vignale

Hi everybody, I have done some experiments with tantalium capacitors and would like to report them. I am writing in this old post because I think this is still on topic. I have finally obtained some 16V 470uF tantalium capacitor and I have deployed them in different locos. The first one is a small switcher with a MX600. It had very little space available for capacitors, so I put only 5 capacitors in parallel (2350uF). The loco had many pickup problems but now runs noticeably better and can keep lower speeds without stalling and losing contact. This is how I wired the capacitors. Regarding the decoder wiring, I followed the guide by Digitrains that has been posted before: white wire goes to the back ground pad and blue wire to the common positive. Works great. I also tried to make a bigger capacitor bank for another loco running a MX633, but I ran into some problems. I wired 20 capacitors (9400uF) in parallel and connected them as per the Zimo manual: blue wire to the positive capacitor side, and gray to the negative side. However, when I enabled track power I got a good puff of magic smoke coming from the capacitors. After disconnecting them the decoder and loco still runs fine, but I am not sure what to do with the capacitors. I am pretty sure I did not make any wiring mistake: the polarities of the capacitors are all the same and should be wired correctly. Only one capacitor got toasted, but I am not sure if the others are still good and usable. Anyway, why did this happen? Track voltage is 12V and the capacitors are rated at 16V. Does the MX633 have a step-up voltage converter charging the capacitors to higher voltage than track voltage? The Zimo manual mentions 16V capacitors, but the drawings show 20V capacitors. Is the 16V rating too low? This is the aftermath. Thanks to everybody who can help me diagnose the issue.

Thanks for the info everyone. I have ordered some 16V 470uF tantalums from China which will hopefully arrive in less than a month (I could not find them in Italy at a reasonable price). I will update the topic when I can test them.

The capacitor supplied by Zimo was an even smaller one (about 600uF). Even with the bigger one I used I am seeing absolutely no difference. I know there are multiple other options, but I first wanted to understand if it was just a capacity problem or I was making an error in the decoder configuration. I have very limited space in some older locos and honestly don't need 6s running time. I remember some discussion on the forum about ceramic/tantalum capacitors being "more effective" than an electrolyte one with similar capacity. Not sure if it's true, but I was looking at some 330uF/16V tantalums which i can put in parallel to fit the space available. I have other locos running MX600 which have more pickup issues, but the new loco was the only one with a 633 which had the capacitor leads already available. I figured it would be mich easier to learn and test the stay alive functionality with this decoder.

Hello everybody, I have just bought a MX633 decoder to experiment with stay alive capacitors. I have extensively searched the forum but couldn’t find a clear answer: I am unsure the benefits a 2200uF capacitor can provide. I have wired the capacitor using the dedicated wires on the 633, but I am unable to see any difference in how the locomotive is running. I must say this is my only new and DCC-ready loco and it was already running great, so it probably isn’t the best test bench for a stay alive capacitor. As it makes great contact with the rails, I tried lifting the loco from the track: the motor stops running as soon as it loses contact. Even when idle, the lights turn off as soon as the contact with the rail is lost. This is how I wired the capacitor. I do understand that the capacitor is rather small, but I expected at least a partial revolution of the wheels at low speed or the lights staying on for a little while. I have checked the capacitor voltage with a multimeter and it charges up to 12V when track power is applied. After disconnecting power it takes a few seconds to drop the voltage down to 3V. I am not sure if I need to modify any CV in the decoder: I set #153 to 100 so it should allow for some running and not modified anything regarding DC running. Is everything configured as it should? Should I just try to increase the storage capacity? Thank you! Kind regards.

