anyway you have the last word on the subject.

Dale H

I'm with Terry on this one. The caps should have already been with the lights he bought.

That was some thread! Interesting read, but most of it went right on over at very high altitude.   Even so, if a piece or two sticks, I'll know more next time.

AHH!! Mr. Bloom do you perform any modifications on your trains??doubtful!!

Thanks for your great words of encouragement to sincere frustration!!!!!!!!!!!!!!

Boo!!!

Dale, John, all

What is the technical reason for installing the 22uh choke and under what conditions is it needed.  I am in the process of addressing DCS signal strength issues at my club (which also has TMCC) and want to better understand 22uh functionality and best places to locate the choke, if needed. 

Thanks,

rad400

The choke is recommended for any car or engine on the loop which has a circuit with capacitance. Mostly this is a constant intensity lighting circuit or an aftermarket circuit where a capacitor is used to reduce light flickering. TMCC, locos with Railsounds units and PS1  locos may have such circuits also and a 4 amp choke can be added to them. Lighted cars only need small amperage chokes. John would know more about this because he uses the system. perhaps he will chime in.

Dale H

I think Dale covered it.  Some, but not all, TMCC and Legacy locomotives will attenuate the DCS signal.  At our club, if we have several Legacy locomotives on the layout, DCS operation is spotty at best.  We seem to have less issues with TMCC locomotives, though a couple isolated instances have popped up.  There are also some Legacy locomotives that don't seem to affect the signal.

For other stuff, if you have a rectifier followed by a filter cap, that can affect the DCS signal.  For that reason, I include a 22uh choke in my lighting circuit, as well as my Super-Chuffer.  My upcoming remote controlled audio & automation board will also have a 22uh choke on the input. We have track powered signals with a diode, cap, and relay, they also get the choke to prevent issues.

It's a cheap component, so if you run DCS, I'd recommend any time you're in doubt if you need it, stick one in.  The choke is right at the power feed into your circuit, it should be the first component.

MTH includes a 22uh choke for installation with their PS/2 update kits for locomotives with a constant voltage lighting board for the same reasons.

The choke is also useful across the power inputs of digital switch motors. They typically have a small cap in the electronics.

This is a common application of users running FasTrack with a DCS system. 22uh works just fine.

Also, signal filter