The brand new Atlas Genset is reported in various forums to have no space for a H0 size decoder. A proposed circumvention is the installation of a N-size decoder.
That is true but can be fixed in three minutes.
The DC version of the Genset has two huge 100 10-9 Farad capacitors parallel to the motor stored under the short hood weight. A first test with temporary plugged in decoders (Zimo MX630 and ESU Lokpilot) gave horrible results at low speed. That is quite normal with that huge capacitors exceeding everything I have ever seen for RFI purposed.
Simply remove the capacitors and install the decoder UNDER the short hood weight at the place the capacitors have been installed before.
No more space problems and the Atlas engine runs as smooth like all the Atlas engines do[:D]
Those big caps are in there and connected on a plain DC loco?
I think we found the problem with the Tech 6. I’ll bet that loco also doesn;t run well on address 00 on those DCC systems that support that. Those caps will be a dead short. I doubt the loco is ‘burning up’ but those caps might be if subject to other than plain DC.
ANd like the smaller caps used for RF supression in Bachmann locos, they will cause havoc with the high frequency PWM drive from a good decoder.
Yes, they are directly wired to the same terminals on the PCB where the cables from the motor are connected to. The DCC decoder was not very pleased with that capacity load parallel to the motor. I wonder the N-decoders did not run havoc too.
That is the usual function of a capacitor at that place. But you are using a much smaller capacitor to “catch” the higher frequencies only. This large capacitors do massive interfere with the motor control of the decoder. It is in some sense like a shortcut for the motor management of the decoder.
ps. I saw your posting in the Atlas forum. I can’t post there because my multiple registrations have never been answered by the admins over there.
Someone on the Atlas forum followed your instructions, he removed those caps and installed a DH123 in the enclosure. He is very please with the way his Genset now operates.
I will let him know I transmitted his thanks to you.
What is the modification that is talked about here? Removing the caps?
I just picked one up with a ditigrax DH165IP decoder. It ran very smoothe and silent in the LHS. I noticed that it has ditchlights which is rare for an atlas. I’ll open it up and take pics later. Maybe even make use of the 6 function decoder with those ditch lights.
I have got two BNSF engines for DC only. They have the large capacitors installed. Both engines got a first test run with a very old DC supply. They ran smooth without problems.
My changes was triggered by reports of to little space to install H0 decoders. I did simply plug one into the 8-pin plug and tried to put the shell back. That did not work due to limited space. Next was a test run with a half open shell. I noticed very bad run at low speed on both engines with Zimo and ESU DCC decoders.
I ignored that problem and started search for a place for the decoder to install. I thought about some milling on the front weight and removed it for further analysis. That was the moment when I discovered the capacitors directly connected to the motor wires at the PCB.
Removal of the capacitors solved both problems. I got space for the decoder and the motor runs very smooth with the decoder.
I have no idea what for those capacitors have been installed with the long wires. It makes no sense to do that for any RFI reasons. You would take capacitors in the Nano Farad range and connect them as close as possible to the motor.
However, because the engine ran fine with an very old DC supply it would be overstated to say Atlas failed to deliver a DC version. Furthermore might be DCC decoders on the market that have no problems with the huge capacitors. I am somewhat carefully to blame Atlas because I did not fully understand what happened. It is just not very unusual.
May be someone with a functional Userid in the Atlas forum can a
I’d love to see the reasoning for those caps. It CAN’T be RF supression, they are far too big to block the frequencies generally involved. My guess it it’s some attempt at doing a keep alive for DC.
It’s might work as-is for a decoder that doesn’t have a high frequency drive, the Bachmanns with the little caps work fine when the decoder isn’t a high frequency one, like the ones they provide for the DCC On-board locos. But those caps are so large they might also interfere with the non-silent decoders as well.
Or maybe the whole thing is a mixup by the Chinese factory - the specs said .01uF capacitors (liek Bachmann uses) and the factory thought they said .01F.
I didn’t note the value of the caps when I removed mine last night, but they were wired in series.
My thought was that they may be filtering the pulsed DC on the motor side for use with DCC systems that use Address 0 for DC. By attempting to filter maybe the value of the caps is tuned to the MRC frequency and looking like a short.
Back to back electrolytic caps make a non-polarized capacitor - since electrolytics are polarity sensitive this is important since they are across the motor terminals and can be either polarity, depending on if you are going forward or reverse.
Depending on just how big they are, there will be an inrush when power is first applied, and when changing directions, although a normal DC pack should absorb this with no problems. On anything with a pulse form of power, it will appear as more or less of a short depending on the pulse frequency. For example, the capacitor supplied with the old Soundtraxx DSX and DSD decoders that went in series with the speaker, blocked low frquencies but allowed higher ones to pass. Parallel to the motor, the capacitors will appear ‘open’ to DC (high resistence) and more of a short (low resistence) as the frequency goes up. The exact resonant frequency is determined by the capacitor value as well as the motor inductance. The smaller the capcitor, the higher the frequency - thus the small caps on Bachmann locos to absorb RF frequencies that could interfere with your TV or radio. These large ones seem to be more of a keep alive or power filtering device.
IN any case, a ‘silent’ DCC decoder uses high frequency PWM so that any ‘singing’ in the motor is above normal human hearing range. The capcitors will interfere with this. They can also block part or all of the BEMF from the motor which also doesn’t help. Not all decoders are the same, if the ‘silent drive’ frequency is near the resonant frequency of the capacitor/motor circuit, or an integer multiple of it, it will be far worse than if the drive frequency was well offset from an even multiple of the resonant frequency. Some decoders, aprticular the European ones, have adjustments for this frequency, to better fine tune the decoder to a particular motor. Others pretty much have just an on/off for the silent running feature.
I’ll wait to see what reasons Atlas gives for incorporating these
The way the capacitors are wired is that the 2 negative leads are soldered together with the 2 positive leads going to the motor. Maybe some have the polarity of one of them wrong and causing the poof. [:-^]