This is also a good suggestion, however I see three issues: If the loco is reversed while waiting at the signal, turning the left signal to green will allow it to reverse out of the platform (and it can be an issue, as I explained before); It only checks for a signal to be green (as BM1 is hard-wired to the red aspect), so if the loco is waiting for the left signal to be clear and the right signal turns green, the ABC will be disabled and the loco will leave the station against a red signal; It only works if there is a single braking section in the middle of the platform. In longer platforms there are 2 ABC sections (one at each end). If this is the case, having the decoder set to obey the ABC signal in both directions of travel will prevent the train from entring the platform. For example, a train coming from S2 into platform 2 would be stopped by the red signal S2R. Again, I am absolutely nitpicking and I will be happy with a compromise (probably completely removing the BM1 module connected to S3R). Regards, Alessandro

I have considered turning on a single ABC module, however it would not be a perfect solution. For clarity, this is my current layout. Signals are placed on the left side of the track. If a train enters platform 3 from the right, I can turn on the ABC controlled by the signal S3L. However, S3R would not have ABC and so the locomotive could reverse out of the section, possibly against a turnout that is closed and not thrown (look at the one near S2R). Current sensing occupancy detectors are fitted in each section and IR sensors on each turnout. I know I am probably hoping to achieve something that is not possible to do. As you suggested, I can choose to only activate the ABC module needed to stop the train in the direction it's travelling, even if it means that reversing out of the block is allowed. It's a pretty edge case scenario and I think I could live with it. Only thing is, the BM1 module is wired directly to the red lamp of each signal so it would mean that I have to show a green aspect (not stop) even is the path is not clear (for example, S3R must show green even if the double slip turnout is turned against the signal). Again, I can live with it, but it would be better if it can be avoided. Regards, Alessandro

Hi everybody, I am fitting an old DC layout for DCC operations. I am using Zimo decoders and implementing ABC to stop the trains in the station's platform. Some of the platforms are bi-directional, so the trains can move from left to right or vice versa. On both ends of such platforms there is a signal controlling the train movement with a isolated ABC section ahead of it. The right rail is fitted with a DIY BM1 module. I have tested it in a separate track and it works great. My issue is that the platform is not long enough to fit 2 ABC sections. Therefore, both rails are controlled by a signal and when the signals are red, they both get connected to the BM1 module. I tried to make a drawing to better explain myself. This is my current wiring situation. From a "theoretical" point of view it's clear to me that this should not work. As both rails have the DCC wave altered, the DCC signal is again symmetrical (although the amplitude of the wave is smaller) and the decoder can't pick up the ABC signal. As comparison, this is what the wiring looks like for a platform that is longh enough tof it 2 ABC sections (not to scale with the previous drawing). And this should work with no issues. Of course if at least one signal is green ABC works as inteded even in the first situation. But, with both signals red, the train can not pick up an asymmetric DCC signal and so it doesn't stop. So my question: is there any way to make ABC working as intended if the situation is the one in the first drawing? The easiest solution is to set the platform as one-way, but I would prefer not to. If anyone has solved a similar problem in the past I am happy to learn. Thank you, Alessandro

Hi everybody again, sorry for replying so late but I didn't have much time to work on the loco. I found some warm white 5mm LEDs (which come with a resistor already soldered for 12V operation) which fit really well in the previous lamp support. I used some sand paper to make them rough and spread the light more evenly, otherwise the light is only in astraight cone and it would have not diffused well. Due to the lack of space I couldn't run the new wires for the LEDs below or above the PCB (the casing wouldn't fit), so I recycled an unused path in the PCB. The overhead pickup must be disabled, but I have no plans to use it in the future so I do not think it is a problem. I also added a small 3D printed platform to prevent the wires from getting stuck in the transmission. For future reference, this is what the wiring looks like. The locomotive looks great with the lights turned on! Thanks again to everyone! Your help has been incredibly helpful Cheers, Alessandro

I feel incredibly silly… of course every locomotive can have its braking distance configured based on its length! It totally makes sense now. As soon as I will have time I will put the required modifications in place and will update the thread. Thank you for your input, it’s been incredibly helpful. Alessandro

Hello, thank you all for the great suggestions. I made the wrong assumption that the motor was picking up power directly from the chassis (yeah, I realized that it is normal for a DC motor to have 0 Ohm resistance between the two power leads). I moved only the gray wire and now everything is working perfectly. This has been really helpful. Currently my wiring looks the same as the top image, except I only connected the red/black wires to the front bogie directly without connecting track power to the PCB. This is because I am implementing ABC braking on my layout and connecting both the front and rear bogie would mean that the ABC braking signal get picked up by the controller when both bogies are in the isolated section, so when the rear of the locomotive gets in the braking zone. This is not ideal when using constant distance braking, as it would mean that a shorter loco will stop before a longer one. Please tell me if my reasoning is flawed. I have yet to test it on the full layout with turnouts and they might pose a bit of an issue if they can’t transmit power properly. I will let you know how testing goes and if needed I might wire back the rear bogie power leads just as in your first picture. I had considered replacing the light bulbs with LEDs, however I find that creating a new solder joint on the old PCB is extremely hard. Solder just doesn’t stick, and this is also why I only connected the new wires where old solder joints were. Any tips for that? Thank for the LED suggestion; I would love to get brighter lights and LEDs are a great way to do it, but also like the warm feeling of incandescent bulbs. Are yours standard white? I can’t seem to find any warm white single LED, they only come in long strips of 5050 bulbs. Also, when running at high speed the transmission emits a tickling sound which is extremely annoying and gets amplified by the plastic chassis when put in place. I believe it comes from the main transmission axle touching something around the rear spring every rotation. I am sure there are no wires interfering with it and can’t locate it exactly. Be aware that the loco hasn’t been run in about 10 years, so it might just need a bit of stretching. I am yet to deploy the old layout where I can take her for a long spin. Thanks again for the lovely feedback, Alessandro

Hello everybody, I have recently gotten back into railway modelling and I am currently updating my old father's locomotives to DCC. I am running Arduino with DCC++EX base station and JMRI for throttles and such. The first DCC conversion I made is a AE 6/6 from Roco. This was my first hard-wiring of a decoder. I opted for a Zimo MX630. After completing the installation, the lights are working great and can be turned on/off and automatically reverse based on the travel direction. However, the locomotive does not move at all. I will give you a brief explaination on the modifications and tests I did on the locomotive. At first the internal PCB was looking like this: From now on I will as the forward direction of the locomotive to the left of the picture and reverse direction to the right. As you can see, the left rail (based on direction of travel as defined before) is hooked up to the chassis and the other rail is connected to the engine via a V shaped flexibile connector near the coils in the back of the engine. The rotating connector is used to connect to the overhead line if it's used for power instead of the rails. The first thing I did was desoldering all the 4 wires and hooking up the decoder to the front ones, red wire on the right and black one on the left. I then connected the orange wire on the right where the orginal black right rail contact was and the gray wire where the original black left rail contact was. The back rail connections were left disconnected. The locomotive was picked up on my programming track and moved flawlessly, as expected. The lights were working as in the original configuration, so they turnend on alternatively thanks to the diodes, but shut off if the engine stopped. I then proceeded to wire the decoder like this in order to gain control of the lights: So the gray wire is still connected to the chassis (it's how the motor picks up power, I guessed) and the orange one to the same V copper connector. Please note that I removed the coils on the right in order to separate the V connector from the part of the circuit that I used as common light connection and the diodes that controlled the lights. This is where I made a mistake while connecting back the PCB: I missed the motor poles with the V connector and instead of clipping to the motor ends (there are 2 ends to connect to, 0 Ohm resistance between them) it touched the chassis. I did not notice this and tried to move the loco, resulting in the DCC++EX station shutting down as I had created a short circuit between the decoder motor leads. After fixing this mistake, I tried the loco again on my main track. The lights now work perfectly (F0 function, they reverse based on the direction selected on JMRI throttle) but the engine does not move. I tested everything I could with my multimeter. Between the V connector and the chassis there are 0V when the train is stopped, but if I apply throttle it reads +7V or -7V (measured in DC) based on the direction of travel chosen in JMRI. I am now pretty hopeless and do not understand at all why the loco isn't moving. I am concerned the coils and circuitry on the back was required for the DC motor to work and I am unsure if I should put them back in. PWM frequency is at 20kHz. I have a basic understending of circuitry and suspect that the copper coils helped converting the PWM output from the decoder to the motor, as before removing them everything worked just fine. Should I put them back in? Or did I fry the decoder when I shorted the motor outputs? Please note that I did not change any CV after it was working with the decoder and just 2 motor leads, as I described before. I have heard old DC motors do not like high PWM frequency. Should I switch to a lower frequency? If yes, what's best? Thank you to everyone who'll help me, it is much appreciated. Please let me know if you need any other picture or information. Best regards, Alessandro